We tested three popular style-matchup theories against no-vig closing prices on the 2026 season. Left-handers win 46.3% against right-handers versus 46.8% implied by the odds, and across 802 players with 20 or more priced matches we found 26 significant mispricing outliers where chance alone would produce about 40. Playing styles are already in the price.
Style-matchup claims are everywhere: Tatjana Maria's slice troubles power players, Mannarino's flat lefty game gives some opponents fits, Tsitsipas' one-handed backhand suffers under heavy topspin. For betting, the question is whether the market already knows. We tested against no-vig closing prices for the 2026 season, roughly 20,400 priced matches.
We labeled handedness for 182 of the most active players on the odds panel, only where it is publicly documented, and took every 2026 match between a labeled lefty and a labeled righty: 201 matches.
| Lefty vs righty (n=201) | Win rate |
|---|---|
| What lefties actually won | 46.3% |
| What their odds implied | 46.8% |
Half a point, well inside noise. The sports-science literature points the same way: left-handers are heavily overrepresented in fast-reaction sports like table tennis, but in tennis the elite advantage is moderate, has been shrinking among male professionals for decades, and is nearly absent in the women's game [1][2].
If the market misses specific players' styles, their season results should beat their market-implied win totals more often than chance allows. For every player with at least 20 priced matches in 2026 (802 players) we computed actual minus implied wins as a z-score. At the 5% level, chance alone would flag about 40 players. We found 26, fewer than chance. The top of the "most underrated" list was Iga Swiatek: a cold streak over 21 matches, not a style effect.
| Player (2026) | n | Actual win% | Implied | Gap |
|---|---|---|---|---|
| T. Maria (the slice) | 36 | 55.6% | 47.7% | +7.8pp |
| A. Mannarino (the lefty flat game) | 26 | 30.8% | 39.7% | -9.0pp |
| S. Tsitsipas (the backhand) | 26 | 50.0% | 53.9% | -3.9pp |
Maria beat her prices, in the direction the theory predicts, but on 36 matches at z = 1.1, a run dozens of players produce by luck every season. Mannarino ran nine points below his prices. Tsitsipas landed close to his.
The stronger form of the theory is pair-specific: certain opponents cannot play Maria. The raw records show the pattern. Maria is 4-0 against Jeanjean, 4-0 against Eala and 3-0 against Fruhvirtova; Mannarino is 3-0 against O'Connell, Safiullin, Coric and Broady. Only three 2026 meetings inside those pairs have recorded odds. The dominant side won all three against a combined 58% implied, which proves nothing at n=3. The larger-sample answer is our head-to-head study: leaders with a 3-0 or better record who were underdogs won 37.5% against 41.0% implied. We will rerun this when the odds panel is deeper.
Style matchups are real on court, and the market has absorbed them. Any remaining style edge would sit where the market has little data, such as a first meeting of extreme styles at a level nobody prices carefully, and would be rare and short-lived. Our model uses style as context, never as a reason to override a price.
Method: 20,398 finished 2026 matches with closing odds; implied probabilities devigged by proportional normalisation (both sides sum to 100%). Handedness labeled for 182 top-panel players from public documentation, uncertain players omitted, giving 201 clean lefty-vs-righty matches. Per-player residual z = (actual wins − implied wins) / sqrt(Σp(1−p)), min 20 matches. Case-box samples are single-season and directional only. Related: the H2H study and the fatigue study, the same pattern of a real effect already in the price.
References:
[1] F. Loffing, "Left-handedness and time pressure in elite interactive ball games," Biology Letters, vol. 13, no. 11, 20170446, 2017.
[2] F. Loffing, N. Hagemann, and B. Strauss, "Left-handedness in professional and amateur tennis," PLOS ONE, vol. 7, no. 11, e49325, 2012.
No. On 201 matches between a labeled left-hander and right-hander in 2026, lefties won 46.3% against 46.8% implied by the no-vig closing odds. Whatever advantage lefties have on court, the market already charges you for it.
Fewer than chance would produce. Across 802 players with at least 20 priced matches in 2026, 26 showed statistically significant gaps between actual wins and market-implied wins, while random variation alone would produce about 40. We found no detectable set of systematically mispriced players.
Maria outperformed her 2026 prices by about 8 percentage points, which points in the direction of the theory but is within normal variance at 36 matches (z = 1.1). Mannarino, the popular lefty case, ran 9 points BELOW his prices. Neither is statistical evidence; both show how easily a style story fits whatever the sample happens to say.
See today's picks: published before the match, graded in public →