This is 100% correct. What I would do is create an index based off a fixed FTP and fixed workloads to use as a baseline/comparison for the index for each zone.
Suppose for men the fixed index FTP is 200. Now suppose the fixed workload for vo2 max is 5x5 at 125%. If I complete a 5x5 or equivalent work out according to current PLs, it would automatically set my vo2 index based on the power I completed it at.
As an example, if I do 5x5 at 250 (125% of 200), my vo2 index would be 100. Alternatively, suppose my FTP is set at 300 in TR but the best I can do is 5x5 (or equivalent) at 120%. This is equivalent to doing 5x5 at 125% for someone with an FTP of 288, therefore my vo2 index would be 144.
This is actually how I measure my personal improvements and for the most part stopped measuring FTP. In 2021, my race pace for a 6-7hr gravel race was 200 watts. In 2022, it was 220 for the same race. I don’t know what my actual FTP was for either race, but my “endurance index” went from 133 to 146. Effectively, my endurance paces were equivalent to someone with an FTP of 267 and 293. It might be that my actual FTP was 250 and 310, but I didn’t care about that value.
I’m always curious by this (PLs or indexes), as generally speaking simple inspection of power curves can be quite revealing. For something like vo2 you look at increases/decreases in average interval power of full-gas efforts. Or if I’m more in maintenance, repeatable efforts. For me thats looking at average interval power on 3-min and 8-min intervals.
One of my best tools for that are a series of tables/charts showing best power at a LOT of durations (5-sec, 10-sec, 15-sec, 30-sec, 60-sec, 2-min, 3-min, 4-min, …). TR shows something like this, right? TP has peak performances. In WKO I can pick any duration that I’m working on and see progress in developing power. Intervals has an intervals search function.
The basics are pretty straightforward. Some of the more interesting metrics require some math. With a good analytic tool you can quickly view and compare across any timeframes. We’ve had good human performance models for many many years now.
That’s a bit like Join Level. Would be an interesting metric to implement on a zone by zone basis.
It’d be a terrific way of tracking and comparing fitness actually.
The scale would be the tricky thing, but definitely doable with TR’s database.
Thought experiment - Threshold Level 100 would be TDF pro or something (say 6w/kg for 60min). 90 would be a Keegan type (say 6w/kg for 4 x 15, or 5.5w/kg for 60min). 50 would be a handy amateur knocking out 4w/kg 4x10min with.
@Nate_Pearson I think this is a real thing that would add value (but done properly and established with real data not with my made up stuff)
I don’t think this would make any sense or work well in practice. PLs are tightly linked to workout levels. A PL 7.5 sweet spot workout is the same independently of our respective FTPs. That’s because, well, the workout is designed with relative power levels, not absolute power levels. The latter doesn’t make much sense.
And because workouts are designed with relative power levels, PLs need to be relative, too.
Now you could argue whether we couldn’t compute an absolute PL score from a relative one and your FTP, but also here, I am not convinced this is actually useful: PLs measure a different dimension of your fitness, and capture e. g. repeatability and/or endurance at a certain power level. Even if you computed an absolute number, is it really useful and informative? If my hypothetical absolute threshold PL is higher than yours, but your FTP is 20 W higher, who is fitter?
We should be very careful with metrics. IMHO the best type of metric TR should introduce next is one or two consistency metrics.
Why not?
PLs are a relative measure of your fitness given a specific FTP.
Why does it need an absolute component?
Like I wrote above, all workouts are designed relative to your FTP, so you will get a relative measure of the difficulty. The absolute component is your FTP.
@Nate_Pearson at least on one occasion wrote that internally they have an absolute fitness score. And that they have been thinking about whether it makes sense to create absolute measures of your fitness other than FTP. So far they haven’t and IMHO for good reason.
I don’t think it’d make TR a better product, quite the contrary.
I don’t think it has to be either or, I think FTP and PL/WL would remain, the AL (Absolute Level) per zone would be an additional number at the end that would give you a direct comparison between those FTP’s that are 20W apart.
Simple and hard: you’d have to find a metric that is specific to that effort. But you’d have to find the right metric yourself and monitor it using a tool like WKO.
I’ve read many discussions on performance metrics on the forum. When you surface metrics, their purpose is to allow athletes to judge success. So you have to choose carefully. I don’t think there are many generic metrics. Yes, you can use your FTP (absolute and relative to your weight) as one metric. Workout consistency is another.
But apart from that it really depends on what you are aiming for. Some people use TTE at specific power levels. Is that a good metric? Depends. If you aim for a 10k TT, is extending your TTE from 50 to 70 minutes at a specific power level really useful? Probably not. To address your example: you need to find a measure for “improved endurance” and then judge success based on that. If your FTP falls, so be it — provided you improve in the metrics most relevant to you. However, “endurance” might mean something much more complicated for you, e. g. the ability to ride for, say, 2,000 kJ worth of distance, and then put the hammer down despite your fatigue.
Lastly, the power curve is a tricky metric, because you first need to understand what you are looking for and be sure that the data is representative of your capabilities — you probably have not done all-out efforts for various durations. One of my local hills for hill repeats is about 4 minutes long. My 5-minute power will therefore not be a good metric, for example.
I agree with the diagnosis. But I don’t think absolute PLs will help here, IMHO they will hurt. Ultimately, the issue is that many people would like one simple metric to rule them all when that does not exist.
The only metric I think TR should add ASAP is workout consistency. Consistently following your training plan is the only universal metric I can think of that is nearly as useful as FTP. Apart from that, my best idea so far is for TR to suggest metrics based on the training plan you have chosen — the 40k TT plan should surface different metrics than the crit plan.
