After a big week, you are both fitter and more tired than before. Both are true at once, and that is the whole point of the form model: it separates what training has built from what it is still costing you. Three numbers, drawn from the load of your activities (Training load): fitness, fatigue and freshness.
Banister’s idea: form is fitness minus fatigue
In 1976, Eric Banister and his colleagues proposed seeing each session as an impulse with two effects. A positive one, fitness, which builds slowly and lasts. A negative one, fatigue, which rises fast and fades fast. Performance on a given day would be the difference between the two. In their model, fitted to a swimmer, the fitness effect faded over about fifty days, the fatigue effect over about fifteen.
In the early 2000s, Andrew Coggan turned this into a version cyclists could compute easily, known by three acronyms:
| Acronym | Original name | Our word |
|---|---|---|
| CTL | Chronic Training Load | Fitness |
| ATL | Acute Training Load | Fatigue |
| TSB | Training Stress Balance | Freshness |
The acronyms are the ones you’ll meet elsewhere, the last one often called “form”. Here, we stick to the words.
Fitness: what six weeks have built
Fitness is the average of your daily load over the past weeks, each day weighing a little less than the one after it. Allen and Coggan use a time constant of 42 days: it takes about six weeks for fitness to reflect your training.
It is expressed in units of load per day. A fitness of 60 is the equivalent of 60 load every day, rest days included: 420 over a week, however it is spread.
It rises slowly and falls slowly. One hard week barely moves it; three weeks off set it back clearly. It is the base you ride on.
Fatigue: the load of the last seven days
Fatigue is the same average over 7 days instead of 42. So it reacts fast: a big week sends it up, a few easy days bring it back down.
On a chart, it looks like a much jumpier version of fitness: above it during a block of work, below it as soon as you ease off.
Freshness: fitness minus fatigue
Freshness is the difference between the two, taken as you wake up, before the day’s activity.
- Negative: fatigue is above fitness — you’re building.
- Near zero: you’re maintaining.
- Positive: fatigue has dropped below fitness — you’re fresh.
Coach Joe Friel gives these markers:
| Freshness | What Friel says |
|---|---|
| Above +25 | Training is too easy: fitness is being lost |
| +15 to +25 | Usually the best range on the day of a major goal |
| −10 to +10 | A transitional zone, where two weeks or more is seldom a good sign |
| −10 to −30 | Hard, focused, productive training while building |
| Below −30 | A greatly increased risk of breaking down |
They are markers, not rules: Friel himself points out that the chart doesn’t remove the differences from one athlete to the next.
Ramp rate: how fast to build
Perhaps the most useful number day to day: how much your fitness rose over the past week. Friel finds that a rise of about 5 to 8 points a week is about right for most athletes. Beyond 10 is a “crash” period some can manage, but he advises keeping it to a week or less before backing off to recover.
Building too fast means piling up fatigue your fitness can’t keep pace with (Building your load without going into the red).
Reading your chart
Three situations come up again and again:
- A block of work: fitness rises gently, fatigue climbs above it, freshness sinks below zero. An easy week brings it back up (The recovery week).
- A taper before a goal: you ease off, fatigue falls much faster than fitness, and freshness rises. That is the whole idea: arrive fresh, having lost as little fitness as possible (Arriving fresh for your goal: the taper).
- A break: everything falls. Freshness turns strongly positive, but fitness leaves along with the fatigue.
What the model can’t know
- It only sees load. Sleep, stress, an illness coming on don’t enter into it. Positive freshness doesn’t guarantee you’ll feel good.
- It depends on your thresholds. An FTP set too low inflates every load, and every curve with it (What FTP is, and how to test it).
- Its constants are the same for everyone. 42 and 7 days are averages; some athletes recover faster than others. Banister’s original model fitted these constants to each athlete, and a study showed how uncertain its predictions remained from one athlete to another.
- It needs history. Starting from zero, it takes about six weeks before fitness can be read.
So the model doesn’t predict your performance to the watt. It tells you a trend: whether you’re building, overdoing it, or arriving fresh. That is already a lot.
In Komeak
The Form tab answers “where do I stand?”.
- Today: your fitness, fatigue and freshness, with freshness put into words — from “Very rested” to “Very tired”, on Friel’s ranges — and your fitness rise over seven days, falling, steady, rising or too fast, with the advice that goes with it. Komeak treats +3 to +8 a week as reasonable; beyond +8, it suggests one or two easy days.
- The form curve: three charts on one time axis — fitness and fatigue, freshness over its ranges, and each day’s load. Pick the period, from six weeks to a year. Tap a day to read its values and its activities. A marker shows each change of FTP, threshold heart rate or maximum heart rate, which explains a jump in the curve.
- On the home screen, the Form card gives today’s state, and warns you when your fitness rises by more than 8 in a week.
- On an activity’s page, its effect on your form: your fitness before and after, and your freshness on waking, that day and the next.
- Under the curve, your recovery: what the model can’t know, your body tells. Two minutes on waking with your heart rate strap give your resting heart rate and your HRV; the Recovery section follows them day after day, the heart rate against its current value, the HRV averaged over 7 days against your zone, with a reading in words and advice when one is called for.
Fitness −3 over 7 days: falling. That is normal after a few easy or rest days.
Every activity feeds your form, whether it was measured by power, heart rate or perceived effort. For the first six weeks, your form is marked as building, with the date it will become reliable.
Sources
- Calvert T. W., Banister E. W., Savage M. V., Bach T., “A Systems Model of the Effects of Training on Physical Performance”, IEEE Transactions on Systems, Man, and Cybernetics, SMC-6(2), 1976, pp. 94–102.
- Allen H., Coggan A., McGregor S., Training and Racing with a Power Meter, 3rd ed., VeloPress, 2019.
- Friel J., “Part 3: Training Stress Balance… So What?”, 12 July 2015. joefrieltraining.com
- Friel J., “The CTL Ramp Rate”, 15 May 2015. joefrieltraining.com
- Hellard P. et al., “Assessing the limitations of the Banister model in monitoring training”, Journal of Sports Sciences, 24(5), 2006, pp. 509–520.