in particular if you are only doing TR indoor workouts
3 Likes
How about simple method like… if someone failed(or paused a lot) 2 or more consecutive threshold or ou workout, TR automatically detect “possible overestimate of ftp via ramp test”. Then notify one to adjust his ftp or edit that 0.75 multiplier…
Yes, this is my story. once I used ramp test result as it is, I barely finished SSB LV when I was newb to structured training. Never bailed during workout but have to pause in order to finish the threshold or ou workout and the trend kept on. I eventually hit burnout and stopped training a while. Now I took 0.95 to ramp test result(roughly 0.72 multiplier for ramp test) and happy with current progress on my SSB MV.
I wish TR has some way to adjust ones multiplier based on ones workout history or let person to choose the value. for example, I might set my custom multiplier as 0.72. When I loose pedal in ramptest, TR prompt me two possible ftp estimate. one with 0.75 and the other with custom value of 0.72. Then I accept one of them. I know I can manually adjust my ftp value but this seem provide a little bit of seamless user experience(I think). After all, one’s ftp could be 0.7 ~ 0.8 of MAP.
If ftp is overestimated, it will affect subscribers loyalty as “I’m not fit to TR” and bail. In reality it’s simple as adjust down ftp value and you can see your progression. But that knowledge(adjust ftp!) require some time in structured training. Feature of adjusting multiplier would be good for both UX and loyalty of newbs.
2 Likes
this thread looks very interestings, if there will be a automatic, i think it would be helpful to syncronise your body weight too from external sources like Garmin Index scale, or?
1 Like
you mean like the 4DP Profile?
SUF acts as if they had invented it, in reality it is just a special type of marketing, but an implementation in this direction could be very interesting, because it helps you find out your weaknesses easier and of course TR could offer workouts for these weaknesses (by automatic or so), or?
1 Like
it could be a combination of 2 Tests (linked together), you have to ride on 2 different days or so, as soon as you have finished both tests, the final power profile can be calculated or so?
Before I forget, a power profile should also take into account the age, I have not seen it often, but in WKO5 or intervals.icu you can see it quite nicely age adjusted, so the driver type should also be able to be determined, such as:
All-rounder
sprinter
Time trialist
Climber
Steady-state rider
Pursuiter
I also like the evaluation in power-Meter.cc
1 Like
I don’t think weight is a particular factor in this idea. Power available on a given test day and the recent history power curve are what they are regardless of rider weight.
We do have an open. Feature Request for weight tracking and sync with smart scales. So that is already covered.
I would like to see the AI move in the direction of not needing tests all the time. There would need to be an initial test and possibly 1 or 2 throughout the year, but if TR is looking at our workout/ride performance it already knows what we’re capable of. I would like to see FTP adjust based on performance. Not performance of a single workout/ride but of numerous recent efforts (14-30 days).
Much like how TR has a pop up after a ramp test to “Accept New FTP”, this could be worked into the software to do this periodically. An example: “Based on recent TR Analysis we recommend a new FTP of XXX Accept/Reject”. TR could go into further explanation and give data points on how they derived this “new FTP”. Then the user could determine if it was accurate representation based in stress, sleep, fueling, etc.
6 Likes
I think they should do both of these, or something similar.
The goal would be to figure out the % changes in ramp test derived FTP to set the right targets for 1. Threshold/SS work and 2. VO2max work.
You’d do the ramp test to get a baseline, and then these two additional tests to tweak from there.
You’d only need to do the set of three tests periodically - maybe just once per year - As the longer you ride with TR, you get to know how you respond to different workouts, and you get an understanding of how to best adjust ramp test FTP to inform targets for different workouts.
And you could do the ramp test more frequently by itself, and use the adjustment factors from the initial set of three tests.
The additional two tests could in fact be incorporated as actual workouts early in each plan. That way they would feel like workouts and not tests.
This is imminently doable - it’s just a question of priorities, and having the right combination of data analytics and understanding of physiology to design the best approach.