You write “realistic representation of [my] fitness”, and that’s exactly it: there are so measures of fitness even if you stick within one energy system. The more specific your demand is, the better you can measure fitness. And as soon as you have more than one measure of fitness, the question becomes how you weigh those measures. What do you think many athletes will think if [additional measure of fitness] increases whereas their FTP stagnates or slightly decreases?
And it is not about errors bars, it is about what that metric represents and what factors it weighs to what extent. Whether it is crowd sourced or not plays no role.
IMHO the best options for TR are one of the following:
Add consistency metrics. These are the only metrics that I think are a universal positive. Staying consistent is what gets you faster.
Allow some third-party tools to plug into their workout and ride data. I am pretty sure TR doesn’t want to do this for the legitimate fear of competitors vacuuming up their data (one of TR’s crown jewels).
Implement additional metrics that are specific to the chosen training plan. This is tricky for all the reasons I gave, but IMHO absolutely essential.
Allow for arbitrary metrics in a TR Pro tier (think WKO-type — you are on your own, but you can do essentially anything).
That applies to any situation where you track more than one performance metric. Mathematically, that is because you are then dealing with a partially ordered set, i. e. there can be situations where e. g. (x,y) is neither less than or greater than (x’,y’) where (x,y) could be any two (or more) performance metrics.
There is also a very subtle point about PLs that is quite important: it is not a performance metric, but a measure of your progress. Once I understood that, it totally clicked with me:
The name is a clue: progress and performance both start with the same letter. Yet, TR chose to call them Progress Levels, not Performance Levels.
The strongest clue is what they were primarily designed to do: to allow AT’s algorithms and TR’s athletes to choose the correct workout.
Progress Levels are in arbitrary units, i. e. the difference between Sweet Spot PL 4.0 and Sweet Spot PL 8.0 is Yes, the latter is much, much harder, but e. g. it does not imply athletes that manage PL 8.0 sweet spot workouts can hang in twice as long or so.
Progress Levels cannot be tracked over time. This is completely deliberate, because I think TR does not think of Progress Levels are a proper performance metric that athletes should aim to maximize. (And it was in discussions on performance metrics and allowing athletes to track progress levels where things clicked for me.)
That’s why IMHO absolute PLs will not happen, because just like relative PLs, they wouldn’t be a good performance metric.
PS Just to be clear, I have repeatedly and vigorously argued that TR should add much better performance analysis tools. It is a real weakness, especially for more advanced athletes whose performance is not well-reflected in FTP changes. For beginners and to a lesser degree intermediate athletes, FTP is IMHO a good and probably the best metric: it is easy to understand and going from e. g. 2.7 W/kg to 3.5 W/kg will probably imply that you have improved every aspect of your fitness.
I don’t race but participate in group rides and gran fondos. I like getting faster and like many have had successes and failures. My most structured training is always from October till April and then I ride primarily unstructured for the rest of the year four times per week for a total of ten to twelve hours per week. (ideally I would at least use train now twice a week in the summer but grandkids and life gets in the way)
I love the comparison tool for comparing year to year and I have little charts/notes/stats scattered throughout my life to motivate me to do better. This year my training compliance has been pretty good (first year in last three that I have not been recovering from surgery) and I was thinking what would really motivate me and came up with I would love a prediction of what my ftp could be in April if I maintained a high compliance to TR training. I know this would have a lot of variables but maybe something simple that could generate a number if, for example, I rated every workout as a moderate effort (this has been my pattern for the last nine weeks). The would motivate the crap out of me as I love chasing a number and if I could ever get to see that my one big goal is possible I would be all in.
As with everything this has probably been suggested before but I wanted to make sure I set this thought free to the universe.
How motivated would be in April if you’d maintained high compliance and didn’t hit this predicted FTP? It’s near impossible to predict ftp improvements. What worked last year is not guaranteed to produce a similar improvement a year on etc.
This conversation is certainly thought provoking and so productive! You guys are awesome!
It would be cool to observe the subtle details in the gradual progress you make in different training zones over the course of the season, relative to the phase of training, and your current FTP. I will point the team toward this thread that has abundance of compelling ideas and perspectives. As it stands, this is not on our roadmap. I raised this idea to the team, so the concept is on our radar.
Thanks again guys- we appreciate the thoughtfulness of this discussion .
I think I’m understanding that you’re looking for a singular metric that combines both FTP and Progression Levels, so you can compare your fitness more easily YoY. While Progression Levels are only relative to your FTP at the time, you’d like better visibility of your comprehensive fitness changes, which I can see the value in.
The good news is: we have explored a tool for this internally!
We would love to build better PR chart visibility, and that development process is ongoing as we continue to expand and build upon what that tool should ultimately look like. We’re excited to work on this, and I’ll be sure to provide updates as they become available!
There are YoY PR charts available now to view power PRs, but I hear you that you’d like Progression Level and FTP YoY visibility too.
Just to add one thing: don’t be boxed in by our limited imagination. Personally, I am not sure whether PLs are a good performance metric. But on the other hand, you might find others. You have infinitely more expertise in this area than I do.
One type metric that I do think TR should add ASAP is consistency/adherence to a training plan. AFAIK being consistent is a universal good. And once you track it, you could make targeted recommendations. E. g. if someone does MV but is not very consistent, AT v4.0 could suggest to the athlete to drop volume to LV and maintain higher consistency.
I’d add in that it would be useful to have some recommendations for our max effort at different durations. Whether it’s TTs or hill climbs, or even thinking about take a 1k flyer, those power/durations are really helpful to have as a guide.
Yes, the latter is much, much harder, but e. g. it does not imply athletes that manage PL 8.0 sweet spot workouts can hang in twice as long or so.
.
.