A bit of a different episode today going over some of the recent analyses I’ve been running over the last few weeks. There’s essentially endless questions we can ask of our database, so if anybody has any questions / curiosities they’d like for me to explore, let me know!
// TOPICS COVERED
00:00 Introduction to Data Analysis in Cycling Training
01:22 Evaluating TrainerRoad AI Performance
11:03 The 50% Rule in Training
22:57 Managing Big Rides Effectively
23:23 Understanding Athletic Age and Performance Metrics
29:55 The Impact of Structured Training on Performance
32:46 Analyzing Training Consistency and Performance
42:14 Comparing Training Methods: On-Offs vs. Intervals
50:47 Conclusion and Future Insights
Great episode and I loved the webapp design to convey the story @Jonathan
Question for a future analysis: What is the ‘realistic bounceback rate’ from a spring/summer of unstructured riding when one might go back to structured training in the fall/winter? You’ve mentioned before how TR is a fairly seasonal product, and use is lowest in the nicer months as people go outside.
I’m sure there are fitness gains from general volume while riding, but nothing compared to structured training. I’d be interested in reviewing the FTP or actual power achievements of two cohorts.
Those who maintain structure year round but still go out and smash or do fun rides in the spring/summer.
Those who truly abandon any form of structure for half the year.
Perhaps finding individuals in similar initial FTP buckets and volume over the course of a few years?
How does the progression of those two cohorts relate or differ? I have my assumption, though i’d love to see what the data shows.
He addresses a couple times in the pod. At one point, in regards to the 50% rule he states: “…it’s observational analysis that draws correlations as opposed to causation” but provides more of an explanation of this later in the pod.
Some durability stuff. Yeah, I know it’s an overused term, but it’s a thing. And I don’t think many of the tools have good metrics on it. I’d love to see metrics on what power PR’s look like after someone has pushed 150tss, 200, 300, etc. through their legs. 1’ - 60’ power PR’s are interesting. 1’ - 60’ power after pushing 300tss through your legs is often a better predictor of race performance in long races.
I like the 50% rule and I’m actually surprised things don’t go more sideways even at lower percentages. I’ve love to see some similar analysis on when athletes push volume within a training block. This time of year, many of the pros and high level amateurs are talking about their “unbound block” where they push a bunch of volume in a final block before unbound. There are lots of flavors of this with some keeping a regular schedule with longer rides, some are skipping rest days and doing more rides, etc., but the common theme is a significant bump in volume. Would it be possible to look at the outcome of an entire training block rather than a single workout or week? When I do a big volume block before Unbound, I’ve added a 4th week to my block (basically pushing my rest week). I think it’s worked well the last couple years, but honestly not sure if there is a best practice here (or if a volume push even makes sense). If a volume push does make sense, is there a good rule to set limits (25%, 40%, etc.)?
Major props on the evolution of the trainer road product! I happen to be one that the ai doesn’t seem to work well for so i was wondering if my crosstraining is a big reason for that. In the past few months, a very small percentage of the suggested rides on my scheduled hard days were at the appropriate level. The last week or so the suggestions have been making more sense though.
It is nice to see some validation of a basic rule that had been instilled in my training for running also applies for cycling. For running, barryp from slowtwitch had been instructing new athletes to make sure their long run makes up less than 50% of weekly volume, ideally consisting of about 30-35% to ensure consistency of training.
Did you all consider outside runs as part of the athlete training volume for those queries? Since you had a minimum volume for a few.
Based on that fascinating podcast, it would be great if the calendar/plan could show alerts if you skip/adjust/add/remove (or plan to do so). E.g. “your lowest TSS weeks are trending down… that’ll impact your progress” or “that ride’s TSS was very high … you can expect a drop in FTP”.
I’ve kind of been thinking that same thing. I’m also wondering if that’s why my recovery weeks didn’t really seem like recovery weeks from a TSS perspective when people did a deep dive on my calendar.
It definitely seems like TR prefers the 70-75 FTP for endurance vs 50-65 range. That said my long endurance rides(2 hours plus) would be more in the 55-60 range.
It is really interesting seeing all this data and does make me feel better about my training. I think a lot of times we get stuck at the pointy end of the spear and miss the forrest for the trees.
Really big kudos on this one @Jonathan ! A lot of us have clamored for this a long time. It’s such a great credibility and confidence builder when you can point to the mountains of data.
+1. I am curious if the “50% rule” (observation) applies across all volumes or is skewed more toward the median. I think it was mentioned it was for athletes that rode 4.3 hours per week or greater? You would think that someone who averages 300 TSS can do a 150 TSS long endurance ride and manage ok. But for someone who averages 1000 TSS and does a 500 TSS ride that training disruption would be greater. But maybe not?
One thing I would like to hear more about, is workout compliance, especially in indoor vs outdoor training.
For instance, do you recommend to always do (shorter?) endurance workouts indoors, to avoid micro breaks? And what about intervals?
Also, how much/little deviation does it take to significantly reduce the quality of the workout? E.g. turning sharp corners, or not hitting the power target close enough. I would guess that it also depends on the type of workout.
I have picked up some info on this from previous episodes, but it would be nice with a more thorough walk through on do’s and don’ts.
My guess is the model is trying to maintain some intensity while reducing volume for the de-load weeks. Riding high Z2 for 45 minutes doesn’t feel like “endurance” but theoretically you should bounce back pretty quickly due to the low volume.
I definitely enjoyed this episode and I’m glad I took a rare opportunity to watch it on YouTube rather than listen on a dog walk.
My point about the correlation , which Jonathon did repeat several times, is that when he was talking about the 50% of TSS causing a skipped workout.
Well, couldn’t some of those 50%+ scenarios include people completing their A race? I’m just thinking that in my situation, my A race greatly exceeded my average TSS, let alone 50%, and I definitely took some time off afterwards for recovery and also to repay some family time. Maybe those skipped workouts were taken into consideration and I missed the qualification?
I’m sure there’d be a correlation between riders completing their A race and skipping workouts, and that might also explain some people like myself having weeks when I didn’t ride at all. So having a very low low.
https://www.instagram.com/TrainerRoad