For decades, cycling was ruled by numbers like threshold power, performance management charts. But what if I told you those metrics might actually be holding you back and there's a better, newer way to do it. Today, I'm sitting down with one of the most exciting young coaches in the world, Alex Welborn, who's pioneering cutting-edge research into W Bal and critical power.
Don't worry if you don't know what those are, you will after this conversation. Because together we're going to explore how this new science is rewriting the rules of training. Why it matters more than you think and how you can use this to build a perfect winter training plan.
By the end, you'll see why the old coaching world is fading and the athletes who don't adapt will soon be left behind. Welcome to the podcast, Alex Welborn. Alex, welcome to the Road Man podcast.
No, thanks for having me. I mean, early, but thank you. >> Yeah, thanks for thanks for flying across.
Always better in person. We're going to try and bridge the world between old coaching philosophy and new coaching philosophy today. And because I'm old and I've been ingrained and WKO.
Like I remember one of my first coaches like emailing on my SRM files because TrainingPeaks didn't even exist as a cloud-based solution. And TrainingPeaks has been something I've relied on. And not just the platform, but the vocabulary and the framing around how I talk about intensity for a long time.
So this I'm hoping this conversation that brings me into the new world of all the crazy new terms, critical power, W Bal. But I want to start in the basic and then advance as we go to try and build up our understanding. So even the time of year we're at now, everyone is starting to think about winter training plans.
And the first book I ever read, and I've been looking up to have him on the podcast a couple of times, and it was a real fanboy moment when I had him on the podcast, was Joe Friel. I was like the first ever training book I had. I think that I think I know which I think Yeah, yeah.
>> Cyclist's Training Bible. So good. Yeah.
I had a copy of it. I don't even know if it was my copy or like between a bunch of friends. And we like every page highlighted, scribbled out.
And some of the stuff in it's still brilliant. Like I love his morning check-in still. Cuz like even with all the data like your heart rate variability, sleep scores, they're all proxies I think for how you feel.
Friel really got to the essence of how you feel with the morning check-in. Like rate your muscle soreness, rate your mood, rate your willingness to train this morning. I love all that stuff.
But the Don't worry, I'm getting to the question here. So let's start with a winter training plan and taking some of his assumptions down and we'll see if you think they're still valid. So when we talked about building winter training plans back in the day, we used to build it in base with a one-week, two-week build on week one, third week build on week two, fourth week decompression week.
Then we move into the second block and he built it in this block phase. Has thinking on periodization changed or how do you approach building winter training plans now? I think a lot of it is depends on the athlete that you have.
I think depends on their constraints within that. And with everything you've got tools in the toolbox. I don't tend to fit like riders so let's say reverse periodization or periodization.
Or polarize, sorry. I think periodization is exactly what we're doing with training. It's not like we're going to take this approach with this athlete.
It's like, right, what tool do we think is going to best fit them? So we might take slightly different approaches. There might be a big holiday in winter, so we might build up a little bit more aggressively knowing that we're going to have a lot more time to recover.
But I think the basics of principles of training are still there. We still need to build. Needs to be progressive.
Needs to have overload and then recovery. It's kind of still very set in this in this day and age. It's just yeah, doing it the most appropriate for the athlete.
Like you might in winter I might do slightly longer blocks because the intensity is not as much that they can cope it. Providing that they've got good feedback within that. I know what's going on.
It's like, right. Cuz training plans are dynamic in their very nature. Like actually how are we feeling?
You know, and then go We get to the end of block like last few days, I'm feeling a bit tired. It's like, right, that's pretty much exactly where we want it. Whereas actually we might get to still feeling okay and then we get to the next week.
It's like, right, we'll bring the recovery week forwards for that. So I think yeah, the basics are still pretty much you can't reinvent the wheel, I think. Yeah, that's actually a interesting direction to go with it because I had Dan Lorang on the podcast and he was talking about athlete comes in the door, very first thing, whether it's Joe or Jane who's working a full-time job and balancing family or it's Jan Frodeno, the triathlon goat, or one of the Red Bull guys, Roglic, the same thing happens.
He sits down with a blank calendar. Doesn't put any races in this. Puts life events that have to that are can't miss these life events.
Obviously for a world tour rider like, you know, Primož, those life events are they're going to be quite minimum impact or disruption. It's not like he could say, "Hey, I'm at a friend's wedding the week of the Tour de France." That's not really going to work.
There's contractual obligations around that. But I think that gets to be a more interesting starting point when you think about your amateur athletes who's because now you can Friel's periodization model for me was brilliant as a student when I had absolutely no obligations. You know, week two builds on week one, week three builds on week two.
We're going to build up block by block. Yeah. But then even when you got into exam time, you're like, "Oh, I'm meant to be doing my big block here, but exams are on.
This doesn't really work." But if you start with Lorang's model of you put your exams in, you put your family holidays in, your trip with your girlfriend in, now it starts to kind of go, "Well, it doesn't make sense to try and do my most intense week when I have my biggest week of outside life stress. Maybe it makes more sense to have that as a recovery week.
" Yeah, I think that's that just goes with the kind of what I was saying earlier with like you get to know the athlete and what's the best approach for them. Like I work with quite a few students as well. So we have some times where actually there's minimal amount of lectures.
So let's make the most of this time. Whereas then like you said, they might go into exams. Well, there's actually there's no point adding additional stress.
Like one a couple of riders I've had where I've tried to explain it's like a each day you have a capacity for stress. So you've got stuff that goes within your life, then you've got your training and you want to typically have a little bit of space left over because then that's kind of your safe space. So if work gets a little bit longer or kind of you push on and do a couple of hero efforts, which you shouldn't, but you do those in the efforts.
I love that expression, hero efforts. Hero efforts, yeah. I'm guilty of that one.
We we all have you know, you got your first like five by five and you're not done five in a while. So you know what? I'm going to do my five-minute all-out effort today and then struggle and then go to my coach, "I can't I'm not hitting my sets.
" And it's like, "Yeah, because you did a five-minute max and then that's why." Like And I think it's just it's treating the athlete as a person. Like I said, what stuff is going to get in the way?
It's not even necessarily what stuff you're going to be doing. Like is there a social ride that you really that's really important to you? Because I think cycling is such a social sport and actively kind of encouraged it.
Like I got one rider in the States where they have a chat lap. They call chat laps on a Friday around one of the parks in New York. So that's pretty much integrated into the training plan.
If that gets a bit too hard, then we adjust based on that. It's that continuous input from the athlete. >> So through that lens where you're talking about you're looking at stress as a broader concept to include life stress, family stress, exam stress and you're throwing training stress in there or maybe even maybe you'd bracket in nutritional stress if someone's eating terrible food into that same.
How do you think about approaching recovery days then through that lens that we normally use as a time-crunched athlete where if I have a busy work day, I'll often use that as my recovery day. But cortisol and stress is still going to be quite high from non-physical activities. Is that really the best place to put a recovery day?
No, I think it's I think everyone's been guilty of this. Like they see recovery day as a day that they can go and do anything they want. So it might be you go out for the long walk with the family.
That's three, four hours and you go shopping and then you go out for dinner. You've been on your feet all day and you become ruined the next day. Like I used to when I used to lived in the South Coast, I used to go paddle boarding on my rest day thinking, "Oh yeah.
" And then the next day I'd start training, my legs were absolutely shot. I'm like, "Why?" And obviously paddle boarding and obviously the cyclist in me at the time was a bit competitive.
I was like, "Right, let's see how fast I can go within this." And I think really it's you want to reduce that entire stress capacity for the day, but it's been most optimal for training as well. Like Saturday, let's say, would probably be a great day.
You could probably do very little, but actually it's a great day to get four, five hours in. So it's the stress on the single day basis, but the weekday basis and then the monthly basis of that kind of overall consistent picture. Like what's the bigger picture within it all?
Do we have a way of quantifying the stress load on our bodies? Like if if we take a the goal of the adaptation I'm looking for today on a Friday is to recover. But now I look at my diary.
Okay, the steps in the paddle boarding example, I think you could kind of intuitively go, "Okay, my my median amount of steps is 7,000 a day. Maybe not a good idea to do 14,000 steps, 100% extra on my recovery day." That seems kind of intuitive.
But when we think of more passive stress like I've extra sales calls. I've, you know, more stressful conversations to have around work. At what point do you would you no longer say this is an environment to get the adaptation we're looking for today, which is recovery?
Yeah, I think it's more a case of, right, there's going to be some constraints that we can't like you're not really going to be able to not do those calls. You're still going to have to go to work within those things. So, it helps us acknowledge then what other things can we look at?
Can I actually get a bed a little bit earlier the night before? Can I focus on my nutrition leading up to it? Like, we know the constraints, what else can we do to factor in on those things to make an extra conscious effort within that?
Cuz you might find actually after that stressful day, you still want to go and ride and train cuz that's and almost can reduce the stressness of your calls even though you put your stress capacity up from your training session. It's that fine balance between the athlete actually acknowledging what's going on. Like, do you know that it's stressful or do you think it's not?
And then how are you responding within that to training within that and your motivation? It's kind of the whole thing is a picture. There's no kind of like, oh this one variable will allow us to do.
It just helps us make that informed decision with I think that's where a good coach and a good athlete relationship helps you just be black and white and be honest with that. Like, mate, stressful day. Like, I'm stressing about training.
Like, half of my job is a to manage stress for athletes. Like, actually no, let's we can make it up. We can change a few things so you're still going to get the overall volume.
Just it'll be the same over the month just not in the day or the week. And this is where coaching comes in. This is the difference between having a coach that understands you as a person and being coached by an AI.
Being building your training plan on chat GPT. It doesn't have that context. Even when I'll chat with athletes now or Sarah culture, I was like, I coach a very small number of people but I like to get to know them intimately.
I was like, I want to know your wife's name. I want to I want to know what she does. I want to know if you're, you know, having domestic problems at home.
I want to know if you have to drop the kids to swim in the morning because all this feeds into your frame of mind going to a session that evening. Yeah, it's you it's getting to know your athletes. Not just a number.
It's a person at the end of it. Like, you want to know I've had, you know, pictures of wedding days when kids have been born. Like, it's as a coach you have such an influence on their day-to-day life.
>> More than a friend. Yeah, to a certain extent it is but like they have that trust within you and I think this is where AI will never replace a coach in that respect. Like, I don't ask, you know, my riders, oh what do you think of the session?
Like, how did you feel? How's it going? How's things?
Like, I can see the data. I know if you completed the session. I know what watts you did but I want to know what's the story behind those numbers.
Yeah, the context to the session. You know, like I might see a session, let's say 4 by 10s, 300 watts. You know, one day you've hit it, the other day you've hit it.
But actually one you had an absolute terrible day. Like, you know, really bad day at work. You've had some family issues, etc.
And one day you've hit it really well. But that would be in two different completely different contexts. You might need to adjust based on those two things.
It's just getting to know the athlete that's turning those pedals. So, building up this idea again of building the training plan for the listener. So, we've they've got down their unmovable objectives for the season.
Now we're kind of getting in if we call that maybe the macro plan and we're zooming in a little bit to the micro plan week to week and we're trying to figure out, okay, I've I've 10 hours or 12 hours a week available to dedicate to the bike. Would you always dedicate 100% of their available training hours to the bike or would you peel some away and say, hey, you say you can train 12 hours but two hours should be admin of preparing meals, preparing kit or is it always full allocation? I think I tend to work off like a little triangle.
Like, what they need to do. Like, what they can do and what they will do. Let's say you've got two hours after work.
Most riders say they've got the two hours but there's them doing faff time. So, you've got half an hour there. The shower afterwards, the food within that.
So, really they've got an hour. So, I'll kind of make sure that that there's a quality hour within that. And if they do have the time, we'll push a little bit of volume.
I kind of let athletes sometimes dictate the volume a little bit from sessions. So, we get the core quality in. And then if they've got the time, we can either pad out a bit more recovery if that's suitable for the session or add a bit more volume providing they're not adding on the intensity.
They let me know and then I can make sure that fits within and adjust accordingly within that. I think for some if you're really time crunched, you might as well make use of that if you've got eight hours, you there's no point building four hours, five hours, six hours. Rest six, seven, eight.
Like, finding they can handle those eight hours >> Yeah. then use it. Yeah.
Um how do you think about distributing that eight hours or 12 hours or whatever they have available? If we think about, you know, we normally have three levers that we can pull. We can pull intensity, we can pull frequency or we can pull duration.
How do you think about arranging those chairs around the table? I think it depends on their training history in terms of like training age. So, if they're very new to the sport, well, we might as well just push with volume cuz that's quite an easy thing that they can do.
And as they kind of progress, we'll kind of manipulate those in the way that suits them best and how they manage those sessions within that. I think I don't try to build too much volume and intensity at the same time depending on the athlete. So, some I might build every other week within terms of let's let's say 3 by 10s for nice example.
It might be every other week we do two minutes rather than two minutes on or push power up within that. I think it's kind of getting an idea of what they've done before, the feedback from the session. So, it's not one rule fits everyone.
Like, like let's learn the athlete. What can they do and how much those non-negotiables and work impact those as well. One of the the models that a lot of people will default to is Stephen Seiler talks about 80/20 distribution of 80% of your stuff.
Pareto obviously more famously than Stephen Seiler talked about this. 80% of your stuff should be quite easy. We can debate what that definition of quite easy is in a little bit.
20% of it quite hard. Is there a minimum effective dose? Like, if I've six hours a week available to train, should I still be looking at this 80/20 or do I need to dial up the intensity lever a little bit more?
I think it can vary between athletes. Your genetics will come into play a little bit. Your kind of life stresses as well.
It's kind of what are your training goals within that? What do you want at the end of the year within that? >> I'm trying to get with that question and phrase it brings is like, where's the guard rails?
Where's the point at which you're like, no one should be doing that much intensity. No one should be doing that little intensity. Again, I think it's with how the athlete's responding to it.
I think there's definitely a case of too much intensity. Like, you know, if you've got six sessions a week, it shouldn't be six hard sessions. You know, I think two to three hard interval sessions per week can fit most people and then the volume kind of is moved around that.
With kind of more full-time athletes, it does naturally kind of go to a polarized approach because they might be doing 20, 30 hours. So, if you're sticking, you know, six hours or whatever like of high intensity, they're going to be pretty cooked. >> That's a lot of intensity.
Yeah, so it's that kind of balance within that I would say. Yeah, cuz you do see and I think we've I'm always cautious of and I don't want to hammer on any particular platforms. So, we're trying a few together like Training Peaks with My Whoosh where they have the workout plans.
