TennisEdge's AI model finds no betting value in S. Baez vs M. Kecmanovic (Kitzbuhel on 2026-07-20) — the market is efficient, with M. Kecmanovic favoured at 1.72.
═══ S. Baez vs M. Kecmanovic — Kitzbuhel (clay, 1/16-finals) ═══
MODELS vs MARKET
Book: Baez 2.10 / Kecmanovic 1.72 · Betfair: 2.14 / 1.80 (exchange = truth)
Implied Baez (Betfair, comm-adj): ~48.0%
ELO 49.1% A · ML 56.6% A · Blend 53.9% A
Alignment: ELO ≈ market (coin flip); only ML disagrees → single-model edge, needs qualitative validation.
RECENT FORM (last 5 each)
Baez:
07-17 Bastad QF L Rublev (1907, Elite) 6-1 6-2 — blowout
07-16 Bastad R16 W De Jong (1787, Strong) 6-4 6-2 — DOM
07-14 Bastad R32 W Dahlin (1296, Weak) 3 sets — GRIND vs Weak (red flag)
06-29 Wimbledon L Struff (1669) 5 sets, grass
06-25 Exhibition W Shevchenko grass exho
Kecmanovic:
07-16 Gstaad R16 L J.M. Cerundolo (1840, Strong) 3-6 6-3 7-5 — tight vs hot player
07-13 Gstaad R32 W Feldbausch (1720, Strong) 3 sets
06-29 Wimbledon L Sinner (2013, Elite) 5 sets — competitive
06-25 Mallorca QF L Marozsan (1553) straight sets, grass
06-23 Mallorca R16 W Sonego (1652) 3 sets
Full-window read: Baez Prostejov clay title 5-0 in June (TITLE 43d ago, clay) but poor May clay
(losses to Michelsen 1582, Gaubas 1686, PCB, Bublik). Kecmanovic Valencia clay TITLE 64d ago
beating Tabilo 1861 / Ugo Carabelli 1817 / Vallejo 1824 — comparable clay pedigree right now.
FATIGUE
Baez: 3 matches/7d, 9 sets, 333 min, back-to-back x1, last 07-17
Kecmanovic: 2 matches/7d, 6 sets, 306 min, last 07-16
Court-time diff 27 min, matches diff 1 → roughly equal (slight extra load on Baez).
NARRATIVE
Let-down: none for either (no recent Finals within 7d).
Venue: no Kitzbuhel history for either player.
Titles: Baez Prostejov 43d (clay, Challenger) · Kecmanovic Valencia 64d (clay, Challenger).
ALT MARKETS
Handicap A +1.5 @1.93 / B -1.5 @1.97 — model cover probs null, no independent margin read → skip.
Total 22.5: O 2.00 / U 1.92; model U 53.2% → edge ~1.1% — thin, pass.
VERDICT: LEAN Baez, no stake.
ML model likes Baez (56.6%) but ELO sides with the market and the qualitative layer does not
validate: 3-set grind vs a 1296 player, 6-1 6-2 blowout by Rublev, weak May clay, slightly
heavier schedule — against a Kecmanovic who just pushed an in-form Cerundolo to 7-5 in the third.
Own estimate Baez ~52% → edge ~4% at effective 2.083 (min 2.02) — below the 5% bar.
All alt-market edges thin. Single-model disagreement without confirmation = no bet.