TennisEdge's AI model finds no betting value in A. Fery vs A. Zverev (Wimbledon on 2026-07-10) — the market is efficient, with A. Zverev favoured at 1.12.
Independent estimates of who wins, each from a different method. The blend is what the bot prices against the market.
| Method | A. Fery | A. Zverev |
|---|---|---|
| Elo ratings-based, surface-weighted |
29.2% | 70.8% |
| Machine learning 29-feature gradient-boosted model |
29.2% | 70.8% |
| Serve / return point-by-point simulation |
51.5% | 48.5% |
| Blend what the bot actually prices off |
30.6% | 69.4% |
| Market implied de-vigged from the opening odds |
15.7% | 84.3% |
We publish how well these score against the closing market, including where they lose: how accurate are tennis predictions?
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 19, 2026.
A. Zverev leads 1–0 over 1 meeting. Full record.
| Date | Event | Surface | Winner | Score |
|---|---|---|---|---|
| July 10, 2026 | Wimbledon | grass | A. Zverev | 3 - 0 |
| L | A. Zverev | 0 - 3 | July 10, 2026 |
| W | F. Cobolli | 0 - 3 | July 8, 2026 |
| W | G. Dimitrov | 2 - 3 | July 6, 2026 |
| W | Z. Bergs | 2 - 3 | July 4, 2026 |
| W | O. Virtanen | 3 - 1 | July 2, 2026 |
All July 10, 2026 results & picks →
═══ A. Fery vs A. Zverev — Wimbledon SF (grass) ═══
MODELS vs MARKET
Odds: Fery 6.00 (imp 15.7%) | Zverev 1.12 (imp 89.3%) [Betfair ML: 7.0 / 1.16]
ELO: Fery 29.2% | ML: 29.2% | Blend: 30.6% → value flag on Fery (+14.9 pts)
All three models disagree with market on Fery — but see qualitative read below.
RECENT FORM (last 5 each)
Fery (5-0 Wimbledon run, ALL GRIND):
07-08 W Cobolli (1772, Strong) QF — GRIND (TB set, 3-setter after two 5-set wars)
07-06 W Dimitrov (1834, Strong) R16 — GRIND (5 sets, deciding TB)
07-04 W Bergs (1770, Strong) R32 — GRIND (5 sets, two TBs)
07-02 W Virtanen (1839, Strong) R64 — GRIND (4 sets, TB)
06-30 W Dzumhur (1440, Weak) R128 — 4 sets
Zverev (5-0 Wimbledon run):
07-08 W Fritz (1999, ELITE) QF — DOM (straight sets, no set past 6-4)
07-07 W Lehecka (1844, Strong) R16 — GRIND (4 sets, TB)
07-04 W Giron (1670, Strong) R32 — GRIND (straights w/ TB)
07-02 W Royer (1661, Strong) R64 — GRIND (straights w/ TB)
06-30 W Blockx (1589, Avg) R128 — GRIND (4 sets, TBs)
FATIGUE (14d, precomputed)
Fery: 5 matches, 15 sets, 858 court min (171/match!), last 07-08
Zverev: 5 matches, 15 sets, 740 court min (148/match), b2b 1, last 07-08
→ 118-min court-time gap: Fery clearly heavier. Fatigue edge ZVEREV.
NARRATIVE
Let-down: none in window (Zverev won RG Final 33d ago — TITLE: French Open 33d ago, clay; settled, no letdown).
Venue: both 5-0 at this Wimbledon.
Fery ELO/model prob is inflated BY the run itself — upset wins pump live ELO while the
market prices class + accumulated fatigue. Cinderella at 171 min/match vs a top seed
who just straight-setted Fritz is exactly the spot where the model overrates the dog.
ALT MARKETS
Handicap: Fery +6.5 @1.81 / Zverev -6.5 @2.20 — no model cover anchor (null); no
independent margin conviction: grass TB variance argues +6.5 lives, fatigue blowout
argues it dies. No edge I can quantify.
Total 35.0: O 1.96 / U 1.98. Model over 54.3% at 34.5 → at 35.0 that shaves to ~51-52%
vs 51.0% implied. <2 pt edge. PASS.
VERDICT: PASS (all markets)
Model-market disagreement exists but the qualitative read invalidates it rather than
confirming it: DOM-over-Elite vs all-GRIND run + worst fatigue profile on the slate.
Book 6.0 on Fery is also worse than exchange 7.0 — the "value" side is shaded.