TennisEdge's AI model finds no betting value in T. Droguet vs C. Ugo Carabelli (Estoril on 2026-07-20) — the market is efficient, with T. Droguet favoured at 1.90.
═══ T. Droguet vs C. Ugo Carabelli — Estoril (clay), 1/16-finals ═══
MODELS & MARKET
Book: 1.90 / 1.90 (no-vig 50/50) | Betfair ML: 2.10 / 1.88 (exchange leans B ~53%)
ELO: 47.6% A | ML: 49.1% A | Blend: 48.4% A → all mildly favor B (51.6%)
Value pick B, no-vig edge +1.6% — but raw implied for B at 1.90 is 52.6%, so raw EV is negative.
RECENT FORM (last 5 shown; all fetched matches used)
Droguet (A): 5-3 last 8, brutal stretch
07-16 Umag QF L vs Merida Aguilar (1757) 6-1 6-3 — flat loss
07-15 Umag R16 W vs van Assche (1710) 3 sets — GRIND
07-14 Umag R32 W vs Blockx (1861, Elite) 3 sets — GRIND, quality scalp
07-12 Iasi FINAL L vs Piros (1792) w/ TB — GRIND
07-11 Iasi SF W vs Royer (1728)
Profile: tiebreak-heavy grinder; 1 DOM (vs Heide 1796 07-10), most wins GRIND.
Ugo Carabelli (B): 2-1 Umag run, then gap to Wimbledon
07-16 Umag QF L vs Burruchaga (1798) 3 sets
07-15 Umag R16 W vs Carreno-Busta (1850, Elite) 3 sets — GRIND, quality scalp
07-14 Umag R32 W vs Topo (1544) 3 sets, 2 TBs — GRIND vs Weak
06-29 Wimbledon L vs Merida Aguilar
06-25 Exhibition L vs Bellucci
Profile: clay-solid, beat an Elite last week, but grinds even Weak opponents.
FATIGUE (precomputed)
A: 8 matches/14d, 24 sets, 908 court-min, b2b streak 5, last 07-16
B: 3 matches/14d, 9 sets, 532 court-min, last 07-16
Diff 376 min → heavier side (A) disadvantaged, BUT both have had ~4 days rest since Umag, which blunts most of it.
NARRATIVE
Let-down: A lost Iasi Final 07-12 — 8d ago, outside all windows → None. B: no finals → None.
Venue: neither player has Estoril history in fetched window → no venue signal.
Title recency: A — Won "ATP Madrid - Final" 90d ago (clay, data-quirk entry); too stale to matter.
VERDICT: PASS
Every signal points the same direction: models lean B, the sharper Betfair exchange leans B harder
(1.88), and the fatigue differential favors B — the market has already priced all of it. Book no-vig
edge is only +1.6% and raw EV at 1.90 is negative. Alt markets add nothing: B -0.5 model edge ~0.7%,
Under 23 ~3% with no independent margin read (both players are tiebreak grinders, which if anything
argues Over, killing the thin Under lean). Full model+market consensus = no edge.