On 22,048 matches from the 2026 season, the betting market's favorite won 71.5% of the time, and the market's probabilities are honest at the low end while slightly underrating heavy favorites (a 90-95% favorite actually wins 96.4%); our own model picks winners less accurately than the closing market, which is normal, because nobody honest beats the close at prediction, and any tipster advertising 80% accuracy is choosing easy matches, not seeing the future.
Every beginner asks the same question: how accurate are tennis predictions, and what percentage should I trust? It sounds like it should have one number for an answer. It doesn't, and the reason it doesn't is the single most useful thing to understand before you believe anyone's accuracy claim, including ours.
Here is the same 2026 season, the same market, the same "prediction method" (just pick the favorite), sliced three ways.
| Only bet when the favorite is... | n | "Accuracy" |
|---|---|---|
| Priced under 1.30 | 8,549 | 85.3% |
| Priced 1.30 to 1.60 | 8,822 | 66.7% |
| Priced 1.60 to 2.00 | 4,677 | 55.2% |
Same sport, same season, same zero skill. You can be an "85% accurate" tennis predictor today by only picking heavy favorites, and you will still lose money doing it, because a 1.15 price already assumes you are right 87% of the time (our favorite-longshot study covers what those bets return: minus 3% and worse). So when anyone advertises 80%+ accuracy without showing the odds behind it, the number is compatible with pure menu choice, and you have no way to tell the difference. That is not an accusation, it is arithmetic. Independent tipster reviewers reach the same conclusion from the other direction, warning bettors to ignore anyone promising 80-90% win rates.
The fair version of the question is: how good are the best publicly available predictions, meaning the closing odds? We stripped the bookmaker margin out of every closing price (no-vig, both sides scaled to 100%) and checked every band of matches: when the market says 70%, does the favorite win 70% of the time?
| Market says | Matches | Actually won |
|---|---|---|
| 50-55% | 2,335 | 53.2% |
| 55-60% | 3,498 | 58.7% |
| 60-65% | 3,435 | 63.3% |
| 65-70% | 2,601 | 69.1% |
| 70-75% | 2,557 | 73.2% |
| 75-80% | 2,918 | 80.8% |
| 80-85% | 1,598 | 84.9% |
| 85-90% | 2,056 | 91.9% |
| 90-95% | 950 | 96.4% |
Two things worth staring at.
First, the market is startlingly well calibrated where matches are genuinely uncertain. When it says 55-65%, reality lands within a point or two. That is what an efficient prediction machine looks like, and it is the machine you are betting against. (We found this the annoying way. An early version of the script counted the exact even-odds matches as favorite wins, and the coin-flip band came out about 11 points too good before we caught it and threw those pairs out. There is no favorite in a 50/50, so counting one is just scoring a coin toss as a hit.)
Second, the drift at the top is real and one-directional: heavy favorites win MORE than their implied probability, by 4-5 points in the top bands. This is the famous favorite-longshot bias seen from the favorite's side. The margin gets loaded onto underdogs, so favorite prices sit a touch too long. Before you get excited: the vig eats that gap at the prices you can actually take, which is exactly what our longshot study measured (even heavy favorites returned minus 3% flat-staked). The bias is real. The free money is not.
Overall the market's favorite wins 71.5% of matches, with a Brier score of 0.187 (lower is better; a coin-flipper scores 0.25).
Same test, on the 4,477 matches our forward pipeline has predicted and settled this season.
| Predictor | Winner-pick accuracy | Brier score |
|---|---|---|
| Devigged closing market | 67.6% | 0.204 |
| Our model blend | 62.8% | 0.225 |
The market wins. It should: the close is the aggregated opinion of everyone with money and information, minutes before play, while our model prices hours earlier from historical data. (Note the market's accuracy is lower on this subset than the 71.5% above because our slate skews toward closer, harder matches. Same yardstick, harder exam. Also worth saying plainly: some of our model inputs are partially anchored to market prices, so this is not a fully independent horse race, which is one more reason we do not claim it as one.)
If a prediction site tells you their model out-predicts closing odds over thousands of matches, ask for the graded record. We publish ours, including the part where the devigged number got worse: the devig-CLV study.
Method: 22,838 finished 2026 matches with closing odds; 22,048 after excluding exact even-odds pairs (no favorite exists) and a handful of unpriceable quotes. Implied probabilities devigged by proportional normalisation. Calibration is the devigged favorite probability, bucketed, vs actual favorite win rate. Model comparison on the 4,477 settled forward predictions with stored model probabilities, market Brier computed on the identical matches. Related: favorite-longshot, devig-CLV, head-to-head and fatigue studies.
Depends entirely on the matches. Picking heavy favorites gives 85%+ with zero skill and negative profit. On genuinely uncertain matches (both players priced 40-60%), the market itself manages about 53-63%, so anything sustainably above that range on those matches would be remarkable. Judge predictions against the odds, never against a bare percentage.
Not at the closing line, on the evidence so far. Our own model picks winners at 62.8% where the devigged close manages 67.6% on the same matches. The realistic goal for a model is beating a subset of prices early, before the market corrects, which is measured by closing-line value, not accuracy.
Easily, and that is the problem. Our data shows picking short favorites yields 85%+ accuracy with zero skill and negative profit, because a 1.15 price already implies 87% wins. High accuracy is cheap to produce honestly, so the number proves nothing by itself. Independent tipster reviewers warn bettors to ignore anyone promising 80-90% win rates and put realistic tennis strike rates at 40-60%, which matches what we measure for the market itself on competitive matches. If a claim comes without the odds and a full graded history attached, you cannot tell skill from a menu choice.
See today's picks — published before the match, graded in public →