How to play big points in tennis
Spot the big point before it's played. Choose a plan you can execute.
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
- 2018
For 2018 ATP break points, Sackmann reported mean leverage of 10.5 percentage points assuming equal players and 6.5 using each match’s observed service rates. He chose 7.5 as a working approximation. These are different assumptions, not an uncertainty interval.
Jeff Sackmann, Measuring the Impact of Break Points (2019)
- 2015
At 15–30, winning reaches 30–30 and losing reaches 15–40. Sackmann’s 2015 analysis compares these branches using theoretical hold probabilities and match data.
Jeff Sackmann, The Pivotal Point of 15-30 (2015)
- 03
Match leverage depends on the set, game and point score, serving order and assumed point-win rates. A score label alone does not determine one universal value.
- 04
The interactive match explorer assumes independent points and constant service win rates in a best-of-five match. It is an illustrative scoring model, not a calibrated player forecast.
Why it works
Leverage is the difference between a player’s match-win probability after winning the next point and after losing it. At 30–40, the two branches lead to deuce or the end of the game. The set and match score determine what that game means.
The scoring model identifies stakes; it does not tell you which serve or rally shot has the best chance of winning. Our practice task is to choose a rehearsed target and next-ball plan, then adapt to the actual return.
Compare execution across targets and watch how the receiver adjusts. Rehearsal and a stated plan do not guarantee that pressure will leave execution unchanged.
The line to say on court“The score tells you it's big. The pattern tells you what to do.”
Drill: The big-point board
60 min · with a partner- Before play, choose one serve target with an intended next shot and one return plan. Agree how your partner will count the reset and any change prompted by the incoming ball.
- Count the first ten points whose pre-point score is 30–30, 30–40, 40–30, deuce, or a set or match point. Count each point once, including repeated deuce points, errors and double faults. These are practice triggers, not a ranking of model leverage.
- Before each selected point, reset and name the plan within the normal pace of play. Afterwards, award one process tick only if you reset, named the plan before play, and can describe keeping it or adapting to the actual ball. Missing or unclear checks earn no tick. Count shot execution and the point result separately; an early ending may leave no next-shot decision.
- The process ticks are a score only after ten selected points. If the match supplies fewer, a partial block doesn't count toward the number to beat; do not rescale it to ten. Start a new block after the tenth point.
More drills like this: mental game
Faults and fixes
Questions players ask
What are the most important points in tennis?
Compare the two possible next-point outcomes at the full match score. Break, set and match points can create large differences, but the leverage also depends on the assumed serving strengths.
Should I go for winners on break point?
The leverage model cannot answer that. Compare the shots you can execute from the actual ball and position; the lesson’s exercise starts with a rehearsed target and next-ball plan.
What should I serve at 30–40 on a second serve?
Choose a target you can execute and prepare an intended plus one. A body serve is one practice example, not a universally superior choice. Read the return before committing.
Is being clutch a real skill in tennis?
These calculations do not measure clutch ability. Player summaries need context, sample sizes and a suitable baseline before making a claim about persistent pressure performance.
What this doesn't cover
The scoring model assumes independent points and fixed service win rates. Full-match rates use later results, so the private clutch sample is retrospective. A ten-point process score records the agreed reset and plan check; it does not measure shot quality, pressure tolerance or a stable clutch trait.