TennisEdge's AI model finds no betting value in L. Noskova vs E. Mertens (Wimbledon on 2026-07-08) — the market is efficient, with L. Noskova favoured at 1.60.
Independent estimates of who wins, each from a different method. The blend is what the bot prices against the market.
| Method | L. Noskova | E. Mertens |
|---|---|---|
| Elo ratings-based, surface-weighted |
60.5% | 39.5% |
| Machine learning 29-feature gradient-boosted model |
48.3% | 51.7% |
| Serve / return point-by-point simulation |
62.9% | 37.1% |
| Blend what the bot actually prices off |
60.5% | 39.5% |
| Market implied de-vigged from the opening odds |
57.9% | 42.1% |
We publish how well these score against the closing market, including where they lose: how accurate are tennis predictions?
| Player | Surface | Elo | Record | Matches |
|---|---|---|---|---|
| L. Noskova | hard | 1892 | 107–53 | 160 |
| E. Mertens | hard | 1855 | 118–77 | 195 |
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 20, 2026.
L. Noskova leads 1–0 over 1 meeting. Full record.
| Date | Event | Surface | Winner | Score |
|---|---|---|---|---|
| July 8, 2026 | Wimbledon | grass | L. Noskova | 2 - 0 |
| W | K. Muchova | 1 - 2 | July 11, 2026 |
| W | M. Kostyuk | 0 - 2 | July 9, 2026 |
| W | E. Mertens | 2 - 0 | July 8, 2026 |
| W | M. Keys | 0 - 2 | July 6, 2026 |
| W | S. Cirstea | 1 - 2 | July 4, 2026 |
| L | L. Noskova | 2 - 0 | July 8, 2026 |
| W | M. Bouzkova | 0 - 2 | July 6, 2026 |
| W | E. Rybakina | 2 - 0 | July 4, 2026 |
| W | M. Timofeeva | 2 - 1 | July 2, 2026 |
| W | L. Siegemund | 2 - 0 | June 30, 2026 |
All July 8, 2026 results & picks →
═══ L. Noskova vs E. Mertens — Wimbledon QF (grass) ═══
MODELS & MARKET
Book: Noskova 1.60 / Mertens 2.20 (implied A ~57.9%) | Betfair ML 1.67/2.46 (A ~59.9%)
ELO: 60.5% A | ML model: 48.3% A | Blend: 60.5% A | value_pick A, edge +2.6%
Alignment: ELO agrees with market; ML model mildly leans Mertens. Split, not consensus disagreement.
RECENT FORM (last 5 shown; all fetched matches used)
Noskova (9-1 on grass this swing):
07-06 W Keys (1930, Elite) 6-4 7-6 — GRIND, statement win
07-04 W Cirstea (1576) in 3 w/ TB — GRIND
07-02 W Osorio (1578) in 3 — competitive
06-30 W Seidel (1549) 6-4 6-3 — DOM vs Weak
06-23 L Ruse (1761) Bad Homburg
TITLE: Berlin 17d ago (grass) — beat Pegula (1897) in Final, Eala (1911) DOM in SF. Let-down window long expired.
Mertens (4-0 at Wimbledon, 3-3 in warmups):
07-06 W Bouzkova (1838, Strong) 6-4 6-4 — DOM
07-04 W Rybakina (1838, Strong) 7-6 6-1 — GRIND/TB but emphatic
07-02 W Timofeeva (1563) in 3
06-30 W Siegemund (1634) 6-2 6-4 — DOM
06-23 L Osaka Bad Homburg (also lost Berlin R16, Hertogenbosch R16 in warmups)
FATIGUE
Noskova: 5 matches/14d, 15 sets, ~369min, last 07-06, no B2B
Mertens: 6 matches/14d, 18 sets, ~376min, last 07-06, 1 B2B
Court-time diff ~7min, matches diff 1 → Roughly equal. No fatigue edge.
NARRATIVE
Let-down: Berlin Final won 17d ago → None. Venue: both 4-0 at this Wimbledon. No differentiator.
MARKET SCAN (alt_markets live)
ML: my read ~60-62% Noskova vs 57.9-59.9% implied → edge ~2-3%, and ML model (48.3%) actively cuts against it. Thin.
Handicap A -2.5 @ 2.02: sim covers 54.6% vs 49.5% implied (+5.1%). But Noskova's own wins here barely/don't cover (+3 vs Keys, net +2 vs Cirstea) and grass tightens margins → shave to ~51%, edge collapses.
Total U22.5 @ 1.99: sim under 58.7% vs 50.3% implied (+8.5%). Qualitative read contradicts: both players hitting TBs/3-setters here (5 of 8 combined Wimbledon matches went over 22.5); one TB in a straight-set match (6-4 7-6 = 23) already busts the under. My independent under estimate ~45-48% — I don't trust the sim's number.
VERDICT: PASS
ELO and market agree on Noskova; the only model disagreements (ML slight Mertens lean, sim under/handicap edges) fail qualitative validation. Both players in strong form, fatigue equal, no narrative edge. No market clears the bar.