TennisEdge's AI model finds no betting value in V. Gaubas vs P. Carreno-Busta (Estoril on 2026-07-20) — the market is efficient, with P. Carreno-Busta favoured at 1.40.
═══ V. Gaubas vs P. Carreno-Busta — Estoril (clay), 1/16-finals ═══
MODELS vs MARKET
Book: Gaubas 3.00 (33.3% w/vig) | Carreno-Busta 1.40 (71.4% w/vig)
ELO: Gaubas 24.8% | PCB 75.2%
ML: Gaubas 36.3% | PCB 63.7%
Exchange: none (alt_markets null) — book-priced event, 8% edge bar applies
Alignment: ELO leans PCB harder than market (+3.8% at book), ML leans softer
(-7.7%). Models disagree with each other; no consensus edge either direction.
RECENT FORM (last 5 shown; all fetched matches classified)
Gaubas (1-4 last 5):
07-13 Bastad L De Jong (1779) 7.7-6.1 7-5 — GRIND loss vs Strong
07-09 Braunschweig L Diaz Acosta (1811)
07-07 Braunschweig W Basilashvili (1673) — GRIND (TB) vs Strong
06-30 Wimbledon 1/64 L Jacquet (1717)
06-25 Wimbledon Final W Lajovic (1558) 5 sets — GRIND vs Weak
Carreno-Busta (2-2 last 4, quality wins beyond):
07-15 Umag 1/8 L Ugo Carabelli (1809) 6-3 5-7 5-7 — GRIND loss vs Strong
07-14 Umag 1/16 W Dodig (1640) straight sets — DOM vs Avg
07-02 Wimbledon L Jodar (1666) 5 sets
06-29 Wimbledon W Shapovalov (1583) — GRIND (TB)
05-29/25 FO: W Tirante (1811), W Lehecka (1847) — real Strong-opponent wins on clay
FATIGUE
Gaubas: 3 matches/14d, 404 court min, last played 07-13 (7d rest)
PCB: 2 matches/14d, 285 min, 1 back-to-back, last played 07-15 (5d rest)
Court-time diff ~119min on Gaubas but both fully rested now → no live fatigue edge.
NARRATIVE
Let-down: Gaubas first Final = Wimbledon (quali) Final Won 06-25, 25d ago → None.
Venue: no Estoril history for either player in window.
Read: PCB clearly the better clay player right now (beat Lehecka/Tirante/Baez in
the window); Gaubas 1-4 in last 5 with two Challenger-level losses. Qualitative
supports the favourite — but that is priced.
VERDICT — PASS
PCB edge at book: 75.2% - 71.4% = +3.8% (the brief's 7.0% "value_edge" ignores
vig). ML model actually shades PCB BELOW market. No Betfair exchange price →
book-priced rule requires ≥8% edge; 3.8% split-model edge at 1.40 fails it.
min odds for 75.2% = 1.40 — book price is exactly at the floor, zero margin.
No case for Gaubas either (losing form, ELO 24.8%). Efficient line. Pass.