You got to be cautious in the motivation I think of these platforms because they're not a cycling coach. They're a cycling business. And a part of the cycling business is understanding business metrics of lifetime value of a customer, churn rates.
So, when they're looking at metrics like churn rate, how do we keep someone on the platform? It's very different to if you're a coach and me and you're like, hey, you've got nine hours available this weekend. Go and do nine hours of zone two.
If I log into my TrainerRoad account and it's like I've nine hours this weekend, go and do nine hours of zone two. There's a higher temptation to just cancel that account and go, I didn't need TrainerRoad to have me do nine hours of zone two. So, they build in what I call almost the Netflix effect where they feel like they need to entertain you in the session.
There's a lot of like, let's change cadence here. Let's >> Entertainment sessions. That's what I call them.
Like, for some riders I do put entertainment sessions. I call them hiding the broccoli on certain things. Cuz, you know, I've worked with quite a few riders for quite a few years and there's only so many ways you can entertain zone three.
You know, the ultimate like 3 by 10s is quite a hard session to disguise multiple ways. You can add in nits and bits of cadence changes in there and that's just then for a little bit of interest and kind of performance within that. But yeah, there's certainly a market in that sort of stuff.
Entertainment sessions where it's like 4 and 1/2 minutes here, 30 seconds here, then 2 minutes 15 seconds there. I'm like, wait. I I like I like simple.
Simple's nice. Yeah, I remember Michael Barry coached me. I was riding for Team Sky and I was really excited.
Michael Barry's coaching me. It's really cool. I was trying to, you know, make my way in the world of cycling and you're looking at someone at the top level and you're thinking he they must be doing crazy stuff cuz you got to think this is back 2014 I think he was coaching me.
So, this is back cycling 2014 is pre-GCN almost. Availability of information hadn't been quite democratized yet. Nobody really knew what sessions people were doing and we these whispers around Sky.
You know, they've got this pineapple juice. It's like a post like Matt Hayman was telling me they called it the the post-race rehydration solution or something. He's like, it was pineapple juice and water.
But because it had post-race rehydration solution, you thought you were getting this magical stuff. So, there was a lot of that around Team Sky. I think Tim Kerrison was coming across and swimming into it.
And so, there was a lot of mysticism around the type of sessions they were in. And I remember getting my it was an Excel document at the time with sessions and it was just like six hours, look over hedges. Five hours, look over hedges.
Seven hours, look over hedges. And I was like, okay, I guess. I think most people would probably if they have this like fantasy in view of like world tour cycling and coaching, I think most people would be quite disappointed in the plan that they got.
It would be you know, I think that like take Joe Laverick for example. He's put on on public some of the training that I've been giving him over the times. And it was like uh Monday, rest day or cafe easy cafe spin like a proper cafe spin like shorts, shoes probably just roll down to the local cafe.
Tuesday, efforts 3 by 10s 4 hours. Wednesday, 4 hours. Thursday, torque efforts.
Friday, 2 hours strength session. Saturday, 5 hours. Sunday, 5 hours.
Like zone one zone one. And this is his favorite, he loves it. You know, that type of like uninterrupted training, but it's like oh is that it?
Yeah. >> And would an athlete like Joe who's, you know, been on the podcast, he's a good mate as well. He's obviously living out in Girona.
With something like a 5-hour endurance ride, how strict are you on your Joe loves to start his sessions going up hills. Mhm. You know.
Harder ride or else super super easy, you're at least top end of zone two. Going up, but how strict are you on kind of creep in the zone into a different zone? I would say it's one of the first things I normally do with new riders I work with.
I'll give them a like a recovery ride and a zone one ride. I was able to quickly gauge one So, how do you differentiate between recovery and zone one? I think it's Maybe this is helpful for the rest of the podcast I think we definitely need to ground some of the terms because, you know, we talked off air.
You talked to different coaches. You could be talking about two totally different things when you didn't say zone one or zone two or zone recovery. Yeah.
So, recovery I try and frame as the aim of that is to have the lowest average heart rate and the lowest average power. Because if you put a target, let's say zone one, cyclists are renowned for going right top of the target. >> Yeah.
So, I try and frame it slightly differently cuz the idea is, you know, I try and put a bit of description of what the session aims are. And this one is literally just to keep the legs turning over, keep it nice and short. And why?
So, this is a question I maybe wouldn't have asked a few years ago, but Stephen Seiler kind of stuck this in my head. He's like, "What's the adaptation you're looking for? Mhm.
Recovery. What are you doing? Training.
It's like active recovery, it's like it's an oxymoron. So, I think with that like for some riders they just like to turn the legs over cuz obviously like you don't tend to have a rest day on the tour. It just stops them becoming like stiff, stale, blocks, flat, to kind of these words associated with it, but minimize any extra stress from that session.
Like Mhm. Yeah, calling it a recovery session is probably completely incorrect cuz you're not really recovering cuz technically you're adding Yeah. stress to it.
You're adding less stress, but you're still Yeah, and it's trying to keep that as And otherwise you might be like want to go for a walk and that and that overall might add more stress. So, if they're going to do something, why not control it as much as we can and minimize it and see how well they respect the recovery cuz that can be quite an area to work on with athletes. Like it might not be the training quality, but actually are they unoptimizing their recovery?
Unoptimizing, yeah, unoptimizing their recovery. By just doing Double negative Double negative. By doing things they just probably shouldn't do.
Like you know, I look on a recovery ride I'll look at the max heart rate and max power rather than the average like see I want to I don't really want to see anything above threshold at all. Yeah. Like I want to just see nice steady fluid heart rate nice and low.
Like I used to see if I can do my average heart rate about 105 110 beats on a recovery ride and I'd find the closest coffee shop go to it and literally roll back. Yeah, I like the idea on a recovery ride of if we break down the different requirements for something you know for cycling that aerodynamic is one of the requirements. So, it's like maybe now can we ride a TT bike or can we stay in a super aero position for the recovery ride when you're doing 100 watts.
Mhm. But all of a sudden I'm I'm working on something else that's important for the overall picture. I like that kind of idea.
I had an old teammate who used to do his recovery rides and he would say that if he put enough power down to engage his cleat into the pedal he was doing too much power. So, he'd try to hover on the cleat, but he'd be on the TT position locked in on this bike literally riding at 60 70 watts. Like opportunity to kind of do something, be in position, get familiar with that adding stress.
Like I normally say if you think you're going too hard, you are. Yeah. Like if you have to think am I going to Yeah, you are.
Like So, are you using the Coggan 5 model for zones? I'll tend to use a mixed model. So, depending on what I have available with the athlete, I'll normally use kind of a bit size with an LT1 kind of top end of zone two, then kind of within the critical power >> lactates?
Yeah, where I can we'll test lactates. I'll get them in to do parts of the research studies that I've done. We have a we have a nice calculator that does it's a bit there it's about nine plus or minus 9% of the power output to give a rough estimate of where we think LT1 will be, which is just based on a load of collective lab data that I've taken about 60 athletes.
So, it's a good starting point if we don't have anything from that. Kind of then see we'll use critical power and then we'll kind of use CP up towards maximal aerobic power. I'm not a big fan of the term, but it's we define it as the peak 60 seconds as a ramp test.
>> Yeah, I've done those tests, they're pretty miserable. Yeah, they're pretty horrible. Um but it's not Some people call it VO2 max power, but in reality VO2 max power doesn't exist.
It's anytime you go above critical power you're in what we call the severe intensity domain. The problem with the MAP test I find is the steps up are too big. Actually on a test in John Wakefield's lab in Girona last year, much nicer because it goes up like 1 watt.
That's what we do. Beautiful cuz the old steps were like 25 30 watts. >> Horrible.
Like we do absolutely fine, I'm totally cooked. We do 12 watts every 5 seconds. 12 watts every 5 seconds.
Still a big enough jump. >> watts every 5 watts every 12 seconds, sorry. >> Okay, yeah.
So, 12 watts every 4 seconds like that. That's That's even worse. >> in 3 minutes.
Yeah, and that that might explain some of the lab data. Whoops. Um We should So, then and then kind of use W prime, so your work capacity above CP to kind of define the top end zones and then Pmax is the top part of it.
>> Okay, and if you're confused what all them terms are like I am, we're going to get into really defining those because that is what we're talking about the the new age of coaching. So before we jump into that new age, let's talk about the old toolkit that we used to have. Our old toolkit was performance management chart or one of the main things.
So, performance management chart for anyone that doesn't know, it was totally satisfying to watch your little blue line increase blue line. The names always baffled me with performance management chart, why we called them chronic training load instead of fitness or why we called it ATL acute training load instead of fatigue or training stress balance. So, with ATL CTL TSB, what did these try to capture and how did coaches use them?
I think the original concept was it it has moved now being called like this represents your fitness score. As soon as you put that connotation of score, you associate the higher the better. Yeah, it's like a game of fights.
So, the idea is it's meant to I think more appropriate is it's load. Like what that athlete's load is currently under. And obviously >> Volume or volume and intensity when you say load?
Or you don't know? We don't really know within that cuz it's very weighted within time. I think it was Tune Van's up Tune Van up's PhD they looked at kind of quantifying quantifying training load within professional cyclists.
So, we'll technically encapsulate both of those elements cuz obviously the way the maths works with it taken into account like normalized power, your training stress score, then your weighted 42-day rolling average. >> average. So, it kind of encapsulates all of that, but doesn't give a value towards which one that's weighted to.
Cuz this argument I've had with a couple of coached athletes, it's just trying I try to like to inform them as much as possible about the metrics that they're going to see cuz I feel as as a coach, part of my job is to educate. And like you can have two riders doing a weekly TSS of I don't know 700 and eventually gets to a CTL of 80. And they both of them are the same in terms of that score, but one could have done it through just volume, one could have done it through volume and intensity or another one could have done it through just intensity.
The training stimulus from all those sessions will be different. Yeah. The overall adaptations from that will be different, yet they're all at 80.
So, according to the maths, they're at the same fitness score. So, I see the problem with this is and really nice frame to think about this I find is if you differentiate between upstream and downstream stuff. Here you've taken an up made one upstream assumption and it's contaminated a bunch of downstream stuff.
And by that I mean if you've wrongly assumed the calculation on training stress score and now training stress score is rolled into a 42-day total which becomes your fitness score. But if the underlying assumption of how we calculate stress is wrong, now the entire performance management is wrong. And for me not separating volume and intensity at a base level on session renders that training stress score a little bit pointless.
Because I don't know how I got that training stress score. Like if I go out and do I don't know 3 hours at, you know, zone one zone two and I'm accumulating 95 stress points or so for that. Mhm.
Or I do a full gas aero long TT I'm getting roughly the same stress score for two vastly different efforts. Like you wake up the next day after doing a 25-mile TT versus wake up the next day after doing a cruisy 3 hours base ride. Two very different feelings.
Yeah, it's like when I was racing mountain bikes I might have a stress score of let's say 140, but I could do that relatively comfortably over a weekend on an endurance ride. But one, I'd feel like I've been run over by a bus the next day, probably the next day as well. But the other one, I could do the same again the next day, which has always been my I think it's just a lot of it is about being aware of the limitations of that value and kind of not inferring too much from that.
Like just seeing it truly for what it is. It's just it's a point-based based on some maths, which we don't necessarily means it represents your current fitness. I mean, the association that the more you do, the better you'll get is somewhat true to a certain point.
And there's obviously going to be a point where that might not It might be too much, but >> Well, have you ever looked at an ultra endurance athletes performance management chart? If you were to look at that with that assumption that this is a fitness score, they're going to win the tour. You look at Lea Wilcox's performance management chart.
She's pure cess to win the Tour de France this year. It's like I've never seen CTLs as high as them ultra endurance people. And it's even more funny when like Chris, I think you had Chris on the show.
>> Yeah, Chris Great Ladd. You're coaching Chris. Yeah.
Yeah. >> Which is just weird small world cycling stuff. So, the way that even the maths works, so let's say he's done his uh Badlands, I think it's like 39 hours is on the TrainingPeaks or vector file.
>> Still wearing the t-shirt. But it only count that score for that day. Yeah, he's ridden over 3 days.
So, it actually say he's not scored any points on the day after and the next day. Ah, cuz he kept it all as one file. Yeah.
See, that's what happens when you go fast. I never had that problem. I broke down into five different files over five different days.
But Sarah, she tried to keep it all as one file and then at night just powered didn't stop the file, just turned it off the bike computer. And you actually couldn't write this. She lost the file with like 8k to go.
Oh, no. And she had this like illusion of this perfect Strava 818 km loop and all the kudos she was going to get. It's like I got to backfire badly on you.
So, even like a constraint within that will start to then impact the maths. I think it can be used well in certain situations where if you want to compare the build-up of the load of the athlete in the similar sort of condition. So, like your CTL is yours.
It should only be compared to you. >> tool. So, like let's say yours and mine, because of like I said about that stress capacity, it will be a created in a different situation under different circumstances.
So, it's not fair let's say, you know, mine's 120 and yours is 120, could be completely different in how it's been measured in terms of that load. So, it's that situation where we might know with a guy that's, you know, working 45 hours a week, we see illness start to creep in when it's, you know, ramp rate's higher than X or load starts to hovering too much time above 90. >> That's a good application for it, the ramp rates.
>> I think that's where it can be used nicely, but it's all about understanding the limitations of that framework and why are we can't infer certain things like fitness. Well, you know, when I was racing mountain bike, my race was 90 minutes. I didn't have to have a huge CTL.
I was hitting PBs. When I took time off and was racing road, my CT was higher, but certain power ranges it wasn't as high because I didn't need to train that. So, it was it's context-specific for the athlete.
Two of the limitations I found early were riding for a US team with a lot of international travel. I would be cooked from a 2-day trip, airports, cars, you know, low budget teams, you're driving a lot, never getting the direct flight, long bus transfers. You've had 2 days or 3 days of zero training stimulus.
So, your freshness score comes up and your fatigue score drops. So, I'm looking at it and going like I've nice and come up to a zero training stress balance and theoretically I should be peaking here. Mhm.
I can't even turn the pedals I'm so fatigued. That was one limitation. And the next one I came across when I started coaching athletes was the plateau you hit when an athlete has a fixed amount of hours they can ride every week, exact same schedule.
You hit a certain point where the feedback you start getting off the athlete is I'm not getting any better because they're judging better off is the blue line still going up. It's like, well, you you don't have any more training Yeah. I think when I kind of work with an athlete, we kind of try to put down what we think success looks like and then frame everything towards that.
