Why Your Endurance Training Workouts Need Athlete Profiling

Paul Warloski (00:25)

Hello and welcome to the Athletes Compass podcast where we navigate training, fitness and health for everyday athletes. Today, we are in a way continuing a discussion we had about the athlete profile and how the profile can help you improve your performance. First, some background. Dr. Laursen and his colleague, Martin Bucheit at Training Science created a free email profile primer course that we'll link to in the show notes, but then

Paul and Andrea Zignoli, who's one of our guests today, part of the Athletica programming team, wrote a helpful blog post about power profiling that we'll also link to. Today, we get to talk with both Andrea and Paul about how Athletica creates a profile automatically for you, but we're going to get into a bit of the science behind that as well. Paul, let's start out by understanding the athlete power and speed profile. Let's talk about that.

Paul Laursen (01:24)

For sure, thanks Paul. So look, I guess at the end of the day, the importance of this, it was alluded to in the last podcast that Andrea featured on with respect to the workout reserve. But coming back to you, the listener, and assumingly that you're on Athletica, the reason why we need to...

The reason why these profiles are important is because we need to see who you are. We need to see what your individual profile looks like. We need to know how hard you can work, for how long you can work, whether you are punchy and you've got really big sprint power, or whether you are more like a diesel engine and you can go all day long.

So there's so much that we get and it's almost like your exercise signature or fingerprint when we look at your profile. So that's the purpose of it in the end of the day. also, it has so many purposes. It also comes back to now, this is one of the big ticket items that Andrea and his team have worked on in the backend, but that profile actually comes out and forms your zones.

So there's so many important aspects of it. And because it's so important, Martin and I put a free mini course together called the Athlete Profiling Primer that you can take. And we'll put that into the show notes here so you can click through and get a little bit more reading into this as well. But maybe I'll kick it over to Andrea next and he can tell us a little bit.

a little bit of a review, Andrea, about what we're looking at when we look at someone's profile. What are we seeing in there? What does it all mean at the end of the day?

Andrea Zignoli (03:21)

Thank you, Paul. I think that with the profile, as you said, running profile, running speed profile or power profile for cycling or power profile for rowing, we are drafting a locomotive profile, as you call it, which encapsulates the idea that the intensity that you can sustain for a given amount of time

is degradating with the duration of the exercise. So you can look at the profile as the map that can tell you how you are able to sustain a given exercise intensity for that exercise duration. So if you connect those points between the exercise duration and the intensity of the exercise,

Paul Warloski (03:58)

Thank

Andrea Zignoli (04:19)

which is described in running with the running speed in cycling with the power and rowing with the power, you can have an idea on how much you can, how long you can sustain that given intensity. Now for different individuals, we say different horses on the same track, the shape of the

line that you can use to connect these points on these graphs is different. Some of the lines are steeper than others, and you can clearly see that, as Paul suggested, is a print of your abilities and speaks about your ability to sustain a given exercise. You can see like

You can clearly see the difference between some that are able to produce very high power or exercise intensity for a short period of time. And ideally, those are the individuals that are best in sprinting abilities or short distance efforts. Then you can see the bigger engines that are able to sustain good loads or good intensities.

Paul Warloski (05:41)

you

Andrea Zignoli (05:42)

for prolonged durations. So this is a generic guideline, but of course, if we're not speaking to like the best athletes in the world, mean, this is true for like generically the best efforts in all the disciplines, like the sprinters notably are

very good in delivering very high power outputs for very high speeds for very short durations. And the climbers, all the time trialists are very good in keeping those high intensity for prolonged durations. But when it comes to an individual that we do not already know, as for example, Paul was mentioning someone that is signing up on Athletica and all bolding right now, we don't

know anything about those individuals, so we needed to collect enough examples to best describe their ability to sustain the exercise for a given duration. It might not well be that if very high power values are present for short durations of time, we can conclude that, those are

Paul Laursen (06:54)

Yes.

