Athletica AI?

You are so very welcome, Paul. Its a great exercise for us to be able to explain our thinking around how we’ve developed the product.

I’m not keen on testing - it stresses me out! Is this going to be a problem?

This is where Athletica truly excels. Our system offers users invisible monitoring/testing by continuously recording your speed and power during your training sessions. From this data, we can determine your anaerobic speed and power reserve, as well as your critical power (see profile chart above), and track how these metrics evolve over time. Andrea provides a thorough explanation on this point for an Athletica user on our Forum if you want to go deeper.

As long as you occasionally perform maximum or near-maximum efforts in your training (i.e., key sets), your profile and thresholds are automatically updated. This means you don’t need to undergo specific, stressful testing sessions to keep your performance metrics accurate.

Additionally, when forming your training plan, you have the option to ‘skip your test week’.

During the onboarding process with Athletica, if you connect to Garmin Connect, your training history from the past two years is automatically downloaded. This ensures that your updated profile (including any hard effort that would count as a test) is available shortly after onboarding. Note it can take some time for Garmin to push all the data to Athletica.

With Athletica, we strive to minimize the stress associated with testing while still providing accurate and up-to-date performance metrics. This approach allows you to focus on your training without the added pressure of formal testing sessions for those like yourself that would rather not.

Does programming/periodisation take account of differing distances for ‘A races? For example, I have 2x ‘A’ races this year - one is a 70.3 and the other is a standard distance tri. Broad approach for the 70.3 might be an emphasis on highest intensity work further away from race day (maybe a VO2max focus block, into a threshold focus block, into a 70.3 race-pace focus block); whereas for the standard distance race I might reverse that order and/or take account of e.g. how much VO2max style work I’ve done earlier in the year (would another block be helpful or not vs more threshold or sweet spot style work).

Yes, it does, and this tends to be where Athletica excels relative to its competitors (we are told). In particular, the build phases of the periodized plans have specific race preparation sessions (principle of specificity; i.e., pieces of the event at race pace are broken up with recovery phases). Such sessions are progressed across weeks. VO2 sessions are consistently targeted on a weekly basis, but (as with the above thread on SI v LI) these tend to be short intervals as they are lower in general stress but target physiology (FT fibers, high SV) not hit in other sessions of the week. Of course, everything is modifiable using the Workout Wizard (which we’ll continue to iterate) or leveraging your training session library (hot tip - steal specific sessions you like from single sport cycling and single sport running for even more variety and interest), so you and your coach are free to go with the training philosophy of your choice. As we know, there are lots of ways to skin the cat.

Is there a standard tapering approach for A/B/C races? Is there scope to individualise this? Is this something the system can learn over time? How do travel plans figure into the algorithm for tapering?

Yes, there is. For A races, we generally use a 2-week taper, lowering volume but maintaining intensity. B and C races generally have a 1-week taper. There is scope to individualize everything on Athletica. While learning is on the roadmap, the system does not yet learn about the individual nuances of how athletes program or respond to their taper (we are a small team and have larger priorities to build a more robust system for athletes and coaches first).

There is some consideration for travel within the taper as there are usually fewer sessions in the last taper week. However, this is often where athletes and coaches will appropriately delete a session or two to account for the travel aspect.

Of course, if or when sessions are deleted/added in the taper, our loading algorithm (see above) continuously refines daily load to optimize both fitness and freshness until race day.

How is the load of weight training accounted for? I am in my fifties and have concluded that a decent amount of weight training is essential. I have also realised that it’s tricky to programme it in a triathlon plan without something else having to give!

You are right to be including weight training into your program as an athlete in your fifties (welcome to the club!) for many reasons.

From a model standpoint at Athletica, this one is a struggle. If I’m honest, I’ve battled it since the days in the NZ Olympic program when we started using TP across various sports around 2010 that have both metabolic conditioning and S&C in their programs. I’m sure there’s a hundred threads on ST talking about it but the general point we’ve come around to is that there is a collision between two models of load. We have the TRIMP (HR/lactate) model above that everyones familiar with (green/red figure below), colliding with a neuromuscular model (black). This is classic HIIT Science material we teach for coaches across all sports.

https://external-content.duckduckgo.com/iu/?u=https%3A%2F%2Fpbs.twimg.com%2Fmedia%2FDnjlu5MVYAAS-2x.jpg&f=1&nofb=1&ipt=3a954ba3d2eab096194a81afe5acda8edf536644b53b0c9f0ab6fb21a327b04a&ipo=images

So the TRIMP model handles the two variables on the left side (aerobic/anaerobic) but the neuromuscular model on the right is absent. To get you thinking about the neuromuscular load/stress think about the soreness you feel after strength training or speed work or running uphill/downhill hard over short periods as examples. TRIMP (i.e., TSS, two on the left) doesn’t touch it.