Like, you know, I want to be stronger, improve my 5-minute power, improve my 20-minute power, you know, whatever that kind of might be and kind of then direct the feedback to those improvements based on those to kind of take them away. Oh, what about my CTL? Oh, you don't need to worry about that.
Like, you know, if the session's building up, you're hitting the targets, we're progressing them, like you're hitting everything we're expecting you to. Ignore it. And I think and some will kind of believe or buy into that quite quickly.
Some take a little bit longer and it's just having that patience to educate them, make them feel like, you know, you're listening to them and even give them a bit of science within that. Like explain it in nice simple terms, not hiding behind, you know, it's based off the Eric Bannister model. They've got this K1 and K2 within the mathematics.
>> What's What's the Eric Bannister model? I think that's where it was like originally developed from. I think I can't remember the year, but there was a like it's an impulse-based model where um it's in team sports how quickly they were coming back from injury, whether or not they're more likely to be susceptible to injury again based on that load.
>> Interesting, didn't know that. Yeah, I think that's I want to say 1975, but don't quote me that at all, but it's >> Someone will fact-check in the comments down below. >> I'm sure.
Oh, well, I'll fact-check myself onto it and then um I'll just say all the numbers across the podcast. So, hopefully then we can just voice over it and I'll be correct within that. But yeah, I think the original math, I think it is even on the TrainingPeaks article like Eric Bannister where it kind of comes from and then I think Andy Coggan kind of changed into the framework to kind of work within the training platform that we know today as the performance management chart.
Yeah. And even if you reverse how we talk about training stress, I guess the assumption before that, which maybe is the fundamental flawed assumption here is that a threshold test is the best way to set your zones. Because if you get the zone wrong, now the training stress is wrong for the day and that's probably more upstream again than even what I talked about training stress.
Mhm. Did you have a moment when you realized there's a ceiling to setting zones off threshold and this performance management chart? How do you mean exactly?
>> Like a ceiling where you're like there's an inbuilt limitation with this model we have for prescribing sessions, for pacing, for managing fatigue, where like there could be a better way to do this. We just haven't quite figured it out yet. Yeah, I think that goes back again to understanding the limitations of what it is, how it's been testing.
What is it actually that we are trying to test? Like the old typical way was you do a 20-minute test, take 95% of that value and do that. But then the actual original threshold test, you do a 5-minute blowout first.
So, a lot of the research has tried to look at, I'm sure we'll get onto it, onto critical power, but sometimes it misses out this original one, which was to deplete your anaerobic side of things. So, it's more of an aerobic effort for your 20-minute. And then those zones based on that.
But now we know that there's a lot more variance depending on the profile of the phenotype of rider you are. So, that can then influence the zones, which will then influence the training stress scores that you get and it influences CTL. So, it depends on what performance questions you were going to ask based on that information to see and acknowledge the limitations around that approach.
Is there anything we can then do better? How What can we do to change any of that to kind of then back up what we're doing? And then are we happy to progress with that decision?
That's interesting. So, talk to me a little bit about that sort of phenotype of rider. They were having more beneficial or less beneficial outcomes.
So, if >> old protocol? Let's say if you are more of a I'm going to say a W prime-based rider. So, you've got more work capacity above threshold for your shorter power duration.
Some might say more rouleur puncheur kind of anaerobic style rider. I don't I'm not big An all-around athlete type. Yeah.
And so, actually that 20-minute effort is going to be a lot more influenced through that. I think there was a presentation at Science and Sport a couple years ago where Kevin Cain was like CP versus FTP. And actually they put a nice chart what percentage of 20-minute power your CP was.
Yeah, if it was more W prime-based, it would influence that model more to be more reliant on that. So, your 20-minute power was less representing CP if you're more W prime-dependent. Okay, so everyone's probably rightly confused right now.
So, let's start at getting in reverting into I'm going to call it the New World. It's like the Old Testament and New Testament. Let's ground some of these terms of critical power, W prime.
What are we talking about here? So, critical power, even critical power has actually been in the framework. Some people like it, some people don't.
There's multiple definitions that surround that. >> I haven't You're example for me yet cuz I haven't used it enough to really know what's going on here. So, let's I'll I'll take a step backwards.
Obviously, we've got functional threshold power, which was originally meant to estimate 60-minute power. But obviously, there's nothing physiologically different between 58 minutes, 61 minutes, but it's a functional marker to give an estimation of a of steady state. >> And that was because 60 minutes was just too mentally taxing for people to do.
Yeah, I think some of the original work was done from 40 km time trials. So, that was quite popular, how to pace those within that. And then I think some of the original work looked at obviously the 20-minute effort and that and it correlated pretty well.
So, why don't we just take an estimate of that? But is that the theory in this is it's a little bit more repeatable for doing Yeah, there's always going to be a learning effect, but I you know, if I had the option, do I go 20 minutes really hard or do I go 60 minutes really hard? I'd probably pick 20.
But the problem with this is there's a more than physical going on there. Anyone who's ridden a 60-minute versus 20-minute, there's no comparison. A 60-minute TT on your training peak schedule will ruin your week.
You'll be looking at that session on Thursday from Sunday night going, "Oh, I'm not looking forward to Thursday." 20 minutes, uh and then we got super lazy. We're like, "Uh if we do two by eight minute efforts, we can estimate the 20, which estimates the 60.
" It's like at some point you got to go, "What are we doing here?" Just roll up your sleeves and get the 60s on. Yeah, it's a nice term that I with the hour record that 30 minutes is not the same as half an hour.
I don't understand that. Talk to me about that. So, a lot of the pre-run up So, this has come from a I think Johnny gave a nice presentation at Cycle Science a few years ago with this exact title, so he can have all the credit for it.
And that when they were testing of riding for the hour, if you just do 30 minutes a test, it's not the same as doing half an hour. So, as in halfway through the first hour versus 30 minutes. It's a little bit of a play on words, but it is totally different from a 60-minute time cuz a lot of it is a mental battle.
When I was kind of more into the road scene before I kind of left cycling and more focused on my PhD, I specifically trained cuz I was getting really annoyed. I kept missing the break. So, I was like, "Right, let's make sure I can do long efforts and try and go solo.
" And so, I started building up, you know, two 20s, three 20s, two 30s, two 40s, to then doing 60 minutes and really trying to find that line of being comfortable to uncomfortable. Ended up stupidly going off the front on a road race and basically stayed solo away from it, but there's no way I could have done that mentally without of doing those efforts. Like, we may have a project next year to go for an hour record.
And like part of the framework that we'll do is make sure we can ride for an hour at that. So, it's just almost a rehearsal of what we've done. Cuz imagine going into the hour record not knowing that you can do it for an hour.
Yeah. And then crack like Dan's got a paper that he put out to where it showed one of one of his practice runs of where he went halfway and then just kind of fell off. It's a nice published paper.
I think Cody's the lead author on to it. I think Johnny had quite a few of them on it. Johnny's on it.
My supervisors on it. And it's my show of all the different physiological changes where the timing goes off, power drops off. You don't want to have that on a Yeah, I don't know.
Our coach used to give me a session called happier and it was just like warm up hour full gas warm down. It's like I learned it was an ironic name. There was nothing happy about that hour.
I still love to give it to people. It's just mentally you got to go to that place just understanding what a lot of what we're talking about today and you know, I had this conversation last week with a a friend of mine. Like, you you can talk about all these new technical terms, but I think Sean Yates has a brilliant line in the Year in Yellow documentary of Bradley Wiggins where he's like they're going through the TT course for the Tour de France time trial.
He's like, "Oh, you can model all this. You can virtual trainer. You need to get out in the effing wind and the rain and know where the potholes are and know where the ramps are.
" We've two different worlds that are almost butting heads, but as science based as we are, there's an irreconcilable truth that cycling is a hard sport that involves weather, elements, and more unknowns than there is knowns. Yeah, it's such a interesting way of words cuz ultimately you've still got to ride. You've still got to put the power output through it and suffer.
I used to I used to lie to myself. So, I used to set myself like three 20s and go, "Right." In the back of my mind I was going to try and do the full hour.
And I used to do this quite a lot. But if I set three 20s, I know I can have a little coffee. I can have five minutes ago.
Well, I've got to 20 minutes. I can do Let's just go for another 10. And then I could just do two 30s.
Oh, if I get to this 10, I can do the next 10. When I get to the final 20, it's like, "Well, I've done 40 minutes now. Let's do 20.
" But if I'd set out and written an hour, I probably wouldn't. >> I'd love to see some studies on this. There was one lad I had on.
His name escapes me, but he was looking at power differences when you engage the brain in activity. So, it was like lights coming on during the session that you had to turn out on your handlebars. Really interesting because you're younger than me.
You You're probably grown up in a world of Zwift and virtual trainers. Oh, no. I'm I'm I was in the [ __ ] And you're reading the back of a paint can and you're threshold and you're like trying to read the back of the paint can to distract yourself from the threshold and the fact you haven't got a fan.
>> I was just counting the lines of bricks. I was still with the good old magnetic turbos. I think I remember when Sufferfest came out.
And that was brilliant. I was like, "Wow, like Yeah, that was brilliant. Go a counter always attacks.
>> Yeah, that's like go and then they keep lying to you, but you you play the same video again and again. So, you knew it was coming up, but it was just And then yeah, There was a bit of spin class about that. Have you ever gone to a spin class and they're like "Okay, full gas 10 seconds.
" So, you're like 1,100 W and then you get to 10 seconds and they're like "2 minutes more." And it's like I can do 1,100 W for 2 minutes and 10 seconds. You've really You've given me an unrealistic expectation that this is finishing in 10 seconds.
Sufferfest had that about it. Yeah, it was it was nice. It made a change on to it.
But yeah, you still have to do those like longer efforts. Okay, so I think we're talking about critical power threshold. Threshold, we got sidetracked cuz we went down a rabbit hole on it, but threshold was originally that 60-minute power and then we started using these proxy tests of 5-minute blowout and 20 minutes.
So, where does critical power come in on this? How much of threshold is it or is that the best way to think about it? So, I kind of try to treat them slightly differently whether that's wrongly or rightly.
I always have this idea that FTP is trying to be associated with a time. So, you know, it's our 60-minute power, which is meant to represent metabolic steady state. Whereas critical power is the boundary between sustainable and sustainable, between the severe and the heavy intensity domain.
So, when you go above critical power, you start depleting W prime. When you go below it, you can exercise for a longer time. In essence, it's very similar.
Just the way we measure it is slightly different. So, critical power we can do a 3-minute time trial effort, 12-minute time trial effort. You can do kind of anything between in 5 to 20 minutes or 3 to 20 minutes.
But then if you kind of extend it, you can weight the maths a little bit as well. It's still a mathematical construct. >> say sustainable to non-sustainable, it's like how long can I ride at critical power for?
How sustainable is sustainable? Like, can I ride at this for 7 hours? So, they turned it between the main oxidative metabolism from that.
So, in numbers terms, we're looking from the research. And there isn't loads and loads on it. I've got a nice like article from it from my lit review in the PhD cuz this is one of those things that we have to kind of talk about the literature around it.
And it's between 35, 30 to 40 minutes, but the ranges are huge. I've seen some people that it's up to 70, 75 minutes. Some people that could be 25 minutes.
So, it very much differs. And generally the fitter you are, so the more aerobically developed, so the higher VO2 max, the higher CP or FTP, the generally longer you can sustain it for. But there isn't on average I would say 30 to 50 minutes within that.
Okay. So, yeah, I had it explained to me pretty similar. So, I asked someone who knows their stuff to explain it to me like I'm a 12-year-old.
And he explained it similar like a line in the sand that separates something you can do for a long time to something you can't do for very long. Basically exactly the exact same really. But But that that's almost going back to line of what FTP is.
Something you go above it, you can't sustain it for very long. You go below it. I do feel there's a lot of argument in trying to pinpoint it down to the exact what.
When in reality we've got biological variation day-to-day, which can account for a couple of percent. We've got variation within the power meter as well. So, I think it's just being aware of where it's likely to happen.
Some So, why have this new line? I think it's easier to test. I think three and 12 cuz then we can have two points on the power duration curve.
It gives us two data points to explain endurance. So, let's say FTP will give us FTP. Critical power will give us CP and W prime.
So, now I've got two parameters to make decisions. So, you can have two riders with a similar FTP, but two different W prime values. So, their work capacity, particularly for high intense intervals, will therefore be different.
Therefore, they need some different approach to make sure they can complete a session. Okay, so what's W prime? Even that's been in So, I did warn people this is going to be a This is my So, W prime originally came as a anaerobic work capacity.
It's now shifted to kind of a work capacity above critical power. It's not purely anaerobic. You watch a VO2 chart box in consumption.
There's a lot of oxygen being consumed at that point when you're exercising. So, it's kind of mixed. It's both aerobic and anaerobic, glycolytic, kind of all those energy systems, PCRs are heavily involved within it.
And Is it a number? Like Yeah, so you get a value. So, we measure it in kilojoules.
So, let's say 200 J for 10 minutes, not 10 minutes, but 200 J. So, you're working 200 W above CP is 200 J per second. So, your W prime is measured in kilojoules.
So, let's say it's 20 kilojoules. So, from that value, let's say I want to work out my 5-minute power from just doing a 3 and 12-minute test. So, I'll work out, right, that's the max amount of kilojoules I've got.
Let's divide that by the time, how long we can do it. Let's say it gives me a number of 45. So, it's 45 plus your critical power.
That's how long I can maintain 5 minutes for in in that a constant load exercise. What did the Giro d'Italia stage slayer, Mads Pedersen, and half the professional peloton have in common? Well, they're all turning to No Me Ol, the natural performance enhancer proven to reduce lactate buildup during intense efforts.
In the 2025 Giro d'Italia, Pedersen's form was undeniable. The Danish star surged to four stage victories. This was a major leap in form from his previous season.
And a key part of his preparation and performance was No Me Ol, developed by the same researchers who discovered the performance power of dietary nitrate. You know those beetroot shots that half the peloton were using? No Me Ol is clinically proven to lower lactate levels, reduce oxidative stress, improve training adaptations, and deliver a noticeable boost from the very first time you take it.
Riders are reporting bigger threshold power, fresher legs mid-race, and quicker recovery, all from a formula made with just three natural ingredients: broccoli sprouts, lemon, and sugar. Whether you're racing at the front or you're smashing local segments, No Me Ol helps you get more out of every ride. Take it before key sessions or races for an immediate edge, or take your training to the next level and get more out of your hard work.