I take the ramp test FTP and reduce by 5% for example. And TBH, I don’t do the ramp test very often, because my FTP does not change much up or down, and I’ve gotten used to adjusting workouts targets manually if things are too hard or easy.
interesting for a Power Profile over a longer period, for example a avg of the last 42 days?
1 Like
I hear you on that! My simple mind can’t get around how to approach this, though. For example, I like to look at efficiency factor on some types of intervals and I think it could be a good indicator of progress, but a lot of people don’t use HR. And absent other physiological stuff I can’t perceive how completing intervals would necessarily indicate growth, especially for people who don’t have a ton of outside data where they might set some PR’s for wattage over time. Having something that accounts for incomplete data/limitations is hard from my point of view, but I’m sure smarter people than I are working on stuff like this
Most of these “problems” would disappear if people weened themselves off of erg mode and got comfortable doing interval workouts with range/zone watt targets instead of being locked into hitting a single number thrown at them by a computer.
You don’t need artificial intelligence if you just develop your own 
6 Likes
That makes some sense considering the association with CTL/ATL/TSB calculations (based on my limited knowledge). But it may well be something different in terms of evaluating a Ramp test and the “recent” power curve.
With all the data TR has, I suspect they can investigate starting with Ramp tests and the assigned FTP. Then review the workout compliance in the following workouts over the following week. They can also see if people make manual adjustments of the FTP after the fact (as well as the option to NOT accept the Ramp test suggested FTP).
See those results and then look back at the preceding time frame, and see if there is some correlation between the FTP and/or plan compliance, in light of the power profile present over different ranges of time (maybe starting with 2 weeks up to 8 weeks?).
I have no idea if there is a practical correlation to draw, but there is a potential for a TON of data in their hands, that may help them determine when the 0.75 multiplier is and is not appropriate for a given rider. It’s possible there is a correlation for riders that will always apply (based on their heredity, training history and such) or it could vary for riders as they move through training phases (based on their prior phase).
It’s all speculation, but TR knows that the 0.75 is hitting some amount of the middle in the bell curve. My goal here is to better leverage known data (if/when it exists, since “new” TR users may not have a power profile to judge against initially) to the end of getting more useful FTP and training zones, with less guessing and interpretation on the rider’s part.
Another, possibly different angle alluded to earlier, is looking at the rider workout adherence following the test. If they implement a “compliance” function, it’s possible they can review workouts (may a minimum of 2 or 3) and then consider a suggestion of raising or lowering rider FTP by seeing problems in those workout.
This is all far from easy and I do NOT have the answers. I just think there is a real need and worthwhile opportunity to improve the testing evaluation, with the aim to improve training for all riders, and minimize the guesswork as much as practical, within the price and service provided.
1 Like
That is an interesting angle, and one I don’t think matters. If you are following the targets in Resistance or Standard mode, it’s not entirely different than getting similar control via ERG mode.
Falling short of power targets is possible to see regardless of the trainer mode. It may present itself differently, but it’s still here.
- ERG may start with lagging cadence and fighting the increasing resistance, and resulting in the Spiral of Death to the Clinch.
- RES/STD mode may show as a rider dropping cadence and power through intervals and ending with results below the target.
In each case, it’s possible to see there is a issue with the rider and the power demands. Manual control over RES/STD doesn’t magically fix the issue, and ERG doesn’t cause the issue. The core problem is the rider and their inability to hit the target.
The root cause may be a crappy day (bad sleep, insufficient recovery or fueling, etc.) or a new FTP that is out of reach in a practical sense. I just don’t see trainer mode as a relevant issue to evaluating FTP in the way I am asking here.