So to your question, this is still a work in progress for us. But work is being done right now in fact as we look to bring more variety of S&C sessions into Athletica with a leading colleague of mine. We have basic S&C sessions on board already you’ll find across most of the programs but those will be replaced by more advanced sessions in the future (big job, +6 months). For now, we recommend you eyeball the existing sessions as a starting point but make your S&C work individual to your own needs. Record the work done on your smart watch and record your RPE/feel and comments afterwards where we’ll take some form of load marker for the work. As there is massive context involved here when you consider every individual and their needs, this makes for a fun puzzle to try and tackle. And now with LLMs and new sensors, we have a real chance to make a dent.

Does the system give any consideration to whether a single sport block might be the best programming solution for an athlete?

Last one :slight_smile:

The system does not yet do that. The algorithms described above work to maximize individual sport specific fitness and also global fitness (i.e., all combined). So at this point we are not adding more cycling sessions over running and swimming sessions. Simply the load of each session is working to deliver the best theoretical stress to raise your fitness yet not violate common principles around ramp rates and acute to chronic load ratios. While block training might at times be effective, it is also risky from a loading standpoint (illness/injury) and we take a conservative approach instead. Of course, the individual user, athlete and coach, are certainly able to override the system and engage in this practice if they feel its best.

While we’re on single sports, you’ll note that Athletica services the single sports of running, cycling, rowing (beta) and one athlete has in fact used it successfully for multiple swimming events (using user-time constraints feature also in beta). As a multisport athlete, when you decide to do a single sport event, say as a B race, Athletica reverts to programming specifically for it about 3 weeks out, before going back to the (often) multisport A race. Best practice in this case we are finding is to jump on board our user-time constraints beta program that allows you to specify the sport, time and days you want to train. This provides a more balanced training plan for multisport users focussing on a single sport (i.e., marathon, Fondo, etc). Just a few bugs we are working out before we bring that feature to full rollout.

Thanks, Chowders. We appreciate the thoughtful questions you’ve posed. There’s a fair few questions in there so instead of making a long winded reply that becomes hard for anyone else reading to digest, we’ll address your questions point-by-point over the coming days so please be patient as we consider each one before rolling them out. Please note that responses are made in collaboration with our backend modeller and data scientist, Dr. Andrea Zignoli.

Paul - thanks again for considering my questions. I’m not short of other questions / things I am curious about, but I recognise that at some point I would just need to make a start and see how it goes in practice.

It seems like you are continuing to develop the system. Is it fair to assume that will continue for the foreseeable future?

Thanks again,
Paul

It seems like you are continuing to develop the system. Is it fair to assume that will continue for the foreseeable future?

Thanks Paul. Yes, we iterate quickly and our team has no plans of stopping anytime soon…

Paul -

Maybe I’m in the minority worrying about something so stupid, but any plans to change the interface? Or having a dedicated app?

I, like many, are used to trainingpeaks. Granted, TP isn’t creating anything magical by any means, but the current Athletica AI interface is a few steps down from some other competitors.

Overall, I’m enjoying it so far!

Glad you’re enjoying the training Cloy and thanks for the feedback. Those two issues you raise have been coming in loud and clear.

  1. Dedicated app: we have recently signed with a specialist company in Vietnam to deliver a contemporary app to our users. The founders of the development company are both triathletes that use Athletica. Just starting so at least 4 months to launch realistically.
  2. New interface: we are close to revealing our new interface. Levi (former Today’s Plan) has done an incredible job for us. Any coaches reading, Levi has a question:

Working on an updated version of the calendar view and have a question for coaches. Right now when changing between athletes the calendar jumps back to the current week. When compared to other platforms their calendars generally don’t change viewing dates when athlete selection is changed.

So for coaches out there, which feels better to your workflow.
A. Have the calendar always return to the current week
B. Keep calendar on the viewing week when the athlete selected

Would value any coaches’ two cents on their workflow UX preference with their athletes, responding here or over at Athletica to Levi’s thread.