Go to drinknomiol.com, that's n o m i o, and check out this game-changing supplement. Details are in the episode show notes or description down below.
How complicated is that equation to deal with, or how can somebody even to rewind this, somebody that's trying to listen to the podcast now and they're trying to move from what we call termed the old world into this new world. So, to figure this out, I need to establish my critical power Mhm. and I need to establish my W prime >> Yep.
values. So, I'm going to do a 3-minute test, a 12-minute test, and a 10 seconds Yeah, you said the 10 seconds more just to kind of get an idea of the top end. That won't come into the calculation unless you use another critical power model called the three parameters.
>> Okay. Which gives you three points. Well, we'll just use what is most used within cycling is your two parameter model.
So, it just gives you CP and W prime from a 3-minute and 12-minute effort. >> And what am I doing with those numbers when I have them? So, there's a couple of different calculations.
The most simple one is you do the inverse of time. So, 180 seconds, 1 / 180 will give you a value, and then 720 1 / 720. I think it's a One way or another, you'll quickly find out on Excel, and that will give you a value.
You then place that against your power values and do a simple scatter chart. You get two data points. Yeah.
Put in the line equation on Excel. So, um C equation and whatever Y equals MX + C, it will give you your CP and W prime value. Okay, so there's calculators, I'm sure.
It's It's Yeah, nice and simple. If you type on Google, our calculator pops up. Nice and easy.
You type in the numbers, done. >> Yeah, we'll try and link a calculator below for someone that wants to move to this. So, you have your two figures now.
So, I have my I've my 3-minute test done, I've my 12-minute test done. I'm plugging them into the calculator. What's the calculator spit now for me?
So, that will spit out your critical power. So, let's say it's 300 W, and then your W prime, which is 20 kJ. Okay.
So, I'm starting to think about my critical power now almost like like that line in the sand. So, it's separating sustainable and unsustainable. That one's kind of clear.
And kilojoules is again, I think maybe a term that some people are unfamiliar with. So, maybe talk to me what a what a kilojoule is, and then what I'm going to do with this result from the spit out of the calculator. So, in essence, it's just a way of we're measuring like work over time.
So, obviously, let's say we're working at 200 W, that's working at 200 J per second. Okay. >> So, you can do it So, when we talk about they need to ride and do 3,000 kJ, it's basically that amount of work that they've done that.
So, this just captures Imagine it's like a battery. So, it captures how much work that battery is capable of. So, then if you want to work out your what we call your utilization rate.
So, let's say I want to work at 50 W above CP, I'm going to use that 20 kJ battery at 50 J per second, then that allows us to work out then how long that battery can sustain that power for. I like that battery analogy cuz that almost captures intensity and duration into the same kilojoule. Like if you say to me do 2,400 kJ, I can ride hard, I can ride easy, and I'm going to get to that 2,400 kJ in a different time frame.
Yeah. So, okay, I understand what kilojoules are, and you need a parameter to calculate kilojoules. So, anyone with a heart rate monitor is flat out of luck Yeah.
I have my W prime Mhm. score. So, that's going to spit out as a like a What would be a typical 20 kJ, or what's it spitting out as a Yeah, so I think I've I've seen anything between 15 kJ to kind of 35.
Depends on your rider type, I would say. Okay, and this 15 to 20 kJ, can you peer comparison that like in terms of old world, if you had a a threshold power of 4 W per kilo, you're a cat three rider. If you've 6.
5 W per kilo, you're probably a world tour rider. Can I do the same with this kilojoule score, or is it individualized? >> I think it's more individualized, but I think that's the way it's going.
We've got a couple of ideas that I want to develop across the next year or two is like a W prime profile. So, it's basically all it's doing is explaining the shape of your power duration curve. Essentially, it's what everyone knows and is familiar with.
You know, they might be more of a punchy rider for 60 seconds. Well, that's likely to be supported by a high W prime because of the energy systems involved with that. You're not really going to get a good 60-second power with a low W prime.
Yes. Not physiologically going to really happen cuz you can't supply that 60-second power with just your a high CP. And cuz is the W prime calculated off just the 3-minute or the 3-minute and the 12-minute together?
>> The three and the 12. So, in mathematical sense, the intercept of the Y the Y intercept of that line, where it goes through that axis, is critical power, which is why it comes to a a number we can maintain indefinitely or forever, which we know is not the case. But if you go into the maths because it's the inverse of time, we'll technically never ever reach Okay.
that, which is where it came from. Whereas W prime is the slope. So, how is that kind of used within that?
Okay, so I have my W prime score. Now, what's my application of it? What Why would I care about this number?
So, the nice way I frame it with let's say we want to do a VO2 max style session. Typically, we might base that as a 120% of FTP. Yeah.
So, some people might find that's quite an easy completable session. Some might really, really struggle within that. One might have, let's say, a 30 kJ W prime, and one might only have a 15 kJ W prime.
So, one probably finishes the session, let's say, with 5 kJ left. The other one third effort third effort, they've depleted it, and they've got six to do. So, they're always going to struggle with them that.
So, it's starting to allow you to understand the power duration curve above critical power, which FTP doesn't allow you to do. If you consider FTP just that line, that's all we have on that power duration curve. >> Yeah.
With the CPW prime, we get the top end part of it as well, so we can understand that better. >> not do this for us? Like you can look at someone, you can go, all right, he can do 700 W for a minute, he can do 500 W for 3 minutes.
That was like a Was it just not as a nuanced an understanding? Yeah, it's it's Ultimately, we're still trying to explain the same thing, the power duration curve. But if we're basing and then let's say you don't know the athlete, we don't have this background information from it, and we just have this FTP value, Yeah.
then we don't we we don't know what that power duration curve is. Within that, it just gives us a framework to make a more informed decision because we know that athlete can complete that session. Okay.
So, maybe I'm wrong with this, but this seems like a valuable tool now for starting to plan and understand pacing. Mhm. When I get into race scenarios or time trial scenarios or or record scenarios, cuz there's a little bit more texture to the data points than we previously had.
Yeah, it just gives you those known points. If we have a good power duration curve, like so, an argument say you might not even need You can say, well, we don't need CPW prime cuz we've got a power duration curve. But it just gives you the framework and how that power duration curve is complete.
Let's say we've got P max defining the top. We've got W prime, which kind of works down towards CP. Then we've got LT1, kind of how that power duration curve is shaped with those values.
It just gives us a little bit more depth, a bit more understanding of the physiology around how that's created, and we get to know that kind of athlete's power profile type kind of better within that. But as you kind of alluded to, the critical power model doesn't work for periods of intermittent exercise. Yeah.
Because it doesn't understand recovery. So, what happens if we have then So, let's say we were doing this standard 6 by 3 at 120%? Or with with the W prime, we can work out what the maximum 3-minute power might be.
Obviously, you could know that from your 3-minute. Let's go 4 minutes. You're going to have to use the model predictor.
We can then use this other tool called the W balance model, which allows us to understand, well, obviously, W prime gets used up in the battery, but how does that battery recharge between those efforts? So, we can see if that interval set can be completed. >> Okay, this is helpful for understanding a good way to frame this, actually, is I of 20 16, 2017, it was me and one of my training partners.
Very similar power across all durations, 20 minute, 5 minute. I think I won one bike race that year. I'm going to say he won 20, and then signed pro the following year.
I couldn't really understand what they did Was he a bit better than me? Potentially, he understood, you know, race dynamics a small bit better. Mhm.
But now, I'm wondering, is there almost his ability to recharge is was different, and we didn't have a tool for measuring that recharge ability then. So, now we can. So, that's kind of what my first research paper was looking at.
We wanted to look at what W prime was correlated to. So, not necessarily causation, but some of the factors see if we could predict it from other values, and to see kind of demystify the fact it's not just anaerobic like VO2 max correlates with that as well. Like, it's we should really do a multilinear regression see what factors cumulative like correlate to it.
But then, what factors are correlated to how much work we can do above critical power, so our ability to recover. Then, we individualize this rate of recovery, shall we call it? We call it tau, which in essence is qualitative.
We call it tau. Tau. I think it comes from physics, so it's the rate, the time which it takes to get to 63% of 100%.
Okay. So, on a monoexponential model, so there's just one gain term into it. We measure that in time in seconds.
So, in tau in the W balance model represents in how many seconds can we recover at this power output back to 63% of our W prime. So, a lower tau value is faster time. So, that's kind of what we aim for.
And we There's a couple of models around. So, originally we had Skippers first model, which came around called the integral model, Skipper one many refer to. We got Skipper two, which is the more popular differential model.
It's just some differences in the mathematics of how it's compute computed. >> are exploding right now, just so you know. That's And then, we've got the Bartram adjustment and the Pugh adjustment to that, which is based on the Australian track squad team pursuit, and then the New Zealand men's national team, international team, and regional team within that.
And then, you can individualize that. So, they just took that a general time constant for these groups. So, this explains that they recover at a set rate, and then you can plot this and run it in real time with your power file.
You threw out another term that I don't know what it is, W balance. W balance. So, that's your balance.
It's like your bank balance, or your rate on your your charge rate on your iPhone. So, let's say you've got W balance at this point is 93%, or you can measure it in obviously kilojoules versus how much you've got. You can do it absolute or relative.
Okay. Is that 93% depleted or 93% remaining? Remaining.
Okay. So, is W balance my ability Does that explain the difference between me and my friend who's winning a lot of races? It's that recharge power below critical power?
That would be tau. So, imagine balance like your bank balance at any one point. >> Okay.
So, if you look at two different riders, Mhm. what's like the tau calculation to tell you this guy can recover faster than the other guy? Like, what what does that look like?
Is it numeric value? So, that's a whole another So, yeah. Wait, what was I saying?
So, currently there are generalized equations for explaining this this value. So, we measure it at We call it DCP. So, it's the difference between >> Explain this to me like I'm 12.
So, critical power and then your recovery power. It's the difference. So, difference to CP to recovery power.
So, a DCP of 50 is 50 W below your CP. Okay. No matter what it is, a >> And how fast am I recharging at that point?
Cuz if I think, you know, say for me, I don't know what my critical power is. Say it's 350. Mhm.
If I'm 50 W below below that, >> Yeah. I'm assuming I'm charging at a different rate than if I'm 150 W below Yeah. below that.
So, the idea is the currently the model framework takes into account the intensity, so the further away you are from CP, the faster your recovery. So, essentially, if you're freewheeling, you're recovering the fastest rate possible, cuz that DCP will be equal to your threat CP. Okay.
So, if it's 350, obviously you can only ever get to 350 as a DCP. And then, it's the how long you spend there. At the minute, that's what takes into account the recovery.
So, you can individualize it. So, if I wanted to work out what yours is and what mine is, we can do It's two ways. We can do it in the field and do a series of intervals, and then adjust those parameters in the model.
So, if you said you were done after five by fives, and you failed on the last one with two minutes to go, we could adjust the parameters to find out what value tau needs to be to explain that. And that's how we can do it. So, we can reverse engineer as well.
>> Interesting. So, if I've won those Chinese fans doesn't charge properly, like, is that me taking time off? How How inactive do you need to be?
Almost like, what's starting to What are you starting to lose first? Like, I found when I came back to do a critical power test, I could actually bluff the three-minute test Mhm. quite a bit.
My 12-minute power was miles off Mhm. where it used to be. My 12-minute power was like 40 W below or 50 W below where my 20-minute power used to be.
Mhm. But my three-minute power is it's not that far off. I've I've bluffed that quite well by just being generally active and kind of, you know, riding the bike and smashing the to the coffee shop with the lads every now and then.
Does this recharge capacity, is that massively affected by sedentary time? Yeah. So, obviously, if you did that in the three-minute 12 three-minute and 12-minute test, that would weight W prime high, CP low, which would then only give you a DCP range of a set percentage.
We now know that intensity domains, so whether we're in above or below LT1, that impacts the rate of recovery as well. So, I I think the research is moving towards that where there's an optimal rate, which is the fastest rate of recovery, which kind of comes into some time trial bits where I'll get onto in a minute. And then, aerobic parameters are highly influenced by recovery.
So, we know type one muscle fiber type influences CP. We know CP influences the ability to recover. So, it is a trainable metric.
And obviously, you can improve that rate of recovery. Yeah, and does this start now help me to pace a non Like, if I'm thinking of climbs and iconic races here, it's rarely just like 6% bottom to top. It's like a 14% pitch, a slight downhill, an 8% pitch, and then we're steady to the finish.
Do you start figuring out what you can hold on the steeper pitches? You Can you model this backwards? Yeah, so a lot of it will be now I think this is where it's going to advance into kind of the more world cuz I know multiple teams are now looking into it.
Well, Eddie I bought Lance one from Vector, and he was talking about how they modeled Eddie Dunbar's pacing for his Vuelta stage win, and really, really interesting. Cuz a lot of it you can use the information you've had before, which does get a little bit harder as races change over the years with different kind of intensities within that, or they might change certain sections. Like, what I think it was like Was it last year into the Arenberg they put like a dead 90° corner, which will affect >> So, what you can like work backwards So, what If let's say you've got a GC rider, what what was likely to happen to them going into that?
So, can we actually minimize their W prime depletion, so they've got a full W balance going into it to then model what they can do what what they're going to have to do on that first bit with that surge? So, how are you minimizing that? Is that That's just a bunch of efficiency.
So, that would be team tactics and bunch efficiency, making sure that that rider is well placed. So, the team might have an objective, right? We're going to use two guys up for this step for this section over the next 5K, and to protect this rider.
So, you can even review it within the DS's post race. Actually, did these riders do a their job as we expected it? Cuz you can basically overlay six W balance charts and find out the point in that race, right?
These two riders should be depleting here if we're protecting, or at least our GC rider shouldn't be using should be minimizing at their use of it. >> Okay, so you can start to explain why a GC rider is dropped when you didn't expect him to be dropped. You know, like sick setting tempo, and you're like, "Whoa, how's Ayuso gone out the back already?
" Mhm. So, you can use it to try and explain the narrative of what's going on, why it might have happened. It's just asking better questions using the framework, cuz a power duration curve or power profile might not necessarily tell us that from from a race.
So, we can use it to retrospectively look at what happened, use it to then plan ahead. So, let take the classics for example. We know that they might There's going to be a minimum amount of use that they're going to have to do above threshold.
