TennisEdge's AI model finds no betting value in P. Kudermetova vs M. Yamaguchi (Washington on 2026-07-25) — the market is efficient, with P. Kudermetova favoured at 1.28.
P. KUDERMETOVA vs M. YAMAGUCHI — WTA Washington, Semi-finals (hard)
MODELS vs MARKET
ELO: 67.4% A | ML: 59.3% A | Blend: 65.8% A
Book: 1.28 / 3.40 (norm. implied A 72.6%)
Betfair (executable): 1.24 / 4.50 -> A 80.6% raw; B effective 4.325 (23.1% after 5% commission)
Raw numerical gap: model B 34.2% vs exchange B 23.1% = ~11% dog edge on paper.
DATA-STALENESS FLAG (decisive)
This is a SEMI-FINAL, yet neither player has a single Washington match in the fetched
history (A last recorded 07-13, B 07-15; fatigue feed agrees: 1 match/14d each).
Both players have won 2-3 matches at this event that the models never ingested.
The exchange (1.24) is pricing that live-event information; the models are not.
The "model edge" for the dog is therefore an artifact of a stale input set.
RECENT FORM (pre-event, most recent 5)
Kudermetova (23, rank 123):
07-13 Athens Q L vs Sakkari (0-6, TB set) | 07-05 Contrexeville SF L vs Boulard (ELO 1174!)
06-30 Wimbledon R64 L vs Samsonova (1759) | 06-25 Wimbledon (quali) F W vs Charaeva (1645)
06-24 Wimbledon (quali) SF W vs Ngounoue — DOM 6-1 6-2
Pre-event read: 3 straight losses incl. one to a 1174-ELO player — poor. TRANSITION:
last surfaces clay/unknown -> hard. TITLE: Wimbledon quali-final win 30d ago (grass).
But she has evidently rediscovered form this week (SF run unseen by models).
Yamaguchi (27, rank 293):
07-15 W75 Granby R32 L vs Marino (1610) | 07-09 Newport R32 L vs Minnen (1704)
07-03 W100 Cary R16 L vs Koike (1477) | 06-30 W100 Cary W vs Osuigwe | 06-12 W100 Wuning QF L
Pre-event read: 3 losses in last 4 at ITF level, no wins vs Strong/Elite in window.
Also carrying unseen wins this week.
FATIGUE
A: 1 match/14d, 72 court-min | B: 1 match/14d, 95 min (both stale — excludes this event).
Diff 23 min, matches diff 0 -> roughly equal. No age modulation needed (23 / 27).
NARRATIVE
Let-down: none in 7d window (A's last Final win 30d ago). Venue momentum: no prior
Washington matches on record for either. Surface transition: A clay->hard (risk on A,
mildly supports dog ML — but data staleness dominates).
ELO vs RANKING
A rank 123 with 67% ELO edge over rank 293 — consistent, ELO credible in direction.
Exchange is simply far more confident than the models, with better (live) information.
GATES
G1 edge-origin: PASS — model B 34.2% vs exchange 23.1% (~11%) exists pre-narrative.
G2 consensus: FAIL — market (80.6% A) AND ML (59.3% A) both against a Yamaguchi
ticket; no H2H, and the live this-event evidence (both in SF) is
precisely the data the models lack — nothing hard supports the dog.
G3 fatal-risk: FAIL — models missing BOTH players' current-event matches; the 1.24
exchange price embeds Kudermetova's unseen SF run. If she has been
winning comfortably this week (likely at that price), the entire
stale-form fade thesis is negated.
VERDICT: PASS
The ~11% "edge" is a stale-model artifact, not value. Backing Kudermetova at 1.24
offers no edge either (model says 65.8% vs 80.6% needed). No alt markets priced
(handicap/totals null). No bet.