TennisEdge's AI model finds no betting value in H. Gaston vs T. Droguet (Estoril on 2026-07-22) — the market is efficient, with T. Droguet favoured at 1.42.
═══════════════════════════════════════════
H. Gaston vs T. Droguet — Estoril (clay) 1/8-finals
═══════════════════════════════════════════
MODELS
ELO: Gaston 30.0% | Droguet 70.0%
ML: Gaston 55.3% | Droguet 44.7% <-- violent disagreement w/ ELO
Blend: Gaston 30.0% (server leans ELO)
Market: Book 2.90/1.42 | Betfair(exec) 2.52/1.59 -> Droguet ~63% implied
Book overpriced Droguet (1.42) vs exchange (1.59).
RECENT FORM (last 5)
Gaston (A):
07-21 Estoril 1/16 W Dzumhur (1711) 5-0 WALKOVER/RET
07-14 Bastad 1/16 L Altmaier (1729) 6-2 7-5
07-12 Braun. Final L Choinski (1782) 4-6 6-0 6-4
07-11 Braun. SF W Diaz Acosta (1824,Strong) 7-5 TB 5-7 GRIND
07-10 Braun. QF W Moeller (1632) 6-2 6-4 DOM
-> clay specialist, reached Braunschweig final; genuinely good clay form.
Droguet (B):
07-20 Estoril 1/16 W Ugo Carabelli (1809,Strong) 6-0 3-6 6-4 GRIND
07-16 Umag QF L Merida (1757) 6-1 6-3
07-15 Umag 1/8 W van Assche (1710,Strong) 1-6 6-1 5-7 GRIND
07-14 Umag 1/16 W Blockx (1861,Elite) 3-6 6-2 6-0 GRIND
07-12 Iasi Final L Piros (1792) 1-6 TB
-> hot streak beating strong/elite clay names, but ALL grinds, 3rd straight event.
FATIGUE
Gaston: 7 matches/14d, 854 court-min, B2B 3 (today's win a walkover -> minimal load)
Droguet: 9 matches/14d, 1053 court-min, B2B 5, 4 matches/7d, 3-setter yesterday
Court-time diff 199min (>60) -> Droguet (favorite) disadvantaged. Edge to the DOG.
NARRATIVE
Let-down: both lost a Final 07-12 (9d ago) -> no active signal.
Venue: neutral (Estoril new for both).
Fatigue + freshness + clay-specialist all point to Gaston keeping it tight/live.
VERDICT: PASS
The only model edge is ELO's 70% on Droguet -> ~5.9% vs commission-adj Betfair 1.59.
But that edge is CONTRADICTED by (a) the ML model (Gaston 55%) and (b) the fatigue
read — Droguet is the more-worn favorite on his 3rd straight event, Gaston fresher and
a lefty clay specialist. No qualitative validation for backing the favorite; and the
dog's exec price (2.52) sits far above the blended model's 30%. Alt markets thin:
Gaston +2.5 cover 52.8% -> only ~3.5% edge (comm-adj) < 5% bar; totals ~0%.
Manufactured single-model edge + contradicting qualitative = PASS.
═══════════════════════════════════════════