Andrea Zignoli (07:08)

fast-reach athletes very good in sprinting because we might see also very high values for long durations. And the simple answer is that we don't have perhaps a reliable profile because they were not given their best effort for a given amount of time. So to get to know the individuals that sign up from...

today or tomorrow on Atlantica, we need to collect those maximum efforts so we can draw those lines that connect the points and we can really understand what is that they are best at. And eventually, for a specific race or a specific duration, we can guide them in getting better in those aspects that are ideal for that kind of race. It might well be that

A fast twitch athlete, which has very high power values for short durations, might want to take part in a very long trial run race or a very long running race. And there's wrong with that. It's just that something perhaps has to change in the load distribution for their preparation. that's pretty much what we're doing.

in Athletica we are using well-established models that can tell us on average for an athletic population how ideally the different points should lie on the graph. And by finding the parameters for those equations we can extract meaningful physiological parameters for those athletes.

And when we compute the ratio between those values, we can infer something about the individual that we have in front of us. By taking the values of our user base in Athletica, we know what is realistic or not realistic. So if someone is off the chart, it probably means that

Paul Laursen (09:10)

Mm-hmm.

Paul Warloski (09:14)

you

Andrea Zignoli (09:29)

the efforts that we included in the profile were not complete or we just needed some more effort to derive a better profile. So we highlight that in Antarctica and we try our best to understand what's in front of us.

Paul Laursen (09:50)

Yeah. And that's one of the cool features that Andre and his team have created. And you can see this yourself if you're an Athletica user. Just go to your Power Profile or your Pace Profile. it's kind of a little hidden right now. And we need to probably do a better job of bringing forth this important information from our AI to coach you and to help your coaches coach you as well.

Paul Warloski (10:11)

Thank

Paul Laursen (10:17)

we can actually tell you when something is missing in there. And I've recently gone through this whole process myself in my own profile where my profiles told me, Paul, you're missing some 30 to 60 second efforts. Paul, you're missing some two to five minute efforts. Go out and give me that. And I've gone and completed those to construct a better profile and to get so that the...

the AI gets more insight into my own fingerprint, my own signature, so it can help prepare me best. So that is a really cool feature that I don't know is anywhere else. So have you seen that, Marianna?

Paul Warloski (11:00)

Bye.

Marjaana Rakai (11:07)

have. And so as I'm coming

I've done my Ironman training. I've noticed my running profile, especially it's very flat because I didn't do, like I did some 30 second hard efforts on running, but I'm a little afraid to hurt myself doing the running sprints. So that's why my profile is pretty flat and it's giving me.

Paul Warloski (11:24)

you

Marjaana Rakai (11:42)

notice that you're missing those hard efforts. So now that I am done Ironman training, I can work on my profile a little bit more and do some sprint efforts. So that's my goal now to make a beautiful guess.

Paul Laursen (11:57)

That's nice. That's nice. Yeah. And of course, we have the systems in place where those individual high intensity zones are actually catered and looking. They're looking at those numbers that you get in there. So you can actually have individual zone updates for your zone 5, 6, 7, right?

high intensity bandwidths. Those are actually captured in there and get updated accordingly to how strong you're actually going in there. Is that right, Andrea?

Andrea Zignoli (12:36)

Yeah, to your point before, before I forget, we can debate back and forth whether having the complete profile on paper is beneficial for your performance or not. What we, we care that is that you have a complete profile so we don't get our numbers wrong. But, and we know that some high intensity at spring training, Paul, you can add more to that, but it's beneficial also for those that can.

Paul Laursen (12:56)

Mm.

Andrea Zignoli (13:07)

compete in long distances for sure, but we don't have enough evidence to say, if your profile is completed, then your performance will be optimized. It's about putting together the blocks and eventually, most likely, your power profile will be deemed complete, but we don't have the scientific evidence that a complete profile equals optimized performance. Do we, Paul?

Paul Laursen (13:10)

Yes.

Paul Warloski (13:20)

you

Thank

Andrea Zignoli (13:36)

So it's, yeah, okay, okay, that's good.

