Power Meter Concepts

So, I ran across an interesting couple of blog posts by Joe Friel.

Estimating TSS and Projecting Race Readiness

In "Projecting Race Readiness", he describes how you can use your power meter data and software to track your fitness, fatigue and form. From those numbers, you can get a pretty good idea about how physically prepared you are to race. Nice, if you have a power meter, which I don't, and won't for the forseeable future.

In "Estimating TSS", he talks about how you can estimate training stress from RPE or heart rate and volume. Now, it's getting interesting! His intent (I think) was to show this for the triathlon community, where you're competing in multiple disciplines and may not be able to gather data in all of them, like swimming.

For me though, it suggests a crude method of projecting fitness without a power meter. In fact, I tried to do something similar with my own formula a couple years ago, but I couldn't get the numbers to work out to anything useful.

As I read and re-read these posts, and started playing with their implementation in a spreadsheet, I realized that a big part of the value to doing this is in planning your training relative to your races. I guess that's clear enough from the title "Projecting Race Readiness", but for some reason I didn't get it at first.

Anyway, so I put in some rough estimates for my workouts for the rest of the year, and marked my races. What I found was that, according to these numbers anyway, I will only be marginally "peaked" for a couple random races (with my best peak happening at the new Maybury CX race, which I may decide not to even do!). The numbers also show that I will generally not have good form for a lot of the races I do care about. (This actually reflects pretty well how past seasons have gone for me.)

(After just a little work...)

Here's where it gets tricky though. If you decide that you want to change your workout plan so that you get better numbers prior to a race, how do you go about doing it? It's not straightforward, since you're trying to optimize a couple outputs, and the "rolling sum" nature of some of the equations makes them difficult to work with. Still, for an engineer, it's an interesting problem.