You know, there's years and years worth of data. We can work out, "Okay, on average by might be the Koppenberg for on the third time, second time, you know, they've done nearly four, five, 600 kilojoules worth of work above CP. So, it gives us another factor to kind of look in durability.
So, how's intensity impacting our ability to produce power later on? So, actually, can we put fueling strategies in place? Cuz if we're using above CP, it's likely to be carbohydrate dominant.
Are they actually then recovering between the sectors particularly well? If not, why? What's happening?
Can we use that to our advantage to make to shake up team tactics? It's There's lots of opportunity to use it in multiple ways to ask better performance questions. I was going to be going to be my follow-up question.
I had Sam Impey and David Don both sitting in the siege room Hexis there, and we went deep on fueling. Do all these calculations assume optimal fueling? At the minute, like, that's probably where the nutritionist's role would then come in with the kind of sports scientist physiologist within that and the DS.
It probably wouldn't be something cuz obviously nutrition's not my expertise. So, that would be like, what do they think's going to be happening? Kind of work collectively as a performance team to then, what is the most optimal fueling strategy, team strategy?
Where should we pace the Swannies within that to make sure where's critical where they cannot miss a feed within that. It allows you to then have that full performance framework together with s- an objective measured justify that decision. Yeah, so when you're looking at the modeling and you're trying to explain why to use that last example, say of course the set tempo on the front and you so as you're after getting dropped, you're looking at the data and you're trying to explain, okay, he's after getting dropped here because he's had too much time above CP in the run-up to this.
His domestiques haven't done a good job protecting him. You need to have that nutrition layer layered in to that understanding. Like this can't be siloed, which we've seen in cycling for so long.
All these different siloed areas where the sports psychologist are working in a vacuum from the strength and conditioning coach to the DS to the team coach and you're just left kind of in these rather than being collaborative relationships, it's all quite competitive. Mhm. Yeah, and I think it's also taken then a step back.
Like models are great, but most models are wrong. Some are just less wrong than others. And I think, like I said earlier, with some of the concepts around CTL, TSS, FTP, it's acknowledging then its limitations within that, but I'm not sure, okay, is there anything better?
What is actually happening? Are some things just always going to happen? Like, you know, you can't take the art of racing out of racing through a mathematical model.
It just helps us to maybe justify why we need extra support on these situations, justify kind of why this decision needs to be made within the team because it's to protect them within particular areas. And we say let's say we check the wind direction within that. We're going straight into a crosswind.
It's likely going to be dangerous to split it. Let's ensure we use one of our riders to put in a bit more W prime to protect the other riders as well. That's kind of we know that as an approach within teams.
It's not anything new, but if, you know, we need to justify it in a particular way, it's a nice framework to use. So, why might the modeling be wrong? If you look back at it and go, that doesn't make sense.
Like some of the things that are my head, and it's probably sure it's a much longer list is nutrition's gone wrong. Mhm. Core temperature's raised.
Mhm. On a bottle job. Yeah.
Anything else I'm missing that might you looking at the model? Fatigue. Cuz currently, like I said, the model at the minute assumes, one, you can go back to 100%.
So, your battery will always go full. There's then the assumption that it's only the intensity of the recovery power and how long that you recover for. So, let's say you did 5 by 5s, it would the model would assume that your recovery is the same.
So, then let's say you put that into an endurance ride. So, you do an hour 5 by 5s, an hour with 30 minutes tempo in there and 5 by 5s again. The model would assume that the recovery is the same on that versus your fresh effort.
I think intuitively, we know that it's not. In the third week of a grand tour especially. >> Exactly.
So, this is kind of where my and my research is more kind of gone into like where are the We've taken more of a pragmatic approach looking at where does the model not hold true with its assumptions. So, is CP and W prime fixed? The model assumes so.
Well, some research by James Sprague where they looked at, you know, does inten- intensity is one of the most important factors for durability. They found that I think they did like a 3 and 12 test. Fresh one day, they did then like 5 by 8 minutes at 105% of CP and then did a 3 and 12 again.
They found that and then they did one with like an endurance ride in there as well to do like I think 2,000 kJ. They found that intensity then reduced W prime within that. So, for me it's like, "Hmm.
" So, we kind of developed a couple of the research studies which has influenced the new model where we looked at acute fatigue. So, we did a a horrible interval session. So, we wanted to look separately at W prime as a rate of utilization.
So, I want to be rather than 5-minute power, which we can predict with CP. So, let's say someone's got 30 kJ and 10 kJ. If you do 5-minute power, it'll have a different rate of utilization.
That makes sense. >> No, explain that again. So, the way we would work 5-minute power out, but we'd have our CP value Yeah.
and then whatever our W prime is divided by 300. >> Yeah. Obviously, if it's uh 30,000 / 300 versus 10,000 / 300, they're going to be different values.
So, you're using your battery at different rate to do that 5-minute power. So, he's like, "Right, let's look at this battery on just the battery itself." So, we chose 60 J per second.
So, essentially 60 W and 100 W 120 W above um thresholds or above CP and look if that has an impact. So, we then did 40 seconds on at this power, 20 seconds at 100 W below your CP. We did this until you'd done around 70% of your total W prime.
So, we knew they could get through it. And then they went to a time to exhaustion at that power. They then had two Who signed up for this?
>> They then had 2 minutes at 25 W and they went straight back to that power again. They then had 30 minutes easy and repeated the entire thing again. So, at the two different intensities, they were work matched.
So, let's say if they did four efforts at the 60, they would only do two efforts at the 120. So, they started the intervals roughly the same amount of work so we could compare them. And there's a lot of interesting findings from that.
So, we presented at Cycling Science this year. The abstract paper should hopefully be out in the next couple of weeks. And that both intensity and fatigue impact both W prime and the ability to recovery.
So, we put this into a new model to have a dynamic tau which takes into all this all accounts. So, it's not now just based off that, it's based on how many intervals you've done, the duration of those intervals, how much recovery time you spent. The more time you spend above CP, then changes happen to W prime.
So, we've got a far more dynamic model now. But does that dynamic model need to start pulling in sleep, heart rate variability? Like a a bad night's sleep totally impacts your ability to Yeah, I think then the question is how do we quantify the impact of sleep on those things?
So, rather in a performance and let's try and optimize that so they're in the best conditions possible and then try and see what happens within that. Like a lot of the data will work backwards. So, understanding does this model parameter fit the data?
Work out then why This is what we continuously do then work out why it might not be working. So, is it nutrition? Is it sleep that's been happening?
Is it consistency? How variable is it? To then see if we can develop it into a framework where we can go, "Right, for this TT or let's take the opening stage of the tour next year, we've got what's team seven, eight riders?
What's the most way to use their W prime to go as fast as possible using that model framework?" That's cool off. That's cool.
I like that. That's I think where it's and I think within within the time trial environment, it's I think within time trials, you'll be able to use it to predict and pace before performance knowing that rider's physiology. And in its current setting, it sounds like it's probably more useful towards the start of a grand tour and maybe less useful as fatigue and I think it depends on how you use it and what questions you ask.
So, there's certain things are going to happen, let's say, in the third week. But we can probably have an idea of what will happen be happening to that rider in the lead-up to those key moments. So, we can still try and protect them, still try and minimize them because this is likely to happen.
And then kind of work out, okay, from previous data, after the third week, what do we see? So, you can retrospectively use data sets to make maybe better decisions as you progress within that. >> Especially as you get riders doing six, seven, eight grand tours and we have the data from that, you can start to model their fatigue levels.
>> Like you'll be able to like you could put It'd be quite easy to pull now within a team and a data set like what what have they been doing across from their first day to the second day, let's say. Obviously, you've got to take into account team roles, what they've been doing within that, but just making sure I've said quite a few times, it's just acknowledge the limitations of the questions you're going to ask within that. We talked about We started off talking about Joe Friel here.
Mhm. Um one of the uses of this threshold test, well, even if you got broadened out to threshold test, and I'm not sure was Friel or Coggan that started talking about power profiling test. So, we'd go a 10-second effort, a 1-minute effort, a 5-minute effort, 20-minute effort.
So, you might look under the old model, say we take a reverse periodization model and so we're working on 10-second sprints for the first block. We'll test 10-second power at the beginning of the first block. We'll come to the end of the first block and we'll retest 10-second power and go, "Okay, that kind of worked.
Here's our improvements." Under the CP model, are you retesting 3 12-minute power after set periods of time? Or is that useful?
Yeah, I think with anything like with any testing framework, in the grand scheme of things, it might not necessarily matter what it is, but as long as it's repeatable or consistent cuz then you can assume that no other factors have changed it. Like one of the issues sometimes I find is riders go to early training camps, they do the best 20-minute power up a climb, come back to the UK on the flat rough roads and can't hit that power again. That's because they were climbing.
So, if you're going to test I sometimes like to do shorter duration tests. Let's say >> comparing apples with apples, I guess. Cuz you're going to have different torque values.
>> going to be doing your effort hard efforts on the climbs, you might as well test on the climbs. If you're going to do them on the flat, test it's not about ego hunting. It's it's making sure it's an accurate representation of your current physiology to then make sure we are using it in the most informed and productive way possible.
Yeah, I guess even on that, like we've all seen riders who are better out of saddle than in the saddle cuz of the composition of muscle fibers they have. Mike Woods been a notable example coming from running, podium in the world championships. Just a beast when the road went uphill especially over 10%.
Mhm. Mike's data doesn't look the same on a flat road as it does going up the side of a mountain. Yeah, like some of my best numbers are always done out of saddle cuz if I came from a mountain biking background, out of saddle two to three minutes flat out.
Whereas to do that sort of power, there might be a 30, 40 W drop. >> Yeah. So, how stable then is CP values and W prime values over the course of a season?
Like what actually moves them? Is it volume, intensity, ability to process carbs? Yeah, kind of a bit of anything.
I think the shorter durations will be more sensitive to fatigue. So, you can have more volatility within those. I think I think it was even James and Peter Leo put a nice paper out looking at the power duration profile changes across a season.
I think yeah, it was volume sustains kind of like CP and the intensity will and fatigue influences kind of the W prime kind of factors within that. And the zones change day to day, I've noticed on vector. Like you'll log in and it's looks slightly different.
It's re-estimating critical power off recent sessions. I'm not used to seeing like you know, you look at your training peaks and your threshold values change from like 370 to 360. It doesn't happen like What's with the drift in zones?
I think that's just taking into account what you've done. It's basing that on what you've done previously. I can't remember what the time frame is, but might be the last 90 days what's been happening within that and then telling you roughly but obviously there's going to be variation day to day like you said based on sleep and nutrition.
This is trying to capture what is likely that physiological state is. Again, with all data-driven models it depends on the data that you've inputted into that as well. But you just got to be careful of not training for your power duration curve.
You'd want to be training for >> Yeah. your event demands and what you're aiming to do. Yeah, and Do you remember how you used to look at So, if you had an athlete come in and they do a power profile test.
We say with that we stick with the 10 seconds, 1 minute, 5 minute, 20 minute. You could look at that power profile. You could say, "Okay, your 10 second, 1 minute powers are rubbish.
Rubbish relative to your peers. You're a cat two rider, but you're actually cat four on 10 seconds. 5 minutes, you're starting to be getting up on cat three, but on 20 minute you're actually a cat one even though you're a cat two rider.
Now, let's look to events that suit your particular characteristics and we'll reverse engineer the demands of an event and maybe double down on that strength of trying to increase your 20 minute power. When you look at this new modeling, are you How do you distinguish between or what does a profile look like from a criterium rider to a time trial rider? Yeah, so it depends on Obviously, we're still trying to all estimate the same thing.
Can we explain the power duration curve? Can we predict it? And then kind of that rider strengths and weaknesses, they're going to have be more suited to certain things.
But again, it depends on you know, where they are now and what that needs to look like. You know, if you're you're a time trialer with a big W prime, I tend to find you have to be quite careful cuz if you deplete W prime, your recovery is really quite compromised. So, you can look at your CP to W prime ratio or kind of W prime and P max ratio within those things to then try and ensure the weaknesses aren't going to be too much of an issue.
I tend to split criteriums up into three periods. Um the different dominating factors for each of those periods. So, the first 15 minutes, let's say it's hectic as anything.
It's everyone's fresh, keen. So, your W prime is going to be a minimum that you need and a minimum CP. Then the next kind of let's say half an hour is a little bit less hectic.
So, actually that's going to be able to recover right smoothly within the bunch. And then you get to the last let's say five laps or so where you need to have done those things within that profile and now it's actually about positioning, making sure we've not done too much work above CP that our P max is declined cuz intensity will reduce that to then deliver that. Whereas some people quite commonly go, "Enter a crit, I need to work on my sprint.
" But fresh it's like 1,400 W. Yeah, it's like right now what we need to work out is pushing the CP up so we do less work above CP so we're fresher for the finish or you change your tactics up. Yeah, it's maintenance of that battery.
But if you're let's say Another way of describing W prime is like your matchbox. How big and how many? So, this is where the modeling come My new model kind of comes into it.
It takes into account we adjust some parameters. Well, let's say you've got a big W prime rider. They're going to have a couple of big matches, but once they're spent, that's it.
They're they're they're going to be crawling at the back. >> So, understanding that is really feeding into their tactical plays for the day. So, someone like me where I had a reasonable W prime, but I could attack at a decent CP.
I could attack again and again and again, but that initial big surge probably wouldn't be that high cuz my W prime wasn't as big, but I could go again and again and again. So, let's say I was in a two-up. It was me versus someone with a high W prime, I would attack and attack and attack and attack them until So, it might come down still to a sprint finish, but I fatigued them so >> It's really interesting when you start overlaying on tactics and how this plays out in the Yeah, and I think this is where we it's rather than just going from a model's perspective and making those decisions, let's relate it.
And this is how my translation goes to the riders that I work with is to explain it in their terminology within that. You start explaining it like that and goes, "Ah, it's nothing new. Riders have been learning that how these rider types are for the last time 50 years or so.
It's it's just putting something more measurable to it rather than taking away, "Oh, this model says this." >> Well, you're starting to understand what your strengths are. When a lot of time if you're in a break, you're maybe not as aware of your strengths as your game is trying to analyze someone else's weaknesses.
Like we used to joke when I was riding the Irish team Aqua Blue, one of my teammates would say like the song The Gambler by Kenny Rogers can basically explain every bike race. It's like the lyrics to it was like, "I made a life of reading people's faces, knowing what the cards said by the way they held their eyes." And so we'd be like that in the break and I'd go, like is his cadence dropping in this section?
