← Blog

How often should you test speed?

Weekly testing tells you about the weather. Once a season tells you nothing useful in time to act on it.

Coaches tend to land on one of two extremes. Either speed gets tested constantly, which turns every session into a competition, or it gets tested once in September and never again, which means the programme is unfalsifiable.

Why weekly is too often

The week-to-week variation in a sprint test (from sleep, weather, surface, fatigue, mood and where the athlete is in their training week) is comfortably larger than a week of genuine adaptation. Test every week and almost everything you see is that noise. Worse, athletes learn that Tuesday is a test, and a maximum effort every Tuesday is a training decision you did not intend to make.

Why once a year is not enough

A single measurement at the start of the season cannot tell you anything, because there is nothing to compare it to. A single measurement at the end tells you what happened after it is too late to change it. Testing exists so you can act on the answer.

A reasonable rhythm

• A baseline in the first two or three weeks, once athletes are conditioned enough to sprint safely.
• A check at the end of each training block (roughly every six to eight weeks), which is long enough for an adaptation to show above the noise.
• One in the competition phase, to see what the racing itself is doing to the numbers.
• The same test each time. A flying 20 in September and a flying 30 in March are not comparable, and neither are two tests with different run-ups.

Three or four times a season, in other words. Enough to catch a block that is not working while there is still time to change it, rare enough that each result means something.

Hold the conditions still

Test at the same time of day, on the same surface, at a similar point in the training week, and record the wind. A test whose conditions are not written down is a number you will not trust in four months, and you will be right not to.

Keeping the runs is the part that gets skipped. StatLap's speed gates save each run into the training section, apart from meet results, so a season of tests sits in one place instead of across four phones and a notebook. That matters most in the second year, when the question stops being whether an athlete improved this block and becomes whether they are on the trajectory they were on last spring.

Two honest limits. There is no dashboard that charts an athlete's speed tests over time and tells you the trend. The runs are stored, and reading them is on you. And a test done when an athlete is tired, ill or unusually cold is not a data point about training; it is a data point about that day, and the useful thing to do with it is write down why and move on.

Related