TennisEdge's AI model finds no betting value in A. Muller vs B. Van De Zandschulp (Rome on 2026-05-06) — the market is efficient, with B. Van De Zandschulp favoured at 1.66.
Independent estimates of who wins, each from a different method. The blend is what the bot prices against the market.
| Method | A. Muller | B. Van De Zandschulp |
|---|---|---|
| Elo ratings-based, surface-weighted |
51.6% | 48.4% |
| Machine learning 29-feature gradient-boosted model |
54.2% | 45.8% |
| Serve / return point-by-point simulation |
33.9% | 66.1% |
| Blend what the bot actually prices off |
46.6% | 53.4% |
| Market implied de-vigged from the opening odds |
42.9% | 57.1% |
We publish how well these score against the closing market, including where they lose: how accurate are tennis predictions?
| Player | Surface | Elo | Record | Matches |
|---|---|---|---|---|
| A. Muller | clay | 1637 | 103–90 | 193 |
| B. Van De Zandschulp | hard | 1749 | 108–84 | 192 |
Our own ratings, computed from match results since 2018. Players with fewer than 5 rated matches are omitted rather than shown at a provisional starting value. Last updated Aug. 3, 2026.
Level at 1–1 over 2 meetings. Full record.
| Date | Event | Surface | Winner | Score |
|---|---|---|---|---|
| May 6, 2026 | Rome | clay | B. Van De Zandschulp | 2 - 0 |
| Feb. 22, 2023 | Doha | hard | A. Muller | 2 - 0 |
| L | G. Den Ouden | 2 - 0 | Aug. 3, 2026 |
| L | M. Poljicak | 1 - 2 | July 31, 2026 |
| W | M. Cecchinato | 2 - 1 | July 29, 2026 |
| W | S. Napolitano | 2 - 0 | July 27, 2026 |
| L | M. Navone | 0 - 2 | July 20, 2026 |
| L | J. Faria | 0 - 2 | July 21, 2026 |
| L | D. Vallejo | 1 - 2 | July 15, 2026 |
| W | T. Daniel | 2 - 0 | July 14, 2026 |
| L | R. Safiullin | 3 - 2 | July 1, 2026 |
| W | A. Kovacevic | 1 - 3 | June 29, 2026 |
All May 6, 2026 results & picks →
Math-only prediction: no LLM calls used