We'll give a little tester here. You know, how fast is he to respond to that? And you're putting together clues for the first, you know, 80% of the race to try and take those clues and piece them together to unlock this puzzle in the last 20% of the race.
And you know, if you've multiple teammates in the break, this gets even more nuanced where you can you can play your cards a little bit differently. But it's really interesting to understand that and profile someone go, "Okay, he's he's the fellow that can go one or two big attacks versus he's the one who can go with 50 smaller attacks." Yeah, and you can use it to then try and work out okay, what situations do I need to avoid to make sure I don't get dropped?
And you can also use it then to try and play some really dirty tactics. And that's what I've done. So, like one of them it's So, you obviously let's say there's a break of five of you and there's one particular person that you want to drop.
So, obviously you you're on the front and you know and they're behind you. So, you do a good like you try and put yourself a little bit in the red on this one. So, you give them a bit of gas and make them push.
So, you might go through on them a little bit hard onto it, but you know their turn's next. So, then you go to the back as you do on like a a team time trial setup and he comes back. So, you soft pedal and let that gap come.
You got to make sure you don't screw yourself over. You know he's just depleted let's say a lot of W prime. He's going to then try and accelerate to get back on, but there's now a gap.
He's probably not realized it. So, you attack him. You just ride him off the back.
But then you got to be careful not to do that to yourself. >> Yes, the art skill bringing yourself out the back. >> Yeah, so it's just kind of understanding how physiology might relate or a a framework to explain physiology can be used to make those tactical decisions.
Let's say I've had a rider where I've worked with if we know that they get to like close to full depletion in their W prime. So, they might be too keen at the start or I wanted to test the legs. Like no, don't test the legs.
We need to preserve this as much as possible so then you can actually race and not feel like you're in the red and slip back and then get dropped. So, we can use it to help something that they should know intuitively which can be hard for some riders quite new to the sport >> Yeah. to allow them to understand what might be happening, why performance has happened in this particular way.
>> Well, it speeds up that whole apprenticeship a little bit. Like that idea of le métier of my French director would have beat into me. It's just like the apprenticeship of cycling that I think understanding tactics is a part of that apprenticeship.
Like you would have been intuitively the fastest guy in the group that you'd know, "Okay, he's the fast twitch. He's the guy I don't want to bring to the line." I'd want a position where he's coming through on a turn on me and that you know, the the five person classic single pace line.
As I pull to the left and he's exposed into the wind, you're kind of trying to look nonchalant. Maybe you're rooting in your pockets. Your turn's at 400 W.
But as soon as he exposes himself to the wind and he comes through to try and do a turn, you're not coming off the gas. You're staying on that. Maybe open that 480, 490 and he's driving racing you through or trying to come through and that's just depleting in your new terminology.
Or you know, we've had I think I analyzed a file for someone where they went through into a break in a crit and then you know, they got to leading three and they committed to the turns rather than saying, "No, I need like a little bit just to give me five turns and then I'll come through." So, we pretty much stayed at like zero, close to zero kilojoules or zero percent battery and then went out back cuz he stayed there, couldn't pull a turn. So, obviously you just get dropped.
Cuz cuz you can technically speaking have it on your head unit on certain ones and you can literally say a percentage wise how much you've got >> So, what do you put on that What would be on the head unit? >> So, you could I think there's a couple of different apps you can have. Again, the mathematics is kind of more the old size so it's not as accurate now, but you can have a percentage or how long it's going to take you to recover, how long you've got left at this power output.
Those are those are the things you can have with relatively good confidence left just to make those decisions. >> And again, as a newbie, that's quite helpful. It's just like you know, you're attacking someone who just commits full gas to it.
Obviously, you can't. You want to just get a gap quickly and then settle into a power. Like if you're solo, you don't want to go too hard on the climbs that you blow like it just Some riders can feel it.
They know it far better. Some not as well. It just Cuz there's a confidence also to like and that confidence comes with experience.
But if you have a data point to give you that confidence. Like you know, someone's gone super hard. You've tried to follow and then he's flicking you through.
The experienced rider is going to go, "Bro, I can't ride through that speed." Which is nice cuz it's like a veiled compliment, "Oh, you're super strong. You're killing me.
" But it's also you recharging in the you know, the new world. But a newer rider will, once he flicks through the elbow, will come through at maximum to do his turn and then get dropped, you know, 40 seconds later because they were never recovered enough to do that turn. But having that data point for them maybe changes that scenario.
Yeah, and it then looks at like you can see then how like your recovery's going within that. Like just to make, you know, you've done a couple of good turns. Let's say there's a couple like three of you and actually I don't know I've got 30% left.
You've been pushing on. The gap's probably steady. Now you can probably soft tap a little bit and then keep it pretty consistent within that rather than both all of us dying a hole and then realizing we've got to knock the pace off by a couple of kilometers an hour and then we're going to get caught and it's a wasted effort and then we screw the rest of the race up for ourselves.
So, I'm assuming recovery isn't linear. Like how does W prime recharge? What shapes that recovery?
Is it power, duration, cadence, terrain, drafting? So, the main masses is a curve linear like a we call it a mono exponential model. There's a more recent suggestion that it's a bi-exponential model.
So, there's a fast recovery phase and a slow recovery phase. Um which we've got the both modeling for. But yeah, imagine it's like quite fast at start and then starts to plateau.
So, a bit like an opposite of a lactate curve. Okay. So, like recovery is like that sort of shape.
So, you've got DCP along the bottom and tau at the top. And what's causing my recharge? Time below critical power?
Or and it doesn't matter how I get there, whether that's you know, smart drafting, whether it's you know, just coming down. Like it doesn't really matter how I get to Yeah, so at the minute it's the power output that's happening which does the DCP function. So, you know, 50 watts below.
So, you recover technically from the model a faster at 100 watts below CP than you will 50 watts below CP. And then how long you spend there will be based on each So, you basically recover second by second. And obviously you'll recover more in 10 sec in 30 seconds than you would do 10 seconds.
I can remember vividly a December training spin. I was on the backroads. There was fog, wet leaves, potholes everywhere.
I was riding on a descent 100 trying to stab this tiny little mode button and praying the battery would last until the final climb of the day. You know that anxiety, the is it going to die anxiety. It totally ruins your ride.
And the truth is for years was normal. Every season I got into this repetitive cycle. I'd end up buying a new set of lights in the winter.
Normally plastic lights, plastic brackets which would invariably snap, charging ports which would fail and batteries that would fade. Most lights feel very disposable. They're shiny for the winter and then straight into the bin at the end of the winter and we repeat again the following winter.
That all changed when I got my first set of Exposure lights. Suddenly this wasn't a consumable anymore. This was a piece of kit built to last a lifetime.
That's why I'm absolutely buzzing about the fully updated Exposure range. Reflex 2.0 automatically adjust the brightness of your light.
It dials it down when you're going at slower speeds on the climbs and then unleashes the full power of the lumens on the descent. So, your hands stay where they need to be, on the bars, so you don't actually crash. There's another really cool feature.
It's called reserve mode and it kicks in when run time hits zero. It drops to a low beam and you get another 30 minutes to get you out of trouble and get you home. USB-C charging is a total game changer to my mind.
It means you can get now 70% faster charging. But importantly for me, you can top up your lights from a power bank mid-ride. So, if you're doing an ultra, a bike packing trip or you're going on event like me at Badlands, just plug it into the power bank and away you go.
No need to worry about batteries failing. Exposure isn't just another set of This is the last set of lights you're ever going to need to buy. Go to exposurelights.
com to check out their full range. Okay. So, as a coach, what's the practical uses now of this W prime for setting like interval durations, deciding recovery lengths between efforts?
So, I've used it more so like we stand sub-threshold efforts. We can't we don't use it cuz you're not going above CP. So, it doesn't really come into much framework there.
But more intensity-based sessions. So, want to make sure that the session they can complete. So, if I want to make it a little bit hard for them, push them, it gives me an idea of how far can I actually push them before they're unable to maintain that power output.
So, if you've got something challenging, they might voice, "I'm a bit concerned about the session. It's intense. I'm not sure if I can complete it.
" Imagine then your coach runs this model for you, sends you the picture of the mass. And normally if it's particular ones I'll attach it as a file or send them like a screenshot WhatsApp note. I go, "Mass says you can do it.
" I'd be quite happy. Like, "Okay." Then you come in the framework Well, but then you can ask more questions within it.
Okay, let's understand maybe recovery's not as good as we thought it cuz at the minute we're just using a large majority of it is generalized models which don't work for everyone. We've a paper showing how four different recovery variations have massive variation within that. So, first it's understanding and learning that athlete.
I can then personally individualize it for that person. And then from future sessions we can then make sure it's specific to them. If they're improving, we can actually isolate the ability to recover between efforts away from CP and W prime improvements.
And can you personalize the recovery? So, if you take a a 2 by 20 threshold. Okay, so you do a 2 by 20 threshold.
Like there might be a few different reasons you give a 2 by 20 threshold. Okay, straight up we want to improve your 20 minute threshold power. One.
Second application, maybe I have a shorter recover period between the two of them. So, I don't want lactate back to baseline. I want lactate start from maybe a slightly elevated point for the second interval.
Another application in the same session could be, hey, you're doing 2 and 1/2 thousand kilojoules first and then you're doing your 2 by 20 for sort of durability fatigue resistance type. Are we more accurately able to estimate what the gap should be between these sessions now cuz maybe previously it was 2 by 20 with a 20 minute gap for the fatigue for the more lactate clearance one. You're kind of just guessing going, "Well, should I have given them a 5 minute break or a 10 minute break?
" Can we get more accurate understanding of I think yes within the idea of what W prime you have left. If you want to have this inference of other physiological things so like lactate clearance, you'd probably have to measure that first cuz there's quite a lot of variability within that to see what was actually happening. What is their clearance rate um within that in the 5 minutes?
What do they turn to? We've done it with like VO2 before to see what's happening. VO2 can return quite quickly.
And then what sort of state is that second effort We can also make sure where let's say it is 220s but they're doing them too hard and you let's say they start pushing like 10, 15, 20 watts above it and that W prime's getting pretty low throughout it. So, it's well, it's a threshold effort. You shouldn't like let's say we're working just below it.
We shouldn't be using any of it. W prime. So, I'll be staying below threshold within that.
For an under-over session, is it a session that is actually completable within that cuz obviously that surge will deplete W prime. You might still be working just above CP. So, actually is there enough Are those surges hard enough?
Are there too many of them? We can then adjust those parameters for that individual a bit more concisely, I think. And can we see then if someone's riding too hard between the intervals, that charge rate?
Yeah. So, if they're let's say a DC 10 watts below threshold for some strange unknown reason. Like not sure why.
I've got training partners like this. >> Like, you know, then it's just not going to recover. It's going to recovery rate will be very, very slow.
So, let's say I think, you know, it could be like say 700 seconds at that power output. So, it's going to take you 700 seconds to recover. So, nearly 700 seconds is what?
Uh 10 minutes and 40 seconds. I was going to say 8 to 9 minutes from my matches more or 11 minutes and 40 seconds. Um so, it's not feasible at all.
Yeah. Very interesting. So, I think I'm starting to get a grip on this new world.
You're working with some great athletes. You know, you're not going to be able to mention I'm sure to say a bit of confidentiality around. You mentioned Laverack already who won't mind you has been loves a bit of free publicity.
Yeah. He's Go check out Joe's medium. >> My training, yeah, he's put I've done a couple of sub-stack articles on like how we've taken winter approaches a few times.
So, we have nice simple trainings. >> I really should have read those. But Sorry about that, Joe.
Pragmatically, when you're planning, like Joe's training, how are you pulling together all these tools? Like how much of these are as Vassilis Anastopoulos said to me like just new words to describe stuff we've already been doing? Or how much is this influencing your coaching on a day-to-day?
I think it's allowing me to ask better questions within that. So, it might not necessarily directly influence certain sessions. Like ultimately for a zone one ride, right?
It's not going to change much. But if you start seeing a zone one ride and you run the W profile and W prime's been utilized a fair old bit, it's not a zone one ride. Okay.
Like it gives you a quick framework to see like you can see a power profile, but actually see that in the lens of separating between above and below threshold. It gives you a quick view to look at it. So, particularly for race analysis, I use it a lot.
We built um an app where you just pop your fit file in. You can check all the models on. It's a free online app cuz I wanted to make sure my research was getting accessible by everyone cuz there's not many You need to know the W Bal framework to be able to individualize it.
Technically, with this model you can. You can have a play around yourself and see what it fits around. So, it'll then look at, you know, your most depleting efforts within that.
You can ask it to say it show you how many top efforts you've done that. So, what sort of key moments in the race within the happening. So, from that we can get an idea of what trend direction might need to go, what seems to be consistently happening, or where performance isn't being as good as it could be.
Is it after a certain amount of work above CP that's doing? So, is it a case of okay, we need to build LT1 up and build that base up more in the winter. Do we see the inability to sprint after things?
So, do we need to drop back volume to make sure we're still got Cuz if you do lots of endurance work, you just sprint, you know, it's a tip and scale. You want to get that at the most optimal point. So, it does influence it quite a lot within the day-to-day.
And just in terms of framework that we that we use within that. So, for the coach that's all in on performance management chart. Mhm.
And that all world, what's that migration pathway look like? I think it's a lot with coaching. You've got to make things quite simple and easy to use, not too time-consuming.
Which is kind of why I built one of those apps, ultimately. You know, let's say I've got 15 race files I need to quickly go through. Let's say I'm working with the team and there's 15 files I need to go through, report to DS's, head of performance.
If I did it manually, that's, you know, download from Training Peaks to WKO, export that into Excel, do that 15 times, then run the parameters on that. Well, I can just upload 15 fit files straight away, done in 30 seconds, if that, from that. Put it into a point where I can do a quick report for each rider.
So, this is what was happening. This is, you know, we used this person at the right time. You can use it quickly.
So, So, got you again. What's What's your workflow? What What software are you using?
And you cuz you mentioned an Excel as well. So, you're using some custom formulas there. Yeah, I'll use a bit of everything.
So, I think Training Peaks for some day-to-day bits. Starting to use more onto Vector cuz they've integrated more, particularly for interval creation. So, I don't have to do that.
It can pop up, use their system, bit of AI, does the sessions for me. So, we're trialing it with Chris at the minute just to kind of see before we do any migration work within that. I use WKO for kind of some more of the custom metrics that I want.
