A "big data" question for TR staff

I think those are good questions!

Just want to mention…according to my limited review of TR data, the major thing that improved workout compliance was answering ‘yes’ to the following question: has the athlete been following TR AI?

Pre-AI compliance was not as good as Post-AI compliance. Or so sez I. Any other effect is likely tertiary.

1 Like



sez TR.

What about those of us that never failed to fail, but got faster? :wink:


Actually, the really interesting question is: what is the optimal workout failure / inability to complete a workout %?

Hear me out before reacting. My logic for an optimal failure % is the following: if you never fail a workout, that means that all of your workouts are within your capabilities, so you’ve never tried to find the edges of your capabilities. So you could do more, you just don’t know how much more. It’s only from trying and “failing” that we truly discover what we can / cannot do


In my experience I f you’re getting stronger week to week there’s no need to find your “point of failure”.

A persons limit is fluid; as their fitness grows their failure point moves. Therefore little can be gained from pushing to failure “one time”. And if it’s required regularly physical/mental burnout is just around the corner.

  • Exception being race day.

Gratuitous thread bump


this is where you I believe you need to @ Zackery Weimer or someone else.

I’m also interested in knowing the answer to this. Until then, I can tell you that yeah, personally I found it much easier to do threashold workouts and others using Resistance mode instead of ERG mode. I also believe most people find it better (if they ever tried it) and/or just doing it outside. just my 2 cents for whatever’s worth.

Hey there! This isn’t something we actively track.

Other variables are more impactful when it comes to passing or failing a workout (what workout was recommended/performed, sleep quality, caffeine intake, nutrition, stress, etc.), which would make it exceedingly difficult for us to extract any actionable insights from this sort of data.

Our general advice is that athletes experiment with different trainer modes to find which one they personally prefer. ERG mode is usually a good bet for most interval sets, but it can be useful to hop into Resistance mode for VO2 or anaerobic intervals that require quick changes in power – especially with trainers that are a bit laggy when it comes to interval/power changes.


Yeah, but you don’t track any of those in TrainerRoad… so why not at least run the metrics based on what you DO have?

1 Like

Because requests like this are a ton of work to satisfy curious internetters while distracting them from their day job?

(Coming from someone who works with analytics and reporting for a living)


Yeah, got me.


Why not have TR switch to resistance mode when there is a short upcoming interval and switch back to erg mode for the rest of the workout. Could even use a different color or just a darker blue so its obvious what is coming in the graphics on screen. Erg mode is still useful for the easy parts.

1 Like

Good search & grab. I had those in this reply and removed them since you added them in your edit.

If anything like this comes along, it needs to be a user-selected option, because some people stick with ERG even for those efforts. Needs a basic on/off control along with some form of threshold where it does or does not activate for those with it turned ON. That threshold may need to be user selected too, for people to set above some % of FTP that they want it to work vs not.

I realized I should have done the research right after posting. I can see turning this feature on and off easy (change the dropdown where you select resistance and erg mode but leave the keyboard shortcut to only switch between erg and resistance) but advanced settings shouldn’t need to be that easy to do. I don’t think it should purely be above a set wattage, but more the jump is above a certain amount and the length of time you’re above that amount is under a set amount.

Thinking at least initially this should be very simple to take care of most cases with as little work as possible. (not claiming this is easy) So no configurability and only apply to vo2 type intervals and shorter so if your power increases by >= 150% from where you are and that power drops by>=25% in under 1.5 minutes.
So if you’re going at 100 watts the power would need to be 250 or more and then drop by 62.5 watts in under 1.5 minutes.
For example Logan:

40% of ftp up to 140% and then 95%. 40% to 140% is a 250% increase, 140 to 95 > 25% decrease…

Would be interesting to see the differences between trainers and might allow trainers to better know how to improve. Plus can give better data to separate wheel on from wheel off trainers. But this might greatly impacted by the money people are willing to spend. If they use a cheap trainer they may not train the same as someone willing to spend more.

1 Like

Personally, I think the best use of this would be to add it as capability into a much needed redo of Workout Creator. That is, besides being able to set the length of an interval, the %FTP, I would really like to be able to select ERG or Resistance, and if Resistance the %.

For non-custom workouts, I think having TR have to create a “simple” level at which workouts would switch from ERG to Resistance won’t be “optimal” for the majority of people, and will just lead to requests for more control. v1 of this should be (in my grand humble opinion) give the flexibility for this for custom workouts, learn if people actually use this and how, and then figure out if it makes sense to add an option to automatically switch modes for TR workouts.

How many people actually create or follow custom workouts? How many of those workouts are generic workouts (i.e. not aimed at someone doing something very specific) how many of those would people spend the extra time to set the mode of the trainer? I’m not saying get rid of custom workouts but I don’t think the user base is large enough to show what you are asking.

What I’m suggesting is a new auto switching mode. If you don’t like it you don’t have to use it. Workouts don’t need to be adjusted at all (no extra data to save in a workout file) so would only impact playback mode and if you want to do an easy test only need it on one platform (no need to adjust the whole platform). Default to not use it but tell people here about it and see who turns it in and leaves it on.

Also you can look at what the competition is doing too. Say looking at how lots of xert users like auto mode there where there is no ability to manually set it

I use custom workouts to tweak standard TR ones (but much less so, now that I’m using Adaptive Training more effectively).

An example of my ‘custom’ workout would be a tweak of Creise. In the last set, I’d increase the rest durations from 0:30 to 0:45. That simple modification gets my old, exhausted body, a fighting chance to finish the workout and hit the power targets.

All this is interesting discussion about custom workouts — but my core question was really about ERG vs. Resistance mode and whether it correlates to ‘successful completion’.

1 Like

We can have an evidenced based argument if you want, and I can use your same argument against what you are proposing: how many people will use this? How many actually care?

On the substance of your argument, I disagree that this is / would be easy for TR to implement on top of existing workouts today. In your system, when TR automatically switches out of ERG mode, what mode and at what setting does it switch into? Resistance Mode or Standard Mode if a trainer supports both? At what level? Should the user be able to select both the mode and the level? Etc.