TennisEdge's AI model finds no betting value in L. Samsonova vs M. Keys (Washington on 2026-07-27) — the market is efficient, with M. Keys favoured at 1.27.
L. Samsonova (A, #21, age 27) vs M. Keys (B, #7, age 31) — Washington, hard, 1/16-finals
MODELS
ELO Keys 80.6% | ML Keys 70.8% | Sim Keys 71.8%
Market (Betfair) Keys 76.3% (odds A 4.0 / B 1.31); book 3.75 / 1.27
Market sits BETWEEN ML (70.8) and ELO (80.6) — models bracket the price = efficient.
Sim win prob (Keys 71.8) converges with ML → inputs trusted, no override.
ELO vs ranking: both sides ELO agrees with rank (Keys #7 fav, Samsonova #21 dog) — ELO credible.
RECENT FORM (grass window; both now transitioning to hard)
Samsonova (last 5):
07-04 Wim L Bouzkova(1811) 6-4 6-7 4-6 — narrow 3-set loss, Strong
07-02 Wim W Shnaider(1672) 3-set GRIND, Strong
06-30 Wim W Kudermetova(1679) 6-3 6-3 DOM, Strong
06-23 BadHomburg L Svitolina(1816) 3-set loss, Strong
06-21 BadHomburg W Siniakova(1774) 6-3 7-5 DOM, Strong
Shape STABLE: quality DOM wins (Kudermetova, Siniakova) offset by narrow losses to Strong; no trend.
Keys (last 5):
07-06 Wim L Noskova(1874) 4-6 6-7(TB) — loss to Elite
07-04 Wim W Anisimova(1887) 3-set GRIND over ELITE — top result
07-02 Wim W Swan(1637) straight DOM, Avg
06-30 Wim W Day(1549) 7-6 3-set GRIND, Weak — one under-performing spot
06-27 Eastbourne FINAL W Maria(1824) straight DOM — TITLE
Shape STABLE-to-strong: Eastbourne title + beat Elite Anisimova; class edge intact.
PATH / RUST
Neither has a Washington match yet (both first round). Both last played Wimbledon 07-04/07-06 —
~3 weeks off → rust risk, but SYMMETRIC. No advantage either way.
FATIGUE
Both well-rested (matches_14d = 0). No factor. No age modulation needed.
NARRATIVE
Let-down: Keys won Eastbourne Final 06-27 = 30d ago → outside 7d window, None.
Title recency: Keys TITLE Eastbourne 06-27 (grass), 30d ago — live form strong.
Surface transition: BOTH grass -> hard (symmetric); both historically solid hard-courters. Directionally neutral.
MARKET SCAN
ML A (Samsonova) @4.0: eff 3.85, implied 25.97%; sim 28.2% / ml 29.2% -> edge ~2-3% (below 5%), ELO opposes. No edge.
ML B (Keys) @1.31: eff 1.2945, implied 77.25%; sim 71.8% -> NEGATIVE edge. No value.
Handicap Samsonova +4.5 @1.94: sim cover 59.8%, eff 1.893 implied 52.8% -> edge ~7.0%.
BUT |line|=4.5 = HIGH-VARIANCE tier -> requires edge >=10% (min 1.84). 7% < 10% -> FAILS bar.
Total: Betfair line 21.0. Sim under21 ~54.3% vs eff-implied 51.5% -> edge ~2.8% (below 5%). Over negative. No edge.
GATES (best candidate = Samsonova +4.5 handicap)
G1 edge-origin: FAIL — 7.0% sim cover edge is real & numeric (no narrative), but on a wide |line|=4.5
the required bar is 10% (wide-AH); 7% falls short. ML edge only ~2-3%, also sub-bar.
G2 consensus: PASS — winner consensus (Keys) not contradicted; dog-cover thesis from same sim. (moot)
G3 fatal-risk: FAIL — Keys is clear higher class, just beat Elite Anisimova & won a grass title;
she has real blow-out upside, directly negating a tight-margin cover thesis on the +4.5.
VERDICT — PASS. Market brackets between ML and ELO on the moneyline = efficient, no ML edge either side.
The only positive-EV line (Samsonova +4.5, ~7%) is a high-variance wide handicap that misses the 10% bar
and carries a live blow-out risk from a higher-class, in-form favourite. Totals edge too thin. No bet.