What is CTL, ATL, TSB? Fitness, fatigue and form explained

What is CTL?

CTL stands for chronic training load and is often called fitness in a training-load chart. It is a longer-term weighted average of daily load, so a sustained pattern of riding matters more than one isolated session. The label describes a model value; it does not directly measure all aspects of your physical fitness.

In Glood, fitness is a 42-day exponentially weighted average of daily training load, fatigue is a 7-day one, and form is fitness minus fatigue, so positive form means fresher.

An exponentially weighted average gives the newest daily load a contribution while retaining a gradually shrinking influence from earlier days. The 42-day period describes the timescale of that response. It is not a box that keeps the last 42 days unchanged and deletes everything older on the next morning.

Several weeks of a different riding pattern can therefore change the longer-term line. A single large session affects it, but does not erase the previous history. Use the line to understand the recent training pattern, rather than chasing a higher value as a goal independent of the riding you want to improve.

What is ATL, and how does it differ from CTL?

ATL stands for acute training load and is often called fatigue: it follows recent load on a shorter timescale than CTL. A shorter exponentially weighted average responds faster to changes in daily input. That makes it useful for describing a recent run of riding, without making it a direct measurement of tiredness.

If daily load increases for several days, the shorter average reacts sooner than the longer one. If the riding becomes lighter, it also falls sooner. That difference in response is the reason the two lines can separate during a demanding block and move closer together during easier riding.

The word fatigue can cause confusion. The chart is summarising training inputs, while your actual experience also includes sleep, work, travel, illness and other demands. A low value does not establish that you feel rested, and a high one does not explain every reason a ride felt difficult.

Look at both the line and the week that produced it. A few long outings, daily short rides and an incomplete activity record can lead to different pictures. Before interpreting a sudden change, check whether the recorded sessions match what you actually did.

What is the CTL, ATL and TSB formula?

The basic relationship is form equals fitness minus fatigue, with each average updated from the previous day and the new daily load. The averaging method and timing must be specified to reproduce a particular chart. A plain arithmetic mean over fixed windows is different from an exponentially weighted average.

Glood’s daily updates are fitness = previous fitness + (daily load − previous fitness) / 42; fatigue = previous fatigue + (daily load − previous fatigue) / 7; form = updated fitness − updated fatigue.

For a worked arithmetic example, suppose the previous fitness and fatigue values are both 40 and the next daily load is 82. Using those updates, fitness becomes 41 and fatigue becomes 46, so their difference is −5. These are invented input values to demonstrate the calculation, not targets or a recommended training session.

If you reproduce a weighted average in a spreadsheet, put daily load in one column and carry each updated average into the next day’s calculation. You also need an initial value and a consistent rule for days with no riding. Different starting values can matter especially near the beginning of a short record.

Daily load is itself an input calculated by a method. These equations do not specify how every possible app scores a ride. Do not assume that equal distances, equal durations or equal numbers copied from separate systems represent exactly the same input.

What does positive or negative form mean?

Positive form means the longer-term fitness value is higher than the shorter-term fatigue value; negative form means the reverse, under the fitness-minus-fatigue convention. The sign helps describe recent load in relation to the longer history. It does not guarantee how you will perform or how you should train today.

Imagine fitness of 50 and fatigue of 40: the difference is +10. Swap the two values and it is −10. This example explains subtraction only. There is no medically meaningful dividing line established by those example numbers, and a value near zero is simply a small difference between the two averages.

Other platforms may differ in sign convention, inputs or when the day’s value is calculated. Check their definitions before comparing charts. A familiar word such as form is not enough to establish that two systems use the same equation or that matching scores have the same meaning.

Avoid treating form bands as medical advice. The chart cannot diagnose a condition or tell you whether exercising through symptoms is appropriate. If you feel unwell or have concerns about starting or continuing exercise, seek medical advice rather than trying to settle the question with a number.

Why can form rise during an easier week?

Form can rise during easier riding because the shorter-term average falls faster than the longer-term average when daily load drops. Their difference can therefore become more positive while total riding decreases. That is a consequence of the model’s response rates, rather than proof that a rest day has instantly improved your ability.

The same relationship helps explain a negative difference during a period of heavier training. The recent-load line can move ahead of the slower one. Neither situation should be judged in isolation: first identify whether the week was intended to build work, allow recovery or fit around other commitments.

A long break can eventually bring both averages down. A positive difference after that break does not establish that you retain the same capacity as before it. Compare the chart with recent rides and your own experience, rather than reading fresher as a synonym for stronger.

Where does the fitness–fatigue model come from?

The fitness–fatigue model comes from Banister and colleagues’ work in the 1970s, which described training as producing positive and negative effects with different rates of decay. Modern load charts use related ideas, but a simple pair of averages is not the full original performance model or a guaranteed prediction of race results.

An early reference is Banister, Calvert, Savage and Bach, A systems model of training for athletic performance (1975), followed by their systems-model work in 1976. The useful historical idea here is that a training stimulus can have effects operating on different timescales. That does not make today’s chart a complete account of the rider.

For a practical review, start with data quality, then look at the pattern over several weeks. Put notes about sleep, illness and how the rides felt beside the chart. A model is most useful when it helps you ask better questions about your recent riding.

How can I read my own training history?

Review the recent pattern, check the underlying activities and compare it with how you felt before drawing conclusions. Avoid comparing one isolated score with another rider’s. Similar numbers do not establish similar climbing ability, endurance or readiness; those depend on more than the recorded load history.

Glood lets you review an activity’s training load and see your own fitness, fatigue and form over time, calculated from recorded activities.

Use the review to decide what you need to understand next: a missing session, a demanding run of days or a break from riding. Keep the purpose of the next week in view. There is little value in increasing the load simply to make the fitness line rise if that no longer serves your riding or leaves enough recovery.