Latest release: Power and pace profile charts upgraded to version 5. Includes analyses of strengths and work-ons below each chart.

https://canada1.discourse-cdn.com/standard30/uploads/athletica/optimized/1X/6a3b7c263b670993765de53cd1df86b8be64aeb4_2_690x354.png

https://canada1.discourse-cdn.com/standard30/uploads/athletica/optimized/1X/e46393dc6fb0a805dc1a2dbc2eecba33707e34be_2_690x377.png

I think I read it on the site that you are working on having it work with Stryd. Any timeframe on that? Thanks!

In our latest update, running power, if you collect this, is now displayed on your session analysis chart.

Other releases include added support for displaying cadence, RPM, and stroke rate data (rowing) in session details and session analysis charts.

Sweet! Thanks. Might have to check it out again and see how it works now that running power can be added.

Is there a reason you can’t pick a weekday for a race? I have a tri that will be on a Friday (due to state park regs can’t be on a weekend) and it’s not allowing that selection.

Trying it out obviously…Will share other feedback once set up.

Is there a reason you can’t pick a weekday for a race? I have a tri that will be on a Friday (due to state park regs can’t be on a weekend) and it’s not allowing that selection.

Trying it out obviously…Will share other feedback once set up.

Just acknowledging this limitation and thanks for flagging. While not ideal, at the moment for the weekday race situation (rare), I am happy to customize your taper week accordingly. Just get in touch via the Athletica Forum or otherwise. This is on our development list and obviously key for the long term success of the product, just some larger lifts ahead of this one at the moment.

Awesome, thanks. I’ll check out the forums and let you know but I can probably figure out my own taper tweaks. And I’m sure weekdays aren’t common but enough to make the eventual improvement list perhaps.

FWIW, I’m only a couple days in and already like it a lot more than TriDot (just finished their free trial), and candidly, that experiment was just due to limited customization when injured and really crappy/low-quality swim workouts in the standard 80/20 plans.

1 Like

Voice Messaging is now available in the Chat tab. This enhancement allows coaches and athletes to communicate more effectively by adding the power of voice to your sessions.

To learn more about the feature and the thought process behind it, check out the latest Coaching Prof newsletter here.

Was very open to giving athletica a shot, but two things just stopped me in my tracks.
#1. I don’t use garmin or strava, I use Hamerhead and Coros (I looked past this opened a strava hated that too and not deleted it)
#2. How do I mark sessions completed when I don’t have a data file?
It is simply ridiculous to think people will wear devices for everything.

Very flawed app so far, hope things improve, but my two week trial was essentially a waste

Well if you can’t give the AI any data what exactly are you looking for it to do?

5 Likes

I tried Athletica AI for two weeks. I don’t have 20 years of racing history, and since cycling is 60% of the race, I need to learn to be as efficient as possible and be a strong cyclist simultaneously.
AAI BIKE TRAINING: I didn’t like having to sign into a separate bike platform membership, and prefer to “train without game” without the social gamification of virtual worlds.
Most notably, I have grown to realize that there is so much more to indoor cycling than following a bike training session in ERG mode riding along just matching power. I greatly missed the Purple Patch Squad bike sessions. Why? The coach’s involvement during those sessions motivated me throughout the entire training. It helped keep me at the optimum pedal stroke, as did staying on top of the power and cadence gauges. They provided in-depth feedback on what success should feel like for that particular section of the live training session. While you could perform the session anytime as a video-on-demand, if you participated live, get ready: the coach will call you out by name to critique your output, your position on the bike, and your target performance in real-time.

What is missing from AAI training? Sessions that focus on the relationship of cadence and gearing, rather than just chasing power output. AAI needs to incorporate when and how to utilize out-of-the-saddle riding, position on the bike, flat and hill gearing choices, managing momentum through the terrain, pacing strategy, RPE and power production, energy conservation, pedal stroke efficiency, mental resiliency, and incorporate all this into terrain management. I very much missed this approach at Purple Patch. It made my training more effective and helped me understand the importance of these factors in improving my cycling performance. I went back to Purple Patch Training because the training isn’t just about making me physically stronger; it’s about transforming me into a smarter, more efficient, and faster rider.

Thats exactly my problem.
I dont swim with a watch, never have. I record all my splits afterward, just surprises me there is no option for this or to simoly mark ws completed and have the algorithm assume it was completed as planned unless otherwis3 specified

I’m sure you’re not the only one in this boat. Not sure if @plaursen has made it over yet but he’d be the one to address this.