Paul Laursen (13:36)

No, no, we don't. We don't know, but look, you want to be that complete athlete, I think. In general, you don't just want to do one type of training. There are different contexts. mean, Marjaana had some plantar fasciitis. She needs to be cautious with high intensity training because she doesn't want a neuromuscular flare up with that plantar fasciitis. So, I think she played it safe and smart.

to train specifically for that event. You know, she's, you know, and we speak about this all the time. She's, you know, perimenopause and an older athlete. She's, but, you know, in her younger 20 something self, maybe, you know, she probably could have afforded to do some more high intensity surges in there. And she probably would have done them as well because she'd been robust enough to, for her body to handle those. So context always don't take the...

Paul Warloski (14:10)

Thank

Thank

Paul Laursen (14:35)

You know, always look inside the recommendations that you're getting from Athletica with common sense, putting it in context of you and your situation. But nevertheless, are, you know, because it loops back with the prescription and, you know, there are those high intensity prescriptions, it is nice to get accuracy within what your capabilities are.

Marjaana Rakai (15:01)

Yeah, that's a really good point. Do you train for perfect profile or do you train for your event demand? And then, like you say, Paul, you have to make...

some critical choices and you can do like you can work on those high efforts on different methods you can do rowing or you can do bike like I did all my almost my my vo2 max sessions on the bike instead of the run or I would run uphill.

Paul Laursen (15:33)

Nice. Yeah, yeah, yeah, no, exactly. You catered it to what you knew your body could handle to give you the training that you needed to get you to toe the line with the best shape possible for worlds. So yeah.

Marjaana Rakai (15:46)

Yes.

Paul Laursen (15:51)

I think one of the other things we really need to bring out in this, when we're talking profiling primer, there's, know, there's, there's a, there's a backstory a little bit in this and it kind of goes something like, you know, we are like Andrea mentioned it, you know, different horses on the same courses. Right. So when you look at everyone, all the horses, that's us on the courses, they're different. And some of us are, yeah, that twitchy.

Paul Warloski (15:51)

you

Paul Laursen (16:20)

Some of us are the diesel and some of us are hybrids. I think the majority of us are hybrids where we have kind of like aspects of both. think that's if you're gonna do bit of an assessment, that tends to be what we are. Where we have these mix of fast and slow twitch fibers. Remember the story, the slow twitch fibers laden with mitochondria, they go on and on and on. So you can go, these are the guys and gals that go all day and they don't fatigue.

Whereas the Twitchies or the sprinters and the hybrids are, you know, they have aspects of that as well. So, and you might ask yourself, well, why does that matter from a training standpoint? Well, it matters because the Twitchy people or the hybrids, they don't tend to recover as well as the diesels. The diesels are great recoverers. They're, the science is very clear.

Paul Warloski (17:11)

Hmm.

you

Paul Laursen (17:17)

where individuals that look like they have a, high, high propensity of slow Twitch muscle fibers that you can hit them time and time again with a lot of, interval work, long intervals, and they tend to just bounce back. No problem. They're the last ones to get over trained. The first ones to get over trained are the sprinters Twitchies. And then the second to get over trained are the hybrids. So.

Paul Warloski (17:20)

Thank you.

Thank you.

Paul Laursen (17:46)

Knowing a little bit about that for yourself, this is again coming back to the profiling primer, just understanding that a little bit. And then coming back to looking at your profile on Athletica, it gives you a little bit of insight into how good you might recover and it helps you potentially cater your own training to how much high intensity work you should have and the types of high intensity work. We should talk about the

types of high intensity sessions that favor both.

yeah, those are some of the key reasons why you might wanna know this.

Andrea Zignoli (18:24)

Yeah, to add a little bit of that. Yeah, I think one of the main purpose was to differentiate the intensity for the prescribed sessions and especially those high intensities. Let's talk about like very high and demanding effort.

Your work Paul about the anaerobic speed or power reserve, talk about that because you might have the same maximal aerobic power or speed, but you have a different peak speed or peak power. So if you prescribe in terms of absolute power values or absolute speed for two individuals with the same.

Paul Laursen (19:08)

Nice.

