TennisEdge's AI model finds no betting value in H. Vandewinkel vs L. Tararudee (Istanbul 2 (Turkey) - Qualification on 2026-07-19) — the market is efficient, with L. Tararudee favoured at 1.61.
═══ H. Vandewinkel vs L. Tararudee — Istanbul 2 Qualification Final (hard) ═══ MODELS & MARKET Odds: A 2.20 (impl 42.3%) | B 1.61 (impl 62.1%) ELO: A 46.4% | ML: A 35.5% | Model: A 46.4% value_pick: A (+4.2% ELO vs market) — but ML says A only 35.5% (mild ~2.4% edge on B) Models point in OPPOSITE directions around the market → no consensus edge. RECENT FORM (last 5 shown; all fetched matches used) Vandewinkel (A): 4-0 this event 07-18 SF W vs Zakharova (1745, Strong) — DOM straight sets 07-17 QF W vs Tan (1500) — 6-1 1-0 (opp retirement) 07-15 R16 W vs Falkowska (1500) — 3-set win 07-13 R32 W vs C.Y. Lee (1500) — DOM 07-10 QF L vs Gao (1630) at W50 Corroios-Seixal Tararudee (B): 4-0 this event 07-18 SF W vs Valdmannova (1695, Strong) — DOM 6-2 6-0 07-17 QF W vs Zvonareva (1500) — DOM 07-15 R16 W vs Knutson (1500) — 3-set win 07-13 R32 W vs Okamura (1500) — DOM 07-01 L vs Alexandrova (1771) Wimbledon, tight 7-5 7-5 Both posted a DOM win over a Strong opponent in the SF — form signals cancel out. B is the stronger recent hard-courter (W100 Tokyo final + Gifu QF runs, hard). FATIGUE (precomputed) A: 7 matches/14d, 21 sets, 449 court-min, 3 back-to-backs B: 4 matches/14d, 12 sets, 276 court-min, 1 back-to-back Diff: 173 court-min, 3 matches → clear fatigue edge to TARARUDEE. NARRATIVE Let-down: none (A's Hertogenbosch title 42d ago — too old; B's Tokyo final loss 84d ago). Venue: both 4-0 at Istanbul 2 — cancels. Title recency: A won Hertogenbosch 42d ago (grass) — stale, different surface. VERDICT: PASS The only model edge on A (ELO +4.2%) is thin AND contradicted by both the ML model (35.5% for A) and the qualitative read (A carries nearly double the court time with 3 back-to-backs; B fresher and the better hard-court form). The reverse case (B ML) is only ~2.4% vs the exchange-less book price with no alt markets to reprice. No side clears the bar: models disagree with each other, and the qualitative signal kills the priced edge rather than validating it.