And then I've obviously a lot of my research is done in Excel, so I can actually check point by point what's actually happening. So, now I've built that into an app, so I can just pop the fit file straight into it and download, and then make decisions. So, it's a bit um cumbersome at the minute, but I think it will slowly cuz some of my analytics are only available with my analytics, if that makes sense.
There's a few things that I've made that's not available anywhere else. Yeah. And how much do you think the athlete needs to understand this?
Cuz, you know, I'm getting to grips with it a little bit. Obviously, I'm playing a little bit of devil's advocate here. I understand some of these terms a little bit better than some of the questions I've asked.
But most of our listeners are totally new to this conversation. They haven't read any of the background paper. They haven't read your research.
If an athlete's coming in to work with a coach who's using this new world view, do you think it's essential for them to understand what W balance is, W prime, critical power? Or is that going to be a conversation that evolves? And just like when I had no clue what training and stress score was when I came into the Training Peaks world, you'll you'll just get that over time.
Yeah, I think it's really good I think to have be open from the start, you know, with the what Cuz some are quite curious and want to have, you know, latest scientific input into their kind of training or what's been happening. And I think I think cuz a lot of people that I work with know kind of the research area that I've done. They know me from like the W Bal stuff.
You know, I've presented the last few years at Science and Society. My name's kind of gone around quite a bit on, you know, W Bal. But it's then my job >> Middle name now, isn't it?
Basically, yeah. It's now my job to kind of educate it to the athletes in a usable way to help explain. It's just another It's like another tool.
It's another framework to view something that's happened. So, let's say we analyze again a race. We have, you know, you can have your peaks.
You can see the power duration curve for that. You can see the power profile going with the elevation. But you put in the W Bal, you can see then how is their W prime used?
How is their battery or that work capacity used through that race? They might have said, "I got dropped at this point." I was like, "Well, you've done four efforts already.
You had 20% left before you started the main section. This is probably why it happened." I guess my hesitancy was, and I probably started hearing these terms or I don't know, 2 years ago, maybe a little bit more.
And I didn't really invest into learning about them because I was like, ah, the the old Coggan Friel stuff has been around and it's time-tested. It's been around a long time. I was like, this seems like it's quite emerging.
Let me just chill a minute and see if it sticks or something comes in to replace it, if it's worth the time investment in understanding this stuff, or it's just going to be usurped by the next fashionable thing that comes in. And maybe I'm still not sure of the answer to that because is are we going to be talking about ventilatory threshold with time where I'm coming in in the next Mhm. few months, and this is a, you know, a redundant framework for thinking about training, or do you think this is here now to stay, and time where it just gets layered on top of this, or maybe you don't think there's even a place for respiratory analysis?
I think with everything it's it goes back to what questions can you ask from it? Does it allow you to ask better questions that nothing else explains? Because at the minute, there is nothing else that measures our ability to recover.
So, actually, it's it's filled a void that's not had something in there. We can intuitively, you know, say, "Oh, yeah, I didn't recover that well." But what is recovery?
What is well in that situation? How, you know, are we trying to explain performance? So, like time trialing.
Well, we know what the maximal 40-minute power might be. You're not going to ride at that 40 minutes. Imagine you've got two climbs, there's two descents within that, and it's an uphill finish.
Where we know that's probably going to be key. Let's say it's 3K to the finish on that. How do we then use the power duration curve to maximize that within that?
Like cuz that power at the end, let's say it's a 5-minute effort. Well, we're not going to do our 5-minute max power cuz we've been riding pretty consistent for Yeah. 30 minutes before that.
So, that model is the only thing at the minute which helps explain how we can our performance is occurring through that. >> Has anyone built out a good pacing model for this? Like Best Bike Split was, you know, widely used for a time trial and triathlon for a long time, but it kind of just gives you a linear pace distribution with a small bit of change based on weather.
Yeah, so, funny you mention that. Now, this wasn't set up at all. >> That's actually I don't even know what you're going to say.
So, Check out my new program. It's 9.99.
Special offer. So, these are the exact questions I've had across my PhD. So, I wanted to rather than go purely down an academic route, what are the questions as a end user is going to have?
Can we use it for this? What sort of things do we need to be aware of? So, I've got a chapter in the PhD which is an applied chapter, which will go into two separate papers.
So, we took, can we use the W Bal model to pace a time trial? And we utilized right, rather than work out all the aerodynamics, I'm going to take a shortcut and do what most coaches will probably do. And then let's stick it in Best Bike Split.
But then let's run with the W Bal model with it. Then let's optimize the W Bal by specs. Best Bike Split with the W Bal model, and then see what tweaks we can do.
So, what we did, we did a 3 and 12 in the lab. Then we did a 3 and 12 on the road, all in TT position. We did a mock TT for 40 minutes of a roughly sort of thing to try and then individualize the model.
So, we then put those parameters through the W Bal model off the Best Bike Split. So, we put it through first, and the split didn't give the pacing pattern that we wanted. So, then we I wrote a script to then pull the pacing plan into a power profile, so into a power like file.
Altered the back half of how we wanted to suggest the pacing. Gave the athlete the pacing plan. Ran it with 14 car set up radio cuz it was a closed road TT, so we did it for the national TT champs.
Did the pacing plan based off that, so taking all the considerations we would do within the team. So, minimal testing rather than three trials to do CP, three trials to individualize it, and a map test. That's like 7 days wasted.
So, we just did a let's say a morning session, then an afternoon session. And then run that pacing plan. But then, how do we deliver that pacing plan to the athlete?
Cuz it's 88 segments. Well, we basically just then did give a voiceover, like keep it in control for the first half. Then actually get over the climb without going into the deep.
>> Okay, you're not doing push notifications through your bike computer? You You probably could, but obviously, I'm not allowed to see the data from I think with UCI regs, I can't see the data that the rider's doing. Real time.
And I also don't want them constantly looking at the head unit. I want them basically locked into position. I'm on the radio in the in the team car making sure they know where to go, when to lift their head up.
>> have heads-up display in helmets? >> Oh, that'd be cool, wouldn't it? It's coming, for sure.
Unless UCI ban it. Yeah, UCI are definitely going to ban it. What I What I want is, I'm on a different tangent here on AI, which I do want to get onto as well.
But for augmented reality, I want a set of sunglasses or I'm going for a Strava segment, full Champs-Élysées crowd on the sides, and I'm sprinting for a virtual finish line, even though I'm out in the countryside on my own. That would be That'd be pretty cool. I'm surprised it's not done.
>> but worth the risk, I find. It'd be good to see it in Cuz in reality, they they could see the head They could have it on display on the head unit as well. >> But heads-up display in a helmet, like, you you know, one of the the POC helmets, a heads-up display on that.
>> of visor space. Much safer than constantly look Like I remember pursuit I'm surprised it's not got a camera in the top and like But I remember pursuit and having the aerodynamics coming in, working on my position. And my feedback after like a couple of test laps was, "I can't see anything.
like I can't see where I'm going. And his reply was like, "Well, "Why do you need to see where you're going? Just look at the black line and go around.
" I was like, "This is terrifying." But that's where it's got like you look at Remco's position. He can't really see anything.
Mhm. But they're calling out in the voice over now as well. Team car will be the eyes and ears.
So, you've got to trust, you know, your DS in the car or the coach that's on the radio, which is kind of what we did. And yeah, we did the full pacing W bar model of best bike split, which did and didn't work. Just we got the time constant wrong through mock testing.
So, we've now got a framework of how we can confidently best optimize it now with basically no testing sessions, which has gone into my new model. So, cuz the best bike split just did it as like normalized power. It didn't take into account the recovery period.
So, it's it's going to be moving to a point where if you put the W bar information into it or the arrow optimization for the speed, the elevation, it's going to come to a point that you can probably do that between riders now. Yeah, the arrow one's tricky. Cuz cuz the CDA really affects the prediction on it.
And even more so in a team time trial. Yeah. And then you've got team rules within like pairing these rules where actually only one rider is whoever when you cross the line you finish.
So, you can So, you can optimize W bar profile. I think we need to be doing more of that with cycling though like playing around with the rules and funny stuff like launching people into four lead outs for the last K. Mhm.
So, it's definitely going to be integrated into particularly from a team's perspective of that and kind of your everyday coach. What's quite interesting on this is I spoke to the Wahoo CEO actually just last night. And he's talking about my criticism of GPS head unit on a sweater it's the you know, I'm using a Hammerhead Karoo brilliant bike computer.
Wahoo he's talking about the Ace cool bike computer. I was like, "At a point like I don't know what iPhone is that? I haven't a clue.
Do you remember at the start you got an iPhone it's like, "Oh, it's music on it. You got the next one I'm definitely getting the new one there's a camera on it." That you're definitely getting the new one you can do email that's got social media.
At a point I'm not sure where it was iPhone 7. What's the difference between iPhone 7 and iPhone 9? And iPhone 4.
You know what I mean? Like at some point we just all went iPhone could bring it unless they bring out something that's projecting holograms in the new iPhone. I'm just staying with that one until the battery dies cuz they're basically all the same.
So, I was asking about our our we our we there with head units like they basically do everything I need my head unit to do now. It's like a new Even gaslight you on some of them. A new firmware updates and it's now it's showing you your heat sensor or whatever you need to be.
Well, the hardware is not really changing. And he actually disagreed completely. He thinks there's a whole era of aerodynamics which are going to come into head units now.
The new Wahoo Ace has vents in it to figure out air speed. That they're only scratching the surface on starting to build the software to use that hardware. And that's going to be the next frontier in head units.
So, this again again I've got that that's a like a real time coefficient of frontal drag Mhm. They're starting to get interesting. They're obviously not UCI legal but to play around with because if you're in an echelon you can understand the wind's not always coming directly from the side.
It's like if you move it back here versus centimeter to the right versus two centimeters to the left you're changing your wind exposure totally. And that's going to be really interesting for modeling in team time trials. I think it's it's going to really kind of change.
I think one of the things that surprised me that hasn't happened yet is the recording frequency. It's down to still 1 hertz. So, one data point for every second.
So, as soon as you go above 60 rpm you're starting to estimate. Uh does SRM not go a bit more than that? >> So, SRM is 20 hertz but that's done the SRM science which you need then the head unit to be able to capture.
That head unit is money as well. It's like you look like a boss. >> you can still get on eBay which are like the PC 7s which is sometimes Oh, which is 7s to old school.
I started on PC 5. PC 4s out. Wired was unreal.
Everyone was hating on wired. Like it worked 100% of the time. If it didn't work your wire was [ __ ] Yeah.
Now it's like go out connect my parameter 92 parameters discovered it's like, right. And my battery's flat so I can't ride today. Anyone know what what what's the last three digits of your parameters?
Like don't know. It's like, "Ah, yeah." Wired gold.
Yeah, I we had it on some of the testing bikes and some of the old SRM science which I think >> Yeah. Some of the original like PC head units which could record. I think so the new PC 9 frame boards on SRM can do like 200 hertz.
So, great for looking at sprint data. Not so great if you need don't know where you're the way home cuz you've no map. How is it no map?
I had a really nice PC 8 for a bit it's like shiny red. It was lovely. The screen was great.
The num- the clarity was fantastic but yeah, no map. I can't believe there's no maps. I also kind of love it.
Like it's like gives you I the you leader founder on the podcast he's like, "We give you what you need and none of what you don't need." I was like, "Yeah, depending on the situation I'm in here I might need a map. Badlands try doing it without a map.
" Yeah, and and they were so simple cuz you could customize the names of the different frameworks. So, I used to have an interval which would like I had some profanity words in there to give me a bit of motivation when I click lap. Nice.
Which you could just change and they it would just they look cool. They did look super cool. They look like when I think of them I think of Greg Van Avermaet as Olympic champion and his SRM 7 or 8.
I'm like, that just looks like he's balling outrageous. Yeah, and these obviously sponsor a lot of the teams kind of with like the HTC Highroad team. And obviously Cav would get like the colored ones and it's just get the yellow SRM I think.
Do you know all the first riders all bought them? Like Greg LeMond bought his SRM. Good bits of kit.
I still keep an eye on eBay just in case I can find like a cheaper one cuz obviously to buy they're quite expensive but they a lot cheaper on second hand now. Cautionary business tale in there as well that maybe SRM and Polar need to look at where like if you have 100% of the market like you really shouldn't be losing 99% of that market share. Yeah, I mean for some people like just having a single sided parameter for a couple of 100 pounds versus one 2,000 pounds which gives more clarity and detail.
Well, do I need that? Yeah. Yeah, it's how you decide to deploy your financial resources.
So, mhm. How would you think about and rider comes into you? Self coached.
Mhm. Wants to move across to this model. Like what's that seven day starter pack look like that you think they should invest in whether it's like resources to go deeper and understand this or software to get started on this or testing you need to do.
Like what's a seven day kind of I think yellow brick road to get them going here. First it depends on what like training platform that they've used before. Obviously if you use Peaks it's going to be quite hard to access the CP and W prime model.
Obviously they've got their own version on WKO5. But unless the athlete has that then that's redundant. So, then it's kind of the other platforms.
So, there's obviously Intervals ICU I think that have the W bar stuff. Vector has it on there as well. So, those are kind of two things if you want to have it accessible as an athlete.
There isn't actually much software online as far as I'm aware I've found where you can just really run the W bar model separately unless you know, it's through our app that we've had. >> through ChatGPT? I don't think the maths is going to be quite correct within that.
Like I've played around within that see you know, kind of what it thinks the modeling is and it comes up with different variations. And then for it to integrate it is quite cumbersome as well. So, unless you're quite a whiz with Excel and you might be able to then create your own.
But then if it's an Excel it's ChatGPT will create Excels for you now. It's quite a laborious process and then you don't know if you're right as well. Cuz that's the problem with it.
Whereas we've got kind of a we have a research grade one which kind of someone helped me develop which then we kind of then altered and changed kind of as we went. >> No, my app is free and within that. So, we can put a link to it up yeah.
Yeah, it's still a bit buggy so it can only work with fit files from it won't work directly from your head unit. It has to come from like the training platform cuz I haven't worked out the fit profile that it reads onto it cuz it can be a bit choppy. But that will run basically you put your CP W prime in your A and B parameters for the model for tau can just be preset cuz it comes from our research paper which I think there's a link to in so, it's an app inspired by my research and it gives you an idea of some of the tools and power that W bar can do.
Or if you're into yeah, Vector or Intervals you can use it within that framework. And so in that seven day for someone getting started like if they're trying to understand how to use these tools for personalizing gaps between you know, typically they're heading out there and they're doing four by four minute VO2. And now they're trying to understand is that four minute gap long enough?
