TennisEdge's AI model finds no betting value in L. van Assche vs T. Droguet (Montreal on 2026-08-02) — the market is efficient, with L. van Assche favoured at 1.80.
Independent estimates of who wins, each from a different method. The blend is what the bot prices against the market.
| Method | L. van Assche | T. Droguet |
|---|---|---|
| Elo ratings-based, surface-weighted |
57.7% | 42.3% |
| Machine learning 29-feature gradient-boosted model |
60.9% | 39.1% |
| Serve / return point-by-point simulation |
46.8% | 53.2% |
| Blend what the bot actually prices off |
54.3% | 45.7% |
| Market implied de-vigged from the opening odds |
52.6% | 47.4% |
We publish how well these score against the closing market, including where they lose: how accurate are tennis predictions?
| Player | Surface | Elo | Record | Matches |
|---|---|---|---|---|
| L. van Assche | hard | 1762 | 101–85 | 186 |
| T. Droguet | clay | 1787 | 133–65 | 198 |
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 2–2 over 4 meetings. Full record.
| Date | Event | Surface | Winner | Score |
|---|---|---|---|---|
| Aug. 3, 2026 | Montreal | hard | T. Droguet | 2 - 1 |
| July 15, 2026 | Umag | clay | T. Droguet | 2 - 1 |
| Jan. 30, 2026 | Quimper | hard | L. van Assche | 2 - 0 |
| Oct. 9, 2024 | Roanne Challenger Men | hard | L. van Assche | 2 - 0 |
| L | T. Droguet | 1 - 2 | Aug. 3, 2026 |
| W | A. Blockx | 2 - 1 | July 26, 2026 |
| W | H. Gaston | 2 - 0 | July 25, 2026 |
| W | A. Rublev | 1 - 2 | July 24, 2026 |
| W | P. Carreno-Busta | 2 - 1 | July 23, 2026 |
| W | L. van Assche | 1 - 2 | Aug. 3, 2026 |
| W | D. Svrcina | 1 - 2 | Aug. 1, 2026 |
| L | H. Gaston | 2 - 0 | July 22, 2026 |
| W | C. Ugo Carabelli | 2 - 1 | July 20, 2026 |
| L | D. Merida Aguilar | 2 - 0 | July 16, 2026 |
All Aug. 2, 2026 results & picks →
MATCH: L. van Assche (A, 22, #97) vs T. Droguet (B, 25, #109) — Montreal, HARD, R1 (1/64) MODELS vs MARKET ELO A 57.7% | ML A 60.9% | blended 54.3% Betfair ML: A 2.02 (comm-adj 50.8%) / B 1.97. Pinnacle no-vig A 51.1%. Models favor A by 4-10pp; both sharp markets (Betfair + Pinnacle) sit ~50/50. RECENT FORM (last 5) van Assche (A): 07-26 WON Estoril FINAL vs Blockx(1890) 6-4 4-6 7-5 (GRIND, title); 07-25 WON SF vs Gaston(1696); 07-24 WON QF vs Rublev(1959); 07-23 WON vs Carreno-Busta(1837); 07-20 WON vs Ferreira Silva(1505). ALL CLAY. Clay specialist, hot week. Droguet (B): 08-01 WON Montreal "Final" vs Svrcina(1735) 3 sets (HARD, yesterday); 07-22 LOST Estoril vs Gaston; 07-20 WON vs Ugo Carabelli(1809); 07-16 LOST Umag; 07-15 WON vs van Assche(1710) [H2H, clay]. PATH & SHAPE (Step 1b) van Assche: SHARPENING on CLAY (won a title) but ZERO hard-court reps in window — pure clay→hard TRANSITION. Last match 7d ago. His ELO/ML edge is almost certainly clay-inflated; models not surface-adjusting. Droguet: has 1 hard win (yesterday). Also off clay but with one hard rep. Beat van Assche on clay 2 weeks ago (07-15). FATIGUE (Step 2) Droguet won a 3-set match YESTERDAY (08-01) -> -4.5pp hangover haircut would apply IF backing B (age 25, baseline). van Assche rested (7d). Neither relevant since no side is staked. NARRATIVE van Assche TITLE: Estoril 7d ago (clay) — possible mild let-down, but on a DIFFERENT surface. TRANSITION clay->hard on A (the side models favor) = RISK, not support. ELO/ranking: both near rank; A's clay-hot ELO is the suspect input here. GATES G1 edge-origin: FAIL(effective) — numeric model edge on A exists (ELO/ML 57-61% vs mkt 51%) BUT is judged clay-ELO contamination on a player with zero hard reps; professional read is the models are WRONG, not the market. No independent edge on B. G2 consensus: stake-tier only — n/a (pass). G3 fatal-risk: n/a (no margin bet). MARKET CHOICE ML: A 2.02 / B 1.97 — no trustworthy edge (A edge contaminated, B has no model support). Handicap: A -0.5 covers 48.3% @2.0 (need 51.3%) no; B +0.5 covers 51.7% @1.9 (need 53.9%) no. Total 22.5: over 48.3% @2.02 (need 50.8%) no; under 51.7% @1.92 (need 53.4%) no. All markets clear PASS. alt_markets present. VERDICT: PASS. Model A-lean is clay-inflated on a clay specialist transitioning to hard with no reps; both sharp markets correctly price ~50/50. No edge on either side.