Try the demo

Free tool

Tennis win-chance calculator

Updated on Free, no sign-up

Enter the current score, who's serving, and roughly how often each player holds serve, and the calculator shows each player's estimated chance of winning the match. It runs the same point-by-point model I use in the Learn articles, in your browser, with no sign-up. Compare the result with the market price to see where you and the market disagree.

Quick presets (serve hold for both players)

How often A wins their own service games.

About 80% for ATP hard courts, 65% for WTA.

Match length

Enter 6 and 6 for a tiebreak.

Who serves this game?

Player A wins the match50%
Player B wins the match50%
As a price: A50¢
As a price: B50¢

A model, not a prediction. It assumes each player keeps winning the same share of points on serve, and knows nothing about form, injuries or nerves.

What the model is, and what it isn't

This is a model, not a prediction. It assumes each player wins a fixed share of the points on their own serve, and that every point is independent of the last. From those two numbers it works out the chance of winning each game, each set and the match from any score. It doesn't know about injuries, nerves, momentum, weather, or who played five hours yesterday. Real matches have all of those, and real prices include them. Treat the output as a baseline: what the odds would be if nothing else mattered.

How to fill it in

  • Serve hold %. How often each player wins their own service games. On the men's tour it's about 80% on hard courts, higher on grass and lower on clay. On the women's tour it's about 65%. The presets set sensible defaults; adjust them if you know the players.
  • Sets and games. The score so far. Use the best-of-five setting for men's matches at the four majors.
  • Points. The score in the current game, as 0, 15, 30, 40 or AD. In a tiebreak at 6-6, enter the tiebreak points instead.
  • Who's serving. The server of the current game, or of the next point in a tiebreak.

How to read the result

The chance is shown as a percentage and as a price in cents, because on a prediction market they're the same thing: a 62% chance is a 62¢ price. If the market has Player A at 70¢ and the model says 62%, either the market knows something the model doesn't, such as an injury or a big gap in form, or the market is ahead of itself. Usually it's the first. Sometimes it's the second, and that's where trades come from. How tennis prices work explains the link between price and probability.

Worked example: a break in the first set

Two evenly matched men on a hard court, both holding about 81% of the time. At 2-2 in the first set with Player B serving, the model gives Player A about 53%. Player A breaks to lead 3-2 and will serve next. Now the model gives A about 68%. That one game was worth about 15¢ on the market.

If you'd bought 50 shares of A at 53¢ for $26.50 before the break, they'd be worth about $34 after it, a gain of about $7.50 before fees. If B had held instead, A would be at about 49¢ and the shares worth about $24.50.

Now put the same score in the deciding set. The break takes A from about 50% to about 81%. Same game, twice the move. That's the single most useful thing the calculator shows.

Limits

  • Serve-hold rates are averages. A player holding 80% for the season might be holding 60% today with a sore shoulder.
  • The model assumes points are independent. Real players get tight, tired or inspired.
  • It ignores retirements, walkovers and the 50-50 rules that can settle a market without a winner. See what happens if a player retires.
  • Prices on a real market also include a spread and, on Polymarket, a fee. The profit calculator covers those.

Use it to build intuition. After a few matches you'll know roughly what a break, a set and a tiebreak point are worth without looking, and you can read more about why in why the price jumps on a break of serve.

FAQ

Is this a prediction of who will win?

No. It's a model that turns a score and two serve-hold rates into a probability, assuming nothing else matters. Real matches include injuries, form and nerves, which the model can't see.

What serve-hold percentage should I use?

About 80% for men on hard courts, a little higher on grass and lower on clay. About 65% for women. If you know a player serves unusually well or badly, adjust it.

Why doesn't the result match the market price?

The market includes things the model doesn't: injuries, form, head-to-head, and how the players look today. A gap between the two is a question to ask, not a trade on its own.

Does it work for tiebreaks and best-of-five matches?

Yes. At 6-6 in a set, enter the tiebreak points. Choose best of five for men's matches at the four majors.

Demo only. 77 Tennis uses demo money with no cash value. Nothing here is financial or betting advice. Prices come from Polymarket's public tennis markets, and I'm not affiliated with Polymarket. Prediction markets aren't legal everywhere, so check the rules where you live. If trading stops being fun, stop.

Keep reading