Is it too long? How do they begin to figure out and answer those questions for themselves? So, obviously you'll then let's say in that first week you've done like a three and 12 test.
So, you've got your CP and W prime and you can then see run the model just as it is. What in terms of your W bar profile does that four by four look at? Are you left with very little left?
So, probably hitting the session quite well. Is it negative and you're not Is it negative and you're still completing it? So, we need to change the model parameters.
Is it actually completable but you're struggling but there's still 50% left? So, again we need to change those model parameters. So, you can use that as like a framework to fit the model to you.
And then any future sessions where I'd say, "Okay, I want to add a fifth effort into it." Is that going to be possible for me? So, it's kind of baby steps on it just asking better questions cuz some athletes might find actually I don't need to change too much in the sessions.
I know myself quite well. It's intuitive. Like it's just that framework.
Like some athletes might know each other quite well. But so is the software packed to answer these better questions? Is that vector Yeah.
and your app or is it vector your app plus something else? So vector for your kind of like day-to-day. So if you create that session, it will run that profile for you.
My app will only run on what's been done so far. So you can use it to kind of like we do have an interval creator which will show all the different models. I think so on my kind of once you find my profile you on the stream that you can see the apps that I've made.
Um I'll I'll send you a link so you can see them. And basically you can run the different models and you might say, "Okay, I've done this before. I know my CP and W prime are these values, which one fits me best?
" So you can use it as a predictor. >> And so can I take my race file down cuz I think that was one of the most interesting applications for me. It's like I can take my start to get learnings from my race file going, "Look, you're not able to follow those attacks.
You're going too deep there or you need to attack more because you're not, you know, you're playing fully to your strengths here." Can I analyze race files in that same way? Yeah, so you just stick the fit file into it.
Obviously input CP and W prime. You can leave the model as it is cuz I've just put on the most common model that fits most most people I should say. And then yeah, it will run through that.
So it gives you a W bar profile. It will then tell you where your top three recovery bouts are, depletion bouts are. It will then just run the W bar model on its own.
You can pick apart how much work you've done above CP, total work done, kind of all the questions I've asked. I've built like a match calculator in there. So how much So for each match, I've defined it as a percentage of CP.
So only counts above CP and how many of those have I done? What's the average intensity of those? What's the average duration and how much W prime am I depleting per match?
That's cool. I like the match analogy cuz it's something that we you know, it's almost the same thing as well. >> Um we're doing a crit study at the minute.
I'm just finishing the data analysis for it. It's probably going to be around in criterium racing in the UK between cat three, two, and one races up from regional A to national B. On average it's each on average every effort is 15 seconds long at 135% of CP as a rough rate.
Well, so if I needed to kind of create a interval structure to be race specific, I might use that to help influence my decision. What's then what's the average recovery duration within that? So we can then look at that.
Go back from time trialing. So it might help justify decisions with particular intervals. So you know, should we be doing 30 15s, 30 30s?
Well, actually most of racing is around let's say or 15 15s. So actually that might suit might suit me better. Might not necessarily be the case, but actually where allows us to make a decision towards that.
That's class. I I really like that. And if you're looking at I know you're not a huge fan.
We talked offer about AI and it's coming along. You're not too enthusiastic that it's going to replace the coach. But if you look at it through the lens of it's an assistant coach.
What ways do you think that's going to add a new layer to our understanding of the massive amounts of data? Cuz we you know, we talked about time where already. So if you start bringing in lactate values, all the stuff we talked about today, critical power, W prime, you layer on that, your heat sensor.
Look, my heart rate strap is getting like something a US Navy SEAL would wear when they're going out like full military full of ammunition cuz now it's got time where on the back, heart rate strap, got your core sensor, core sensor and then your two master flow bio sweat sensors coming as well. It's like we're getting a lot of data here. So to make sense of all that, there's been a huge race in the world for behind the scenes for the last 12, 24 months for our data analysts to really unpack this and Vasilis was talking about this where he just gets a data a package from his data analyst to summarize what would have taken him hours and hours to parse through himself.
What different ways do you see AI being that beneficial sidekick? So I think it allows to kind of put a lot of that grunt work out. So like collecting all the data, aligning all the data.
So you can go actually like we've got this Well, we've put all the input parameters of all the riders. So let's say we want to take it over a grand tour. What factors seem to be influencing W prime depletion throughout the tour?
Where and we can maybe mark up, you know, where riders get dropped, certain things have happened. What factors come into play while rider X, let's say they've done X amount of work, the core temperature seems to be getting to this consistently each time this happens or they're dehydrated at this point and we know that they've missed a couple of fuel stops. So it allows us to potentially you get rid of all that grunt work to do it quite quickly, efficiently and ask it questions quickly.
What across the team seems to be happening within these riders? Could it be core temperature? Could it be, you know, they're were using W prime too much early on in a race within that like There's almost an endless amount of questions you could potentially then ask rather than going through each file trying to mark it yourself, do this, understand that, make a spreadsheet and a chart it.
I think that's where you can get ask questions very quickly to see if there's merit then to explore those to put interventions within teams to and maybe invest in more equipment within a particular area, justify those decisions. We've got five, six Welsh MVP Poggy Yonas four, who have I missed there? You're British, aren't you?
Shall I try and throw Pidcock into that one? Not a hope. Uh look, we've four, maybe five, maybe I missed someone there.
I Do I call them X-Men, superheroes of cycling? They're mopping up 90% of race wins between them. Pedersen maybe you could go close.
Don't throw him into that. I don't think quite there. Is there a worry for any of these teams that this new level of analysis will begin to find a weakness in Poggy?
I'd love to say yes. Like, you know, is it that superior ability to recover? Does that change after that?
So if we can use that to potentially understand what might be happening, how can then we manipulate that? How can we exploit it? And can you train a rider specifically on his weaknesses?
Yeah, and trying to understand then what's happening or I think I did I did some analysis for a team when I gave a presentation and I kind of gave um some synopsis of what was happening in the W bar profile in the UAE tour. So when I think one of the riders attacked, obviously they went with them and they could no longer maintain the power output that they were so they dropped. You saw this happen on the W bar chart.
But two different riders, one had to drop more below CP than the other one to maintain to the finish. So one lost 40 seconds on the other rider. But then would we have been better if this is for GC, not the stage win, of actually going, "I know I can't maintain that for this amount of time.
" So let's say you had this on your head unit, go actually, "I've got 6 minutes and 30 seconds left." But actually I know this is an 8-minute climb. >> seen Stephen Barrett was here H2R at a performance and we've seen Ben O'Connor trying to follow Poggy.
Mhm. And we're all watching Tour de France and Mayes going, "Oh my god, Ben O'Connor can follow Poggy. No one else can follow him.
" And then he takes the cyanide capsule and so out the back, finishes way behind that group of I can't remember who was in the chase group, but you could fair guess it's, you know, usual suspects, Roglic or so. Finishes behind that group. That's an obvious example of all of us looking on, "No way, he's holding Poggy.
That's going to end in bad news." But you could you could almost model that if you had the Yeah, so you could say, "Well, actually this is unsustainable." But then it goes and goes, "Well, you know, some models are wrong.
Most of them Most models are wrong. Some are more wrong than others." And you can't take away that art of bike racing as well.
So I think this is where it comes in nicely with the kind of new influx of sports science within teams is not is for me, let's say if I get integrated within a team, it's not to go, "The model says this." Like this is the likely outcome, but also trust your legs. How should I pace it?
Use feel. Don't neglect those inner sensations. It's just another tool of using it.
It doesn't It's not you know, I think it was like Cav couldn't test particularly well on the SRM yet most amount of Tour de France sprint stage wins. Like you can't take out that desire, that passion, that ability to just suffer. >> Someone said basically every time they test him, he got thrown almost thrown off the British Academy.
And then he just win the next weekend and they're like, "Obviously we can't throw him off now cuz he just won again." Yeah, and it goes back to, you know, understanding the limitations within that model in any sort of performance model to make those decisions cuz a rider wants to go long and you know, unless it's going to cost the entire GC, but sometimes you've got to put your cards down and play them to win. Yeah, cuz there's almost a like there's a there's a heavy sports psychology piece in this.
It's it's Cav, it's it's the goats. It's Cav, it's when it's Michael Phelps won his last ever gold medal by 0.01 of a second.
There's nothing physical about that. That's a he's set a standard for himself. I I'm the greatest of all time.
I win golds. That's the standard. I won't fall one step below that standard on the podium because this is the standard I've set.
There's there's something there that's yeah, it's so small, it's so nuanced that I don't know if models ever But let's say you know, I end up working completely hypothetically, you know, I'm in charge of the pacing for Remco, let's say in a TT. Congratulations. There's no way From Joel Lavery to Remco, you've quite the switch.
But there's no way that I need to tell him how to pace a TT, right? Using the model framework. There's just no There might be able to do it and go, "Okay, mate, it's unlikely to be able to go faster, right?
" But then some riders that can't pace it particularly well, it gives us that framework. >> Well, the great example for Remco there is remember last year he backpedaled on start line and broke his power meter. Mhm.
And he paced that TT to what did he say? Within four watts of his goal. That's just that's it's like Sean Yates in that year in yellow documentary I talked about where the interviewer is asking him, what do you say to Wiggins before you know, a time trial like this.
He's like, you say absolutely nothing to him. He's like, he's going to have wobbles. He's going to mess up.
He's going to make a total balls of stuff in the Tour de France, but it's not going to be today. Yeah. He's like, this is what he does.
He locks in for time trials and he's the best in the world at delivering effort over set distance. Yeah. And it's it's then bridging that sports science with with feel as well, trusting that athlete.
And again, as always knowing that limitation within that. Like, a true you know, if you practice TTing, you get good at pacing it in it. You don't need a parameter.
What was it? Um Taylor Phinney when he used to do TTing, I think he said on his video it's quite a few years ago. He used to go on speed and feel.
Didn't need his parameter cuz it all to me speed and that is your performance currency, not power. I had Lachlan Morton on last week. Doesn't use a parameter or heart rate monitor.
So, it can help, you know, those riders that might be new to it as a good idea of pacing to most optimally do that cuz eventually it should be a case of well, the model should probably align with what they're feeling if they're very good at it. To finish up. If somebody is falling kind of off plan, like they're I don't say over training but they're not getting the adaptation from the sessions.
I guess the historic problem for us is there's a big gap between stimulus and adaptation. There's a big gap between me doing a block of threshold efforts and me understanding if my threshold is actually improved. Mhm.
How do we close that gap? How do we begin to close the gap between stimulus and adaptation to know that 2 by 20 is a better protocol for me than 3 by 12 or 3 by 8? I think that will become from like time and experience cuz obviously there's multiple ways that as the saying goes to skin a cat.
But as a coach to understand that rider where it's periodic testing within that. Should you be seeing then from a sustainable effort RPE going down? Should heart rate becoming more stable, less gain throughout the session?
You know, if you're not responding, okay, what other factors might be happening on in life? Actually, are we optimizing nutrition? I've had things where we've had a nutritionist involved in the rider's training and it's transformed it.
Yeah. It's not the training. It's that stimulus comes from everything as well and adaptation comes from, you know, are we sleeping well?
Are we It's a lifestyle choice and on many occasions within that. Alex, you're going to have to come back for part two. I can.
That's fine. This is going to get uh Yeah, I think I need to go away into the real world and use this stuff and mess up and understand the limitations and have those aha moments where I go, "Ah, now I understand." I think this is a podcast I'm going to have to I rarely listen back to my own podcasts.
I'm going to have to listen back to this one. And I would suggest listeners do as well like a couple of times and really dig into this cuz the transition from as we've termed it the old world to the new world, it's not going to happen overnight. It's going to be a lot of conversations, a lot of trial and error, but I really thank you for traveling across for your time cuz I think it's a super valuable conversation.
Yeah, it's it's it's I think it's just such a nice framework that you can I mean, I've dedicated 4 years of research but I obviously just finishing up the PhD now with within that cuz I think, you know, I've been the end user now I'm the user creating the research to ask these questions of how can we understand things better? Like, take Chris for example on Badlands. We ran his um profile for it and he went over threshold four times and used like 6 kJ of work above CP cuz obviously for him we shouldn't be using anything really He definitely went over it on the first climb, didn't he?
>> But you could see the exact point he says, "I can tell you exactly where that happened on that race." It just gives a bit more context to it to ask those better questions. Yeah, I I we started right at the bottom of a climb and I was just looking at lads going like forget CP, W' bell.
Listen to their breathing on this climb. You're like, it's an 818 km race and you are breathing out your hole 400 m into the course. And we've used it to like Chris is really good at pacing.
We've come to the conclusion to him and for him himself to trust himself. Not get caught out by pushing on cuz actually in the back end of the race, he'll be feeling strong and be able to overtake those riders that have gone like, you know, like Leadville 100 was a short race for him. >> Yeah, oh he's he's a freak.
I It's so interesting chatting to him. I'm not sure if you've had a chance to listen to the podcast or if it's even out. No, I think I have.
Yeah, I've listened to bits of it. Yeah. But my experience and his experience inside the same race are just so different cuz he obviously finished in 34 or 36 hours or something like that.
>> Mhm. I was like 5 days out there. These are two totally different efforts.
But and previously before going into the race, I would have, you know, almost dismissed it. Oh, it's 5 days. It's kind of cruisy.
It brings in this totally different type of fatigue that I've never felt for. I've seen people with like a broom duct taped to their holding their neck up cuz their neck muscles just giving up from being that many hours on the bike. I've ultimate respect for anybody who's in that Yeah, that's long for needing a broom.
We did a I did my undergraduate project on 24-hour mountain bike racing and we came to the conclusion the lineup we had a single-sided bottle cage. So, eventually we had really bad back pain after about 18 hours just from going in the from the same side. It's those little questions that I love about like applied science with it.
It felt the exact same experience. Not with a single-sided bottle cage but I broke my scapula racing in France on So, I default I haven't got brilliant external rotation on my left arm. So, everything I took out of my pockets is on my right side.
I felt it like a tightness, soreness that I've never felt there from just that repeated going into the back pocket with the right hand all the time. I did something with my I did the ligaments in my collarbone after trying to get a bit too sideways on a tabletop jump once and then to the point I could I would struggle to get into my back pocket so I had to use my other side. The things we do.
Alex, I really appreciate your time. Thanks for having me. Till next time.