
What is percentage tennis and how do you play it
Miss less on purpose. Make them hit the great shot.
See it, eight instants at a time.
Every frame is traced from real match footage, so you see the shape of the stroke, not a drawing of an idea.

Key facts
- 33.5%
Borderias and colleagues report 33.5% winners, 43.6% forced errors and 22.8% unforced errors in 21 women’s doubles matches from the 2019 Australian Open, Roland Garros and Wimbledon. The simulator uses these rounded values as its starting mix.
Borderias et al. (2023), Study II in Borderias’s thesis, pp. 60–67
- ~61%
Cross’s historical comparison of 15 men’s matches per Slam, from the last sixteen onward at four events ending with the 2008 Australian Open, reports about 61% of points won through opponent errors at most events and 83% at the French Open.
Cross, Grand Slam Statistics, University of Sydney
- 10.5–10.6%
In that French Open sample, clean winners made up 10.5–10.6% of points won, reported separately for match winners and losers. This is a historical sample, not a current tour average.
Cross, Grand Slam Statistics, University of Sydney
- 10%
In one bootstrap model of Federer–Nadal matches, reducing Federer’s unforced-error rate by 10% changes his simulated match-win percentage from 44.5% to 53.7%. The model assumes his level and style otherwise stay the same; this is not an observed training gain.
Peiris, Epasinghege Dona & Swartz (2025), Table 4
- 72.1%
At Wimbledon 2015–17, the mean share of points per match lasting 0–4 successful shots was 72.1% for men and 65.9% for women. Errors were excluded from the shot count.
Fitzpatrick (2021), Table 7.3 and rally-length definition
Why it works
Brad Gilbert reached the top five without a weapon by attacking weaknesses and imposing uncomfortable patterns. Percentage tennis is that approach with arithmetic. It's aggressive with margin, not pushing. The goal is to make your opponent beat you with a great shot.
Margin means the net window and the depth band, not a line. Clear the net comfortably and land past the service line, and a heavy, deep ball still attacks. Two metres inside the lines is still a winner if they're running.
Where you are decides what's allowed. Split the court into defend, neutral and attack thirds: defend to reset, and attack from the attack third. On big points, take the bigger half of the court, the weaker wing and the proven serve.
The line to say on court“Make them hit the great shot.”
Drill: The error count
15 min · with a partner- Warm up crosscourt with generous net clearance and a target beyond the service line. Agree your ending definitions before counting. A missed serve is not a point unless it is the second fault.
- Play 20 points. Count who won each point and its ending: winner, forced error, unforced error or unclear. Keep disagreements unclear; do not force agreement. Count the unclear points you lost.
- Play 20 more with no line attempts from behind the baseline. Keep the same opponent, court side and serving pattern. Compare points won and your own unforced losses, keeping unclear losses separate. Allow extra time to finish; an incomplete block is not a 20-point result.
More drills like this: point play
Faults and fixes
Questions players ask
Do most tennis points end in errors?
Errors were the majority in the samples shown here. Borderias and colleagues report 66.4% forced plus unforced errors in 21 women’s doubles matches from three 2019 Slams. Cross reports a different historical men’s singles sample. Neither establishes a universal rate for your matches; keep your own count.
Is percentage tennis the same as pushing?
No. Pushing sends short, high balls that invite the attack. The percentage ball is heavy and deep: aggressive, with margin.
Should I go for the line off a mid-court ball?
Not from the middle of the court a step inside the baseline. Hit heavy cross-court deep and follow it in. The line becomes right when the ball is shorter and you're further in.
When is a drop shot the percentage play?
Only when your opponent is well behind the baseline.
What this doesn't cover
The ending studies cover different formats, eras and samples: women’s doubles in 2019 and historical men’s singles. Their categories also depend on observer judgment. They cannot establish your error rate or the best tactic for a particular ball. The simulator is an authored what-if exercise; its equal split between opponents and independent-point scoring are assumptions.
Sources
- Borderias et al. (2023), women’s doubles point endings — Study II, pp. 60–67 in Borderias’s thesis
- Cross, Grand Slam Statistics, University of Sydney
- Analysis of the Impact of Unforced Errors in Tennis (SAGE, 2025)
- Heavy Topspin, the winner-unforced error ratio (2015)
- Brad Gilbert, Winning Ugly (1993)
- O'Shannessy, Grand Slam charting
- Match Charting Project
- Fitzpatrick (2021), Table 7.3 and rally-length definition