TennisEdge's AI model finds no betting value in R. Safiullin vs J. Fonseca (Wimbledon on 2026-07-03) — the market is efficient, with J. Fonseca favoured at 1.42.
Independent estimates of who wins, each from a different method. The blend is what the bot prices against the market.
| Method | R. Safiullin | J. Fonseca |
|---|---|---|
| Elo ratings-based, surface-weighted |
61.3% | 38.7% |
| Machine learning 29-feature gradient-boosted model |
21.2% | 78.8% |
| Serve / return point-by-point simulation |
67.2% | 32.8% |
| Blend what the bot actually prices off |
61.3% | 38.7% |
| Market implied de-vigged from the opening odds |
31.1% | 68.9% |
We publish how well these score against the closing market, including where they lose: how accurate are tennis predictions?
| Player | Surface | Elo | Record | Matches |
|---|---|---|---|---|
| R. Safiullin | hard | 1715 | 107–81 | 188 |
| J. Fonseca | hard | 1954 | 61–22 | 83 |
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 March 4, 2026.
R. Safiullin leads 1–0 over 1 meeting. Full record.
| Date | Event | Surface | Winner | Score |
|---|---|---|---|---|
| July 3, 2026 | Wimbledon | grass | R. Safiullin | 3 - 0 |
| L | N. Djokovic | 1 - 3 | July 5, 2026 |
| W | J. Fonseca | 3 - 0 | July 3, 2026 |
| W | B. Van De Zandschulp | 3 - 2 | July 1, 2026 |
| W | A. Rublev | 2 - 3 | June 29, 2026 |
| W | J. Kym | 3 - 2 | June 25, 2026 |
| W | C. Moutet | 3 - 1 | July 18, 2026 |
| W | T. Griekspoor | 3 - 1 | July 18, 2026 |
| L | N. Kyrgios | 0 - 3 | July 18, 2026 |
| L | R. Safiullin | 3 - 0 | July 3, 2026 |
| W | J. De Jong | 0 - 3 | July 1, 2026 |
All July 3, 2026 results & picks →
═══════════════════════════════════════════════
R. Safiullin (A) vs J. Fonseca (B)
Wimbledon · grass · 1/16-finals · 2026-07-03
═══════════════════════════════════════════════
ODDS A 3.14 (imp 31%) | B 1.42 (imp 69%)
MODELS ELO A 0.613 (value_pick A, edge +0.30)
ML A 0.212 (B 0.788)
MODEL A 0.613 | serve/ret A 0.672
→ violent model split; ELO is the outlier.
FORM (last 14d)
Safiullin (A): 6 matches, 18 sets — grinding.
07-01 W Van De Zandschulp (1585 Avg) 6-0 4-6 6-3 3-6 7-6 5s GRIND
06-29 W Rublev (1744 Strong) 5s GRIND (TBs)
06-25 W Final Kym (1651 Avg) 5s
06-24 W SF Coppejans (1666 Avg) 2-0
06-22 W QF McCabe (1634 Avg) 2-0
→ deep run but every recent match a war; ELO fattened
by a low-tier clay win streak (Oeiras/Mauthausen).
Fonseca (B): 2 matches, 6 sets — fresh.
07-01 W De Jong (1532 Weak) 6-1 7-5 6-4 DOM
06-29 W Bautista-Agut (1634 Avg) 7-6 6-4 6-3 GRIND (clean 3-0)
Season pedigree: beat Djokovic (1991) & Ruud (2032) at RG.
FATIGUE
A: 940 court-min / 6 matches / back-to-back — heavy.
B: 252 court-min / 2 matches — rested.
Diff > 60min → Safiullin clearly disadvantaged.
NARRATIVE
Let-down: A won a Final 8d ago (06-25) → outside window, None.
Venue: Fonseca 2-0 this Wimbledon draw, both in straight sets.
Caveat: grass is B's weakest surface — blown out 2-6 2-6 by
Hanfmann at Halle (06-16); young, can be erratic.
VERDICT — LEAN (Fonseca)
The +0.30 "value" on Safiullin is a mirage: ELO is inflated by
Challenger clay wins and ignores a brutal fatigue load. ML (0.79),
market (0.69), fresher legs, and Elite-level pedigree all point to
Fonseca. But at 1.42 the price already reflects it — no exploitable
edge — and his grass volatility adds variance. Directional read is
clear; a bet is not. Lean Fonseca, no stake.
═══════════════════════════════════════════════