Science and Secrets Part 3: Training + Recovery = Adaptation

By Danielle Farinella

I’ll start off by saying that this article contains some advanced topics which are most definitely not necessary for success at the master or youth rowing levels. However, if you enjoy getting into the math and science of it all, you may find them interesting. You will want to be familiar with the concepts in Parts 1 and 2 in this series before reading Part 3.

So far in this series, we’ve covered the building blocks: workouts. But really, a huge part of training effectively is thinking about how those workouts stack together and how we recover from those workouts.

FOLLOW US ON SOCIAL MEDIA 

for TRRA updates 

The Impulse Response Model

The research is pretty clear that Training + Recovery = Adaptation of our systems, which makes us faster. However, how do we know how much training to do, or how much recovery to do, or when to train and when to recover? Realistically, most athletes will use a combination of precedent (“well lots of people train 6 sessions per week with one rest day, so I guess I’ll do that too”) and trial and error (“oops, I feel really fatigued doing this, I’ll shift to 5 days on and 2 days off”).

But, sports scientists have been trying to quantify and model these values using something called an Impulse Response Model (a name ripped-off from the physics field) (Busso, 2023). There are tons of different models, and they usually look something like this oversimplification:

Performance = Initial Performance + Fitness - Fatigue

Fitness = the sum of your Daily Training Loads over a period of time, accounting for a time constant for fitness degradation and a time constant for fitness increases.*

Fatigue = the sum of your Daily Training Loads over a period of time, accounting for a faster time constant for fatigue degradation and a faster time constant for fatigue increases.*

Daily Training Load = Intensity of workout * Duration of workout (often done by assigning Intensity Factors to each of your training zones, and multiplying your duration in each zone by those IFs). To determine zones, see Science and Secrets Part 2.

*These contestants change across models, but in nearly all cases, your fatigue will “degrade” faster than your fitness. This is (theoretically) why tapers work. There’s also a whole subset of research trying to accurately define those constants, which likely need to be done per athlete to be accurate.

If trying to identify all those constants and values sounds like a lot of work, never fear, there is software that will do it for you. In fact, if you wear a smart watch or use the Strava fitness and freshness graph, these are already being calculated for you!

I like to use Training Peaks, and you can see the model I described above play out in the “Performance Management Chart”. Here (below), we can see the pink line is Fatigue, the blue line is Fitness, and the orange line is Performance. The red dots are the Daily Training Loads. The blue dots are the Intensity Factors for that day. This was from January to February, when I was training pretty consistently.

Okay, so this tells us about the past, but how do we know the future? 🔮

In Training Peaks, you can plan out your workouts in a calendar so that it will calculate all of these values for future sessions. This graph (below) shows the last 90 days of my training, plus an estimation of the next two weeks (dashed lines). You can see I have a taper planned for the week of 4/6, since that is where Fatigue starts to decrease and Performance rises.

Overall, what you will likely find using this method is that you will need to stack multiple days of quality training per week to increase fitness. So not rocket science! But it can help you feel more confident in your training plan.

It’s worth noting that studies have found these predictions to be imperfect when predicting performance, likely due to factors these models can’t account for, such as recovery quality, differences in time constants across athletes dependent on training history, and lack of isolation of each energy system (although this interesting paper is working on that problem -> Kontro, 2026).

Be sure to listen to your body, don’t just listen to the model!

If you’re thinking about getting set up with an Impulse Response Model for your training, it is important not to blindly follow the model! We want to stay in an optimal-ish range where fatigue is not too high, but our fitness is still increasing. How do we know when fatigue is “too high”? If your resting heart rate is higher than usual for a prolonged period, you feel agitated for a prolonged period of time, or your easy UT2 pace has slowed considerably, those all are signs it may be time to take a day or two off, and note your current fatigue value as an “upper boundary” for future planning.

Works Cited:

Clarke DC, Skiba PF. Rationale and resources for teaching the mathematical modeling of athletic training and performance. Adv Physiol Educ 37: 134–152, 2013; doi:10.1152/advan.00078.2011.

Kontro H, Mastracci A, Cheung SS, MacInnis MJ. The three-dimensional impulse-response model: Modeling the training process in accordance with energy system-specific adaptation. PLoS One. 2026 Feb 6;21(2):e0341721. doi: 10.1371/journal.pone.0341721. PMID: 41650016; PMCID: PMC12880663.

Morton, Richard & Fitz-Clarke, J & Banister, E. (1990). Modeling human performance in running. Journal of applied physiology (Bethesda, Md. : 1985). 69. 1171-7. 10.1152/jappl.1990.69.3.1171.

Thierry Busso, Sébastien Chalencon. Validity and Accuracy of Impulse-Response Models for Modeling and Predicting Training Effects on Performance of Swimmers. Medicine and Science in Sports and Exercise, 2023, 55 (7), pp.1274-1285. ⟨10.1249/mss.0000000000003139⟩. ⟨hal-04713855⟩