Andrea Zignoli (19:19)

aerobic capacity, you might not capture the right target intensity for that specific outlet. But if you take the percentage between these two different values, which is like the maximum aerobic power speed and the maximum peak power, you go in percentage with that, you're going to prescribe a much better

session for both individuals at the same time. yeah, I think like in the profile primer, you spoke about that also. And it's an advantage of having Athletica that can capture both of these aspects, what you can do in the very short duration.

Paul Laursen (19:48)

That's right.

Andrea Zignoli (20:12)

like between five and 300 seconds, which is like five minutes or less. And then those longer durations between, let's say, three minutes and an hour. So we are speaking about two different kind of models that can be used to describe how your performance and your ability to keep that intensity is degradating with the time.

especially at Atlantica we implement and we talk about that in the profiling course, it's the anaerobic speed reserve or power reserve and the critical power and we mentioned the critical power model previously in previous episode. And with these models we can detect those relevant physiological parameters that can be used to

best describe your profile in very simple terms.

Paul Laursen (21:14)

That's right. And prescribe accordingly. And again, so to our knowledge, Athletica is the only platform that actually prescribes to that individual level, which is kind of cool. And then you might also be asking yourself, well, what's the best type of HIIT training that I should do in accordance with who I am? And again, different horses. So say you're a diesel engine. Say you're like Marjaana and you just...

Paul Warloski (21:17)

Thanks.

you

Paul Laursen (21:41)

Energizer, I don't know. Are you a you a diesel Marjaana Are you hybrid?

Marjaana Rakai (21:47)

when I say I'm a hybrid.

Paul Laursen (21:49)

Yeah. Yeah. Well, do you know some diesels? Well, I used to coach Kyle Buckingham. So he's, and he's an ex-pro. He won Ironman in South Africa and he was clearly a diesel. The guy could just crank power for days. like again, so for him, he was great at short intervals, but he definitely benefited in long intervals as well. He could definitely handle those.

Paul Warloski (21:49)

Hmm.

Marjaana Rakai (21:54)

Cindy.

Paul Laursen (22:18)

This is the longer intervals. We're talking two to five minutes is your typical long interval. That can usually be tolerated quite nicely by a diesel engine type if you've got that phenotype. But not necessarily for the twitchy or the hybrid. They're going to get fatigued in that one. So in the future, we want to make sure that we have that option to change those. We highly recommend considering the workout

wizard and again, we've got some work to do in that one too, but the workout wizard eventually is going to be able to Switch and swap those different types of hit sessions according to your profile But yeah, so so yeah, those are kind of how you match so we always talk about 30 30s Marjaana always does her her hit session with 30 30s. Those are great for hybrids, which is probably why they're the

the most prescribed hit session on Athletica, but with Workout Wizard, that can go a bunch of different ways. We talk about those. And then, yeah, for diesels, they can handle, in terms of a hit session, more of the longer intervals, and they tend to recover a little bit more. So just important to kind of think about how this all sort of ties back to the profiling primer.

Marjaana Rakai (23:43)

I want to go a little bit back to an athlete who signs up with Athletica. And the second week is always the test week. He'll go do a few tests and those tests go into Athletica's brain. And how does Athletica detect them and build the profile and how, you know,

Paul Warloski (23:52)

Thank

Marjaana Rakai (24:08)

Occasionally you see these pop-ups that show that your thresholds have been changed. Can you explain the whole process how those are linked?

Andrea Zignoli (24:22)

That's a good question.