2 Likes
For weak souls like me who (wrongly) feel any move below 100% on the intensity adjustment is a sign of failure, erg mode is a destroyer of dreams . . . Mentally, erg mode often leaves me with 2 bad choices, "bail’ by reducing intensity or soldier on and die a slow painful death. My rational side knows this is wrong but that side doesn’t always win
I have found great relief going back to resistance mode and doing my intervals hard and not sweating the watt variations. Bad days don’t involve telling the computer i suck (i.e. manually reducing intensity) and on those good days, I get the joy of bumping up against the top of the target zone.
2 Likes
Yeah, our head can do odd things. I am fine with the ERG Workout Intensity adjustment. It is a different form and feel to RES/STD and just letting the watts fall a bit. Ultimately, we know the impact may be the same, but the trip there sure can lead to a different feeling (mentally) afterwards.
2 Likes
The computer gives you a single number target for any interval and your FTP for that day because of constraints flowing from the equipment and software. Neither single digit precision is based on physiology. Its a software thing.
Erg mode can only handle a single number, the FTP number is based on snap shot in time effort run through a model designed to fit a range of humanity and your number by necessity remains static over time. It all works but it is not precise and was never intended to be. The folks who develop all this stuff understand that. That’s where we get zones from. The only “issue” is many users demand a level of precision that is both not possible and that was never intended.
5 Likes
Oh, for sure it’s not necessary to be successful. Some level of precision is needed, but nothing close to the single watt worry we sometimes see (and that I fretted at times in the past).
“Close counts” in many ways, and it may seem I am asking for a tighter level than I really am, with the main request here. But this aims more towards getting the overall rainbow of power levels a tad closer, so we are more properly “centered” in the fader range of our training levels.
2 Likes
This exists already with the 8-min test, a 20-min test, and the ramp test.
I actually prefer the 20-min test when I really want to test my capabilities for a TT as it just feels more accurate and I can get a gauge on how X amount of watts feels for 20 minutes, could I hold that for another 40 in the race and how to adjust that or train it leading up to the race.
What Nate has said is that practically no one was completing the 20-min test within the training plans, thus not having an accurate FTP and subsequently making the training plans either too easy or too hard…this all leads to an ineffective product and people won’t keep paying for that, even though it’s the customer’s fault for not following the plan (even if it’s just the test).
Also of note is that I need an easy day (or even off day) immediately after the 20-min test, but with the ramp test I can do an hour workout that same day and a hard one the next. They have stated this much on the podcast that you can just put it in anywhere if you want to re-access and not really effect your recovery/plan.
And of course they want more customers, so making it easier to get started is of much importance, as well as keeping people engaged in the plans so they stick with it.
This is why I love workout creator so much because to me FTP is truly best used for steady state work so SS and lower, but once you push upwards it changes, so if you just pluck a VO2 workout at 120% based on FTP it’s really meaningless as far as if it’s applicable to you or not. Some people find they can do much more than that, some find it impossible so can create my own to adjust for that aspect.
Nothing is perfect, but TR is doing it right, they just can’t account for every person with one application so we must make our own educated decisions. Even Nate himself is famously known for trying all sorts of crazy things with pseudo-science and learning about it so we can follow that lead and learn more about ourselves which helps improve this platform at the same time.
2 Likes
I think this is a really important skill to have. It not only reduces your reliance on a standardized test (which I think most would agree is a good thing) but also enables you to adjust your expectations day by day, which has positive implications for workout compliance.
The amount of ride data that TR has is pretty revolutionary and can certainly be leveraged to improve testing outcomes, but at this point I feel that self-evaluation is still superior given the amount of external variables at play.
6 Likes
Maybe it would not be bad to think outside the box, when running with Stryd Pod there is an Auto CP which more or less sets your rFTP pretty well without you having to do a special test.
The few things you need, you have to complete so many different runs over a period of 1 to 2 months (more or less a 42 days window), based on this data your rFTP is then determined quite well.
I have been using the Stryd Pod for a year and have converted my running workouts from the TR Triathlon Plan via Training Peaks so that I can send the workouts to my Garmin watch.
I have to say, impressive, my limits are right in terms of feeling