TennisEdge's AI model finds no betting value in E. Butvilas vs A. Bublik (Geneva on 2026-05-20) — the market is efficient, with A. Bublik favoured at 1.33.
Independent estimates of who wins, each from a different method. The blend is what the bot prices against the market.
| Method | E. Butvilas | A. Bublik |
|---|---|---|
| Elo ratings-based, surface-weighted |
45.7% | 54.3% |
| Machine learning 29-feature gradient-boosted model |
35.9% | 64.1% |
| Serve / return point-by-point simulation |
41.4% | 58.6% |
| Blend what the bot actually prices off |
41.0% | 59.0% |
| Market implied de-vigged from the opening odds |
28.1% | 71.9% |
We publish how well these score against the closing market, including where they lose: how accurate are tennis predictions?
| Player | Surface | Elo | Record | Matches |
|---|---|---|---|---|
| E. Butvilas | hard | 1578 | 81–57 | 138 |
| A. Bublik | hard | 1879 | 111–93 | 204 |
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 July 25, 2026.
A. Bublik leads 1–0 over 1 meeting. Full record.
| Date | Event | Surface | Winner | Score |
|---|---|---|---|---|
| May 20, 2026 | Geneva | clay | A. Bublik | 2 - 0 |
| L | M. McDonald | 2 - 1 | July 25, 2026 |
| L | C. Tabur | 1 - 2 | July 18, 2026 |
| L | M. Huesler | 2 - 0 | July 11, 2026 |
| L | F. Ferreira Silva | 2 - 0 | June 30, 2026 |
| L | M. Echargui | 1 - 2 | June 22, 2026 |
| L | Q. Halys | 0 - 2 | July 25, 2026 |
| W | T. M. Etcheverry | 2 - 0 | July 24, 2026 |
| W | A. Molcan | 2 - 0 | July 23, 2026 |
| W | F. Diaz Acosta | 2 - 0 | July 22, 2026 |
| L | Q. Halys | 1 - 2 | July 17, 2026 |
All May 20, 2026 results & picks →
Math-only prediction: no LLM calls used