Most sprint tests I see go wrong in one of two ways. The athlete runs once and that one number becomes the result. Or the athlete runs eight times with thirty seconds between reps, and the test turns into conditioning.
Neither tells you how fast the athlete is today. One rep can be a bad start. Eight short-rest reps measure how tired they are getting.
Here is the version I use with phone timing gates.
Two or three reps, all-out
For a 20 m test, I time two or three reps per athlete. Every rep is a full effort from the same start position. No "just a practice one" that gets timed and then argued about later. If a rep is a warm-up, it is not timed.
Three reps is usually enough to get a fair best. If the first two are close, the third often adds nothing, and that is fine. You can stop at two.
Rest like you mean it
The rest between reps matters more than the number of reps. For a short sprint I give a few full minutes of walking rest between reps, and longer for older or stronger athletes who clearly need it.
Rushing the rest is the easiest way to ruin a test. The second rep looks slower, and you cannot tell whether the athlete is slower or just not recovered.
With a squad this solves itself. Run the athletes one after another down the same lane. By the time the last athlete has gone, the first one has had their rest.
Keep the best, not the average
I take the best rep as the test result. The average punishes an athlete for one slipped start, and a slipped start says nothing about speed.
The best rep is also what you compare next time. Same distance, same start, same surface, same number of reps, same rest. If any of those change, the comparison is weaker.
Stop when the times fall
If a rep is clearly slower than the one before it, after proper rest, the athlete is done for the day. Do not chase a better time with a fourth and fifth rep. A fall in speed on a fresh test usually means fatigue, a tight hamstring, or a bad day, and none of those get better with more sprints.
The same goes for any rep that looks wrong: a stumble, a false start, an athlete who pulls up. Note it, do not count it, and decide whether the athlete should run again at all.
How I set it up
Two phones become the start and finish gates in the SpeedTrackr Android app. The second phone can be another Android phone, an iPhone or a laptop on the web app. It is practice-grade start and finish timing: one time per rep for the whole distance, no split times, not a fully automatic timing system.
Before the session, make sure every Android phone is on the latest version from Google Play, so the start and finish phones pair without trouble.
If you coach a group, turn on Coach Mode and pick the athlete before each rep. Every time then lands on that athlete's page. The Squad tab shows when each athlete last tested and tags a NEW BEST when they beat an earlier result, so after a test day you can see who improved without reading through every rep. Coach Mode works in the Android app and on the web app.
The short version
Two or three all-out reps. A few minutes of rest between them. Keep the best. Stop when the times drop. Do it the same way next time.
It is not complicated, but it is the difference between a sprint test you can compare week to week and a set of numbers that just tells you who was tired.
The Android app is on Google Play: https://play.google.com/store/apps/details?id=com.speedtrackr.reactionpro
On an iPhone or a laptop, use the web app: https://www.speedtrackr.com/auth/login
