More on Training Stress

A few more words on tracking training stress before I get back to bullheaded training.

Turns out my training stress score idea, like most ideas, is not original. "Daniel's Points" is a concept that is very similar to the IF^4 stress score I talked about in the last post. Apparently there was a running coach named Jack Daniels (no relation to the liquor, apparently), probably in the 80's, who made an intensity-to-the-fourth relationship between training stress and, in his case, running speed (which is almost directly proportional to power for runners, since wind resistance doesn't come into account). His runners had a fair amount of success on the collegiate scene.

What I like about it is it builds on the concept of NP in a very natural way, mathematically. Basically, with NP, you are defining a stress function that goes by P^4 (after smoothing to get rid of power peaks). NP is simply the steady state power which results in the average training stress. So, every point in time of the training ride corresponds to a stress that is proportional to P^4. The IF^4 proposal for training stress simply integrates this training stress in time.

Put another way, at every point in time during a workout, your body is exposed to a stressor (proportional to P^4) that it must adapt to. Add up all these stressors and you get an integrated adaptation stress, which is your IF^4 training score. Long, slow endurance rides, necessary to prepare your body for the stress of training, naturally don't score well on this scale - I think this is correct; a long endurance ride will not make you fast, whereas an hour of interval training will. This is in contrast to TSS where a 3 hour ride at endurance pace is given the same or better score as an hour of interval training. Threshold intervals will score equivalently on both scales.

Coggan's TSS is based on TRIMPS (you can google it; stands for "Training Impulse" and was designed for training by HR), which apparently started out as a score directly proportional to the training intensity. When this was found not adequate to describe what athletes were experiencing, another factor of intensity was added, making it an intensity-squared relationship. So, it's a guess with another guess added on.  Not the most elegant thing in the world.

This IF^2 relationship makes for a unique problem with TSS; namely, that TSS is a function of overall workout time independent of intensity. If you just add time at zero power to the end of your ride to make it longer, you can manufacture TSS points with obviously no gain in fitness. The IF^4 relationship contains no problem in this respect, because if you run through the math, it quickly becomes apparent that "overall workout time" cancels out of the equation and you just get a straight-up integration of the stress function.

Anyway, like I said, I am keeping track of both this year. I'll watch the IF^4 stress score with the most interest though, because it makes more sense to me than TSS. If training stress can be captured by a single number, it seems the IF^4 has the most theoretical backing and about the same amount of empirical backing as TSS.

At this point in my racing career, I am just curious about these things on an intellectual level. Comes from my engineering background, I am sure. This being my third year, I am just in "watch and learn" mode as far as these metrics are concerned - keeping track of various numbers and correlating them to my race results. I'm sure someone can design a system of training based on any system of keeping track of training stress. The important part, I would imagine, is the athlete's and coach's direct experience in tying training stress metrics to performance.