A little bit more about the constant fatigue “model”

I got an email from another highly respected coach (he asked to be called that!) asking me to elaborate a little on how to implement in a weekly routine he constant fatigue model that I talked about here, and how it differs from other ways of scheduling training.

In order to talk about this, it might be helpful to define a subjective, qualitative training stress score that I will call TRS, Training Response Score. For a simple definition of TRS, let’s say that the total training load that any athlete can handle in a week represents 700 TRS, which means that the average daily TRS is 100. With this average daily TRS defined, it is easier to qualitatively explain the constant fatigue level model, and even compare it wth other scheduling models. Given that TRS is a measure based on adaptation to training, it is closely linked to the state of fatigue/recovery of the ahlete. Therefore, constant fatigue levels should be obtained by aiming to have daily TRS’s of around 100.

Even though for the sake of simplicity, the time frame of one week was mentioned, we need to remember that adaptation to training occurs in periods of 6-8 weeks. So when we are talking about the 700 TRS an athlete can handle in a week, we really should be thinking of the amount of training load an athlete can perform in a 6-8 week block, 4200-5600 TRS. This training load is something that falls into a very narrow interval.

As we all know, specificity has, or should have, a big impact when designing the training plan. Therefore, specificity requirements might deviate us from the “perfect” implementation of the constant fatigue model. This is particularly true for Ironman training, where for specificity, it is often needed to schedule days which will bring a TRS in excess of 100. To give an example, a training session that is common for IM athletes to do, a 5 hour ride followed by a 40min run, is a session that might have a TRS of over 150 for some athletes.

So how do you couple the constant fatigue model with the specificity needs? I feel that the best approach is to have a hard day/easy day approach, where you seek recovery, hence adaptation, in the short term. In terms of TRS, this would mean that following a 150 TRS day, you would schedule a 50 TRS.

An alternative way to schedule training is to center microcyle planning around 2-3 “big days” in a week. Let’s try to see this with the help of the TRS. A “big day” might represent some 150-200 TRS and 2-3 days can mean a training load of 400-500 TRS. Because we have a “budget” of 700 TRS for the week (or more important 4200-5600 TRS for the block), and because such stenous sessions require significant recovery in the following days, a good part of the week is comprised of low TRS days. The practical consequences of what was stated above are pretty straightforward. There is only so much and athlete can handle inside 6-8 weeks. So if you are seeking adaptation to training, something that apparently is not a goal for some, if you are going to do “big days” or “big weeks” of training, they will be followed by small days or small weeks in order to balance out the adaptation equation. And this is the important detail that never makes it to athlete’s blogs.

Some of you must be asking right now this: What is the difference between the two approaches, providing you respect the same TRS during a 6-8 week block?
The main difference I find is the predictability in response to training that you get by using a constant fatigue/recovery model. By seeking to find the training load that induces a constant recovery timeframe, you can easily answer one of the more important questions when scheduling training: “Is the athlete recovered in order to be able to complete the goals of the next session?”


As mentioned in an earlier post, the training process can be loosely modeled as a series of sequenced processes composed of:

Training Load (session) -> Fatigue -> Adaptation(recovery) -> Performance

By trying to maintain constant the load/fatigue part of the process, we can more easily predict the effect that the scheduled sessions have in the overall training plan and hence the impact in performance.