Service points won: the statistic that explains most of a tennis match
TL;DR. In tennis, the percentage of points won on serve is the statistic that concentrates the most information: it describes serve power, the quality of the first shot after it, and how solid the player is in their own service games, all at once. Aces are spectacular; this metric is what actually explains who controls the match.
Note. Informational and statistical content. It is not betting advice nor a promise of profit. No statistic predicts a single match: they are there to estimate probabilities over the long run. 18+ only. Play responsibly.
Why it beats the ace count
An ace wins one point. A good serve wins many more points than it finishes outright: it puts the returner on the back foot and hands the server a comfortable shot to close with. That chain of points won thanks to the serve, but not with the serve, is what the service points percentage captures and the ace count does not.
The practical upshot is that a player with few aces can post an excellent service points percentage — precise rather than powerful serving — while a big server with a fragile second delivery can post a mediocre one.
First and second serve are read separately
| Metric | What it describes |
|---|---|
| % of points won on first serve | The player's ceiling when their best shot lands |
| % of points won on second serve | Their floor: what happens when the first one misses |
| First serve percentage in | How often they reach that ceiling |
| % of service points won (overall) | The summary of the three above |
The second serve is where most tight matches are decided. A player with a devastating first serve and a weak second lives on a tightrope: every missed first serve makes them vulnerable, and on the big points tension drives the first-serve percentage down exactly when it is needed most.
The surface changes everything
The same player posts very different percentages depending on the court. On fast surfaces the serve dominates and everybody's numbers rise; on clay the high, slow bounce gives the returner time and the numbers fall.
So a season average blending surfaces describes an average player who does not exist. The honest comparison is always within the same surface, and it is why our model keeps a per-surface rating alongside the general one — we cover it in surface statistics.
What the model does with the serve
The probability of a player winning the match comes from an Elo rating blended with their specific rating on the tournament's surface. That rating absorbs service performance indirectly, because serving is what produces the results it updates on.
The direct read of the service percentage is published as context on the match page, so you can see where each player's profile comes from.
Frequently asked questions
Should I look at aces or points won on serve?
Points won on serve. A good delivery wins many more points than it finishes outright, because it puts the returner on the back foot. Aces only count the ones that end on the spot.
Why does the second serve matter so much?
Because it is the player's floor and it is where tight matches are decided. A devastating first serve with a weak second leaves the player vulnerable every time the first one misses, exactly when tension drives the first-serve percentage down.
Can service percentages be compared across surfaces?
Not without correcting. On fast courts the serve dominates and everyone's numbers rise; on clay they fall. An average blending surfaces describes a player who does not exist.