KOM Informatics: Secrets Of The High Responders

By David Brown

KOM Informatics is a web-based application that gives serious cyclists world class tools that provide accurate and comprehensive views of their training workload and how that workload relates to peak performances. We've published a number of articles that focused on the correlation between various aspects of an athlete's workload and their performance in 16 different personal record "time boxes". Here is the first article in the series:KOM Informatics: Raising 20 Minute Power Over 4 Weeks: What Actually Worked.

Just focusing on this first article for now, the findings were that Aerobic intervals (Z4, Z5) dominated the successful interventions, accounting for 52% combined. Anaerobic workloads (Z6, Z6) accounted for a not insignificant 28% which probably reflects the popularity of HIIT modalities. While intervals are undoubtably an important aspect of training, and may well be the primary factor to account for these athletes' success, we've found 0 coaches that advocate doing them every day. There's a reason for this, as anyone who has successfully completed an interval focused intervention will attest: When done correctly, you'll be fatigued for a day or two afterwards, making further interval training problematic. The focus then becomes optimizing the rest of the program.

All of the KOMInformatics athletes that managed successful interventions by achieving a statistically significant correlation between workload and performance did a good job. But some did better than others, judging by the number of watts they improved from the fifth highest to the highest PR in a given time box (WattsDiff). We're calling these athletes high responders. There are potential explanations for this apparent discrepancy that fall outside the data domain of the system. For example, the system has no idea whether an athlete is towards the beginning of their career and thus is more likely to achieve big gains. Since the system can't measure things like this, they won't be addressed in this article. Instead we'll be looking at some commonly dispensed training advice which can be measured, and seeing how much it contributes to WattsDiff. Hopefully this way we'll be able to winkle out the secrets of the high responders.

To achieve this, we'll be performing a series of correlation analyses similar to those we did for the original article. Rather then focusing on one athlete at a time, we'll be looking at everyone. The larger sample size should help to reveal trends that might be hard to spot otherwise. One end of the correlation analysis will be focusing on WattsDiff, that is the difference between the 5th highest and highest PR in a given time box. Here is a look at my data from 2022 and 2023 which should clarify things:

Ride Lots!

Overall KJs vs WattsDiff

This quote is often attributed to Eddy Merckx, and it's certainly seeped into cycling culture. But does riding lots have a statistically significant effect on WattsDiff? We found the correlation between KJs of overall riding effort and WattsDiff was relatively weak (r(274) = 0.1832, p < .05 (statistically significant)) but still statistically significant due to the large sample size. So, if you can't bring yourself to do intervals, grinding out large volumes of hours in the saddle is a slow, but reliable way to log gains. The cycling nerd gives this approach 2 thumbs up!

Do Big Rides!

The big, or epic ride ethos pervades cycling culture. Many of us who follow pros on Strava notice that 5, 6, and even 7 hour training rides are pretty common. And speaking of Strava, those amateurs that follow suit are sure to get a lot of kudos for this type of ride. But how well do big rides correlate with WattsDiff? We arbitrarily decided any ride over 3 hours of moving time counts as a big ride, and then looked at big rides vs WattsDiff a couple of different ways.

KJs incurred over 3 hours vs WattsDiff

The first way we looked at big rides focused on the number of KJs incurred after 3 hours. This approach has the advantage of treating the epic-ness of an athlete's big rides on a continuum. Those that only worked for a few minutes over 3 hours would get less credit than those putting in 4 or 5 hour rides. The correlation was very weak here (r(274) = 0.0598, p > .05 (not statistically significant)).

Big Ride Weekly Frequency vs WattsDiff

Perhaps insisting on a strict correlation between KJs incurred over 3 hours and WattsDiff is the wrong way to look at this. Instead, we decided to look at big ride frequency. A fair number of coaches advise taking one big endurance ride per week (we continued to use the 3+ hour standard). So we divided the periods up into weeks, and gave the athletes 1 credit for each week they adhered to this big ride discipline, and correlated it with WattsDiff. The correlation was negative here r(1368) = -0.0297, p > .05 (not statistically significant).

So, what's going on here? We looked at the possibility that big rides might be causing enough fatigue to inhibit performance over the next few days after that big ride. We found that fatigue from the big ride is not a factor, at least insofar as putting out KJ's is concerned as overall KJ from the big ride + next 4 days > overall KJ from small rides + next 4 days (Small Ride AvgKJ = 2793, Big Ride Avg KJ 4291). We also looked at the quality of the KJ's incurred from both sets of data. Maybe it was the case that big ride fatigue caused athletes to back off of interval training. But, looking at interval KJ on the same basis set forth directly above, BigRide AvgInterval KJ = 711.172236790608, and SmallRide AvgInterval KJ = 601.311999295916.

We can only speculate at this point that the dampened PR performance outcomes experienced by big ride athletes may be due to overreaching. While we can't definitively establish overreaching as a factor with the available data, it's worth considering. Overreaching is a state that occurs when an athlete experiences a temporary decline in performance due to an increased training load. The Cycling Nerd gives big rides a thumbs down in terms of a method to optimize watts gains. But remember, there still is a positive correlation between overall KJs and WattsDiff. So, if you're one of those that only has time to log significant rides on weekends you should keep right on doing so.

Don't Buy Upgrades, Ride Up Grades!

Elevation vs WattsDiff

The headline above is yet another quote from the great Eddy Merckx. There is no doubt that the cult of climbing is alive and well in the cycling world. The efficacy of climbing to build adaptations is backed up by both the experience of coaches and athletes, and by science. However, just like any other tool, climbing rides must be used properly to make a difference. Our athletes didn't do so judging by the tiny positive correlation of elevation to WattsDiff r(402) = 0.0017, p > .05 (not statistically significant). The Cycling Nerd gives elevation, in and of itself a thumbs down in terms of a method to optimize watts gains. Pedaling 120 watts up a hill produces no more adaptations than pedaling 120 watts in the flats. Properly used though, elevation can be a great vehicle for creating adaptations.

Do Lots Of Rides!

Frequency Of Rides vs Watts Diff

Lost in the noise generated by epic rides, and rides with huge elevation gains, are those athletes that keep grinding day in and day out. Their individual rides are not that impressive, but they ride consistently, many once a day, racking up big weekly miles and hours. We've already demonstrated that the former strategies are not that great at producing gains. So, does this decidedly unflashy strategy work better? We found a weak, but statistically significant correlation between frequency of rides and WattsDiff. r(402) = 0.1852, p < .05 (statistically significant). This is actually a little better than just riding lots of KJs (r(274) = 0.1832, p < .05 (statistically significant)). If your schedule permits, daily rides, even if they aren't that difficult can be a slow but reliable way of producing performance gains. The cycling nerd gives this approach 2 thumbs up!

Discussion

This article has provided a pretty clear blueprint on how athletes can optimize watts gain from the endurance portion of their training. Ride lots, and ride frequently. Big rides that impress your Strava followers can be counter-productive. Getting lots of climbing in won't help unless you actually work harder on the climbs.