Shout out to Stefano, which is one of our developers in the back end. He did a very nice job in delivering this for for Athletica and his contribution was mostly around how we can

Paul Warloski (24:26)

Thanks

Andrea Zignoli (24:42)

Bring in the data that the user is providing and extracted the relevant information out of that. In this case, again, we looking for maximum efforts and perhaps those are hidden in long training sessions where eventually a, an athlete did some maximum efforts, but then they did not report that or, they.

did not isolate those sessions. So we needed to take them and collect these maximum points. With this algorithm, we will lay down these points on a graph and we see how these physiological models are fitting these data points. The most difficult part of that

is discriminating the maximum effort from those that are not maximum and to separate the bad data from the good data. So sometimes you bring in some data which is not good because GPS issues or like artifacts in your heart rate or whatever. So you have to delete those outliers. But you need to do that conscious that you might

be losing some important information. So that's the beauty behind the solution that we now implemented in AlphaEdit. It seems like we're doing a good job in discriminating realistic and maximum effort from unrealistic and unreliable efforts. So as previously mentioned, we are using these two models, anaerobic power or speed reserve and critical power. So from the first model,

we consider the points between five seconds each to 300 seconds each, so shorter durations, and we extract the maximum peak power or the maximum sprinting speed and the maximal aerobic power or maximal aerobic speed. So with these two parameters, you already are in the domain of the zone six and seven, so you can get those kind of intensity

thresholds. From the other model, which is the critical power model, you can extract the thresholds for zone four and five. And then eventually, if you

work backwards and you try to estimate what is the separation between zone 1, 2, and 3, you get that first ventilatory threshold or first threshold. So it's an approximation of this metric that you can estimate in multiple ways. As per usual, we're trying to implement the

the methodology that can give us the most robust result. So we are not sensitive to metrics with high variability. So if we use this methodology, we think we are close to a good approximation of these training thresholds without being too sensitive to metrics or variables that display high variability.

due to noisy data or like unpredictable intra-individual variability or

Marjaana Rakai (28:27)

And how often does it do that? Like how often athleticists brain checks the maximum?

Andrea Zignoli (28:34)

that's a good question.

To be frank, this is something that is still debated right now internally because some of us are more of the idea that...

athletes should go close to their maximum more often. Some others advocate for like

a less sensitive approach to the maximum effort. So at the moment, we are considering a six-week window for your effort to be included in the profile to guide your training sessions. Eventually, we can get to a point where different aspects of your profile are changing throughout time less frequently. For example, you're going to have your

basic threshold, let's say zone three and four, settles down for a prolonged period of time. And what is changing more frequently is those high zone six and seven boundaries. And we found that those higher boundaries, let's say six and seven, five and six are

Paul Warloski (29:49)

Thank

Andrea Zignoli (29:57)

more likely to change meaningful over time than the lower thresholds.

But this is just speculation. And again, what we're looking for is for the most robust solution that can provide us a good average behavior on the world user base. And of course, it's not going to be perfect for everyone. But again, there's no

Paul Laursen (30:18)

He

you

Andrea Zignoli (30:27)

correct solution right here, right now. We don't know the answer to that question, so it was not a good question.

Marjaana Rakai (30:28)

Thank

Paul Laursen (30:34)

No, it was a great question. I mean, I'm reflecting if you can go back to the former episode where we started with the, you know, the history of Athletica a little bit, Andrea, you know, I think when we, we thought that, you know, no problem, we'll go and create an AI coach for everyone, right? And it'll be easy. And it's just, these are, these are the nuances that you're just highlighting right now where, you know,

Marjaana Rakai (30:34)

Sorry!

Paul Laursen (31:02)

There's so much context that's involved to optimize for every single individual. And at the end of the day, we have to do what's best for the most individuals that we can. And Andrea is kind of just alluding to that compromise position where we're probably moving into. And that's the life of trying to make a mass market.

solution, training solution for athletes and coaches.

Andrea Zignoli (31:35)

Yeah, agree.

Paul Warloski (31:36)

Does Athletica automatically use the power or speed profile to prescribe power speed for workouts?

Paul Laursen (31:49)

It is, yeah. We're just through the zones, you know, again, the, and the, as Andrea was alluding to as well, the upper zones, zones five, six, seven, high intensity are updating more frequently in the future than the lower thresholds. The lower thresholds tend to be a little more stable. You know, your critical power, critical pace, those are all those things sort of below.

Paul Warloski (31:52)

Okay.

