← Blog

Why your 40 times keep dropping and nobody is faster

If the whole squad improved by two tenths this month, the thing that changed was probably the timing.

It is a good feeling and it should be a suspicious one. Every athlete on the sheet is faster than last month. Not one of them is slower. Somebody is about to put it on social media.

Squads do not improve uniformly

Real training effects are uneven. Some athletes respond to a block, some do not, some were ill in week three, and one grew and is temporarily worse at controlling limbs that are longer than they were. A month of genuine improvement looks like most athletes a little better, a few unchanged and one or two mysteriously backwards.

A whole squad improving by a similar amount in the same direction is the signature of a measurement change, not a training effect.

The usual causes

• A different person on the watch. Reaction times vary between people by more than the improvement you are hoping to see.
• The same person, more invested. A coach who has just spent six weeks on a speed block starts the watch marginally late and stops it marginally early. This is not cheating; it is being human and having a stake.
• A changed start. Timing from first movement instead of from a command, or an athlete who has learned to roll into it, moves everything.
• A different surface, shoe or wind. A tailwind and fresh spikes are worth more over 40 metres than a training block usually is.
• A shorter distance. A tape measure that was pulled tight last time and casually this time is a hundredth or two, every run.

The fix is to remove the person

Not because anybody is dishonest, but because human reaction is a bias rather than noise. It runs in one direction and it moves with expectation. Automatic timing removes it entirely, which is the whole reason sanctioning bodies add a correction to hand times before comparing them to electronic ones.

StatLap's speed gates do this on two phones: one watching the start line, one watching the finish, the distance between them typed in. Nothing about the number depends on anybody's thumb, so a squad that gets faster on gates got faster.

Then hold the rest of it still

Automatic timing fixes the timer and none of the other variables. If you want month-to-month numbers that mean something, test at the same time of day, on the same surface, in similar weather, at the same point in the training week, with the same start command and the same measured distance. Write down anything you could not hold constant, especially wind.

Gate runs are stored in their own training section rather than mixed in with meet results, which matters more than it sounds. A Tuesday fly-in that drifts into an athlete's competition history is a number somebody will eventually quote at a recruiter.

The test to apply to any improvement: could this have been the instrument? If the answer is maybe, it was.

Related