The first serve percentage everyone quotes hides the real choice

A high first serve percentage can be evidence of a good day or of a server quietly refusing to take any risk.

Sixty five per cent first serves in. The commentary treats it as a virtue on its own, which has always struck me as odd, because you can raise that number to eighty any afternoon you like by taking twenty kilometres per hour off and rolling it down the middle. Nobody would call that an improvement. The percentage is not a skill, it is the visible half of a trade you made before the toss went up.

What the number is hiding is the distribution behind it. Two servers can both hit sixty per cent and be running completely different businesses. One is landing a heavy ball wide on the deuce court and living with a handful of long misses. The other is putting everything at the returner's hip because that serve almost never misses and almost never wins the point outright. The first player will lose more free points on the second serve and win far more on the first. Over three sets that difference dwarfs the percentage gap either way.

So the metric I would want, if the data existed in a form broadcasters could show, is points won per first serve attempt. That folds in the miss rate and the reward together and it stops rewarding timidity. A serve that goes in eighty per cent of the time and wins fifty five per cent of those points is worse than one that lands sixty per cent of the time and wins seventy five.

There is a caveat, and it is about the second serve. If your second serve is genuinely poor, the calculus changes and a conservative first serve is a rational hedge, because the punishment for missing is severe. That is the honest reason a lot of players creep upward in percentage as a match tightens. It is not confidence. It is a quiet admission about the second ball.

Which suggests where practice time should go, and it is not on the radar gun. Fix the second serve and the first serve gets braver for free. I have never seen that work in the other direction.

The best servers I have watched all have the same tell in a tight game. They miss the first serve long rather than into the net.

Long means they still swung.