TennisEdge's AI model finds no betting value in D. Kasatkina vs M. Bolkvadze (Washington on 2026-07-25) — the market is efficient, with D. Kasatkina favoured at 1.25.
D. KASATKINA vs M. BOLKVADZE — WTA Washington, Semi-finals (hard)
MODELS vs MARKET
ELO: 74.3% A | ML: 65.8% A | Blend: 66.8% A
Book: 1.25 / 3.75 (75.0% A) | Betfair: 1.20 / 5.70 (83.3% A / 17.5% B)
Only candidate: Bolkvadze ML @5.70 (comm-adj 5.465, needs >18.3%); blend gives her ~33.2% -> nominal ~15% edge.
DATA STALENESS (decisive)
Both players are in a Washington SEMI-final, yet history ends 07-02/07-03 and precomputed fatigue shows 0 matches in 14d for both. This week's Washington matches (2-3 each, incl. whatever run put world #503 Bolkvadze into a WTA SF) are entirely missing from model inputs. The exchange (1.20/5.70) has priced the live week; our models have not seen it. Model-market "disagreement" here is stale-input artifact, not edge.
RECENT FORM (last 5 visible each)
Kasatkina (29, rank 66):
07-03 Wimbledon R32 L vs Osaka (1790) 1-6 3-6
07-01 Wimbledon R64 W vs Tjen (1537) — GRIND (TB)
06-29 Wimbledon R128 W vs Xu (1500) — 3 sets
06-22 Eastbourne L vs Kessler (1619)
06-13 Nottingham SF L vs Blinkova (1635)
Older: Strasbourg title run in May (beat Samsonova 1801, Stearns 1772). Grass swing poor; back on hard this week and clearly winning (SF).
Bolkvadze (28, rank 503):
07-02 Wimbledon R32 L vs Krueger (1847) 1-6 0-6 blowout
06-30 Wimbledon R64 W vs Tomljanovic (1728) — 3 sets
06-25 Wimbledon quali final W vs Vandromme (1558)
06-24 Wimbledon quali SF W vs Kinoshita (1638) DOM
06-10 W50 Guimaraes L vs Jorge (1613)
Older: May losses to sub-1500 ITF players (Adeshina 1386, Appleton 1453, Hibino 1451).
FATIGUE
Precomputed fatigue is 0/0 for both — invalid given both played this week in Washington. No usable fatigue read; age-neutral, no adjustment made.
ELO/RANKING SANITY
Bolkvadze: ELO implies ~26-33% vs a former top-10, yet ranked 503 with recent blowout losses to sub-1500 opposition — inflated/stale rating (farmed at quali/ITF level). The numeric edge on her side rests on an ELO that fails the sanity check. Discounted.
Kasatkina: ELO-fav at rank 66 — plausible for a former top-10; credible.
NARRATIVE
Let-down: none actionable (Bolkvadze quali "Final" 06-25 is a month old). Venue: no prior Washington matches in window. Surface transition: both transitioned grass->hard weeks ago and have played this week on hard (unseen) — neutral.
GATES
G1 edge-origin: FAIL — the ~15% numeric edge on Bolkvadze comes from models fed stale inputs (missing the entire Washington week) and an ELO that fails the ranking sanity check; stripped of the untrustworthy rating, no bar-clearing edge exists.
G2 consensus: FAIL — exchange 83% A sits hard against a B pick and the live this-event evidence (both mid-tournament, exchange has seen this week's matches) is on the market's side; no on-surface H2H or live evidence for B.
G3 fatal-risk: FAIL — Bolkvadze's visible record (1-6 0-6 blowout, losses to 1386/1451/1453 ELO ITF players) directly negates a 33% win claim vs a former top-10; if the ELO is farmed, the thesis is fiction.
VERDICT: PASS. The only numeric edge (Bolkvadze ML) is a stale-data artifact on a non-credible rating, with the exchange holding fresher information than the models. Backing Kasatkina at 1.20 is negative-edge vs the 67% blend. No stakeable market (alt handicap/totals not priced).