Paul Laursen (32:19)

terms of both a heart rate and a power pace, whereas that, you know, your ability to punch is probably, you know, that the changes we're seeing a little more rapidly. This is the other thing we spoke about this today on our, you know, our, backend meeting. And it's like, this is a, everything here is a work in progress. And we are learning every day from our users and how.

how things work across our thousand plus users. It's quite amazing. And I reflect on that as well, Andrea, when we kind of made the switch from academia, doing the studies in the labs exclusively with 10 subjects coming into our lab to now dealing with thousands of users online from that perspective. it's quite, I think the...

speed and frequency of learning is faster for us now as well. So sorry, just a little side

Paul Warloski (33:23)

You know, what do you, what do you see as the purpose? You know, we've been talking about this, you know, in, in, in general terms, but you know, the purpose of the speed or power profile for the everyday athlete. mean, especially one, maybe for one who's not using athletic yet. I mean, imagine someone who wants to run a half marathon and they just started using athletic or maybe they've never even heard of athlete profiling. What are some of the benefits of understanding?

You know, the phenotype, which, you know, were that you might describe real quick, you know, what kind of an athlete he or she is.

Paul Laursen (34:00)

Yeah, well, I would say the, you know, the profile helps determine, you know, what that athlete is. And the phenotype is, are you a diesel? Are you a, are you a hybrid or are you a twitchy sprinter athlete? Right. And the onboarding, if you're just, you know, if you're that half marathon individual coming in and onboarding, we're having a look at your history, right? If you connect up your wearable onto Athletica.

We're looking at that profile. We're also looking at your fitness history. How much have you been moving or training in the past? That goes into another, that goes into the banister model that Andrea spoke about in the last episode. but yeah, that's sort of where it applies in the everyday athlete, I don't think, and this is an important role that coaches play, right? When coaches learn about an athlete, well, they,

They ask all these questions. How much have you trained in the past? Do you have any data that I can look at to determine how fast you can move over a short duration or how fast you can move over a very long duration? And just to get a place to start. Well, we sort of have that now in a more streamlined way. I've never seen it be able to happen as fast as in the past.

where I would use a platform like TrainingPeaks and whatnot. It would take me a lot more time, but right now with Athletica, as soon as you've onboarded and we've got that two years of historical data uploaded on there, wow, that's powerful. We know exactly where you sort of sit, where your abilities are, and then we can move forward from there. So those are my thoughts,

Paul Warloski (35:53)

Perfect. Andrea, what do you think of that?

Andrea Zignoli (35:58)

Yeah, that's a good question. On my side to what Paul just said, I might say this is what is out there like in the scientific literature. And one of the things that I like of Atlantica is that it is, I like to say, hit science and code it. So it takes and builds on the shoulders of

many giants that know and studied a lot about exercise physiology and sports science and we build on top of that a tool that an athlete can use to me to learn about themselves and whenever you learn something about yourself it's forever and if we can be

the connecting bridge with what has been done in the past in this domain and give them to an athlete so they can use to learn something and benefit from it. That's pretty much it. I can be happy with that and live with that. So that is also one aspect of athletic that we care so much about.

Paul Warloski (37:21)

So here's what I understand from this discussion. The power or speed profile takes into account the different capacities of a cyclist or a runner by measuring their power or speed at different time intervals and creating a chart or graph from that. Number two.

The power or speed profile gives us a tool to more precisely gauge our efforts and performance, allowing us to better work on our weaknesses and capitalize on our strengths. And number three, profiling with speed or power gives us a type which enables us to get better for a specific event that we may have to change some of our load distribution, but this is also where

A coach can come in handy and help you understand what you need to do to get better for your event. That is all for this week. Join us next week on the Athletes Compass podcast. Ask your training questions in the comments or on our social media or as Andrea said on the Athletica Forum.

If you enjoyed this episode, we'd appreciate it if you would take a moment to give us a five star review for more information or to schedule a consultation with Paul, Marjaana or myself. Please check the links in the show notes for Andrea Zignoli, Paul Laursen and Marjaana Rakai. I am Paul Warloski and this has been the Athletes Compass podcast. Thank you so much for listening.