TennisEdge's AI model finds no betting value in N. Borges vs O. Luz (Estoril on 2026-07-21) — the market is efficient, with N. Borges favoured at 1.04.
═══ N. Borges vs O. Luz — ATP Estoril 1/16-finals (clay) ═══
MODELS vs MARKET
Book: A 1.04 / B 10.00 (implied A 90.6%)
ELO 79.1% A | ML 82.8% A | Sim 72.2% A → value_pick B, edge 18.4% vs book
Betfair ML snapshot: A 1.03 / B 1.02 — INVALID pair (probs sum ~1.95); no
executable exchange price for Luz anywhere near 10.0 can be verified.
RECENT FORM (last 5 shown)
Borges: L Darderi(1922) 6-7 4-6 GRIND | W Dimitrov(1681) 6-4 6-2 DOM |
W Kouame(1491) 6-4 6-2 DOM | L Sinner(2016) Wimbledon | W Boyer(1532)
→ Competitive vs Elite, DOM vs Strong/Avg. Bastad clay QF run.
Luz: W Tabur(1648) Final YESTERDAY | W Tseng(1662) ret. | L Gombos(1555)
Braunschweig Final | W Sanchez Quilez(1550) | L Onclin(1563) 1-6 3-6
→ 4-1 but ALL vs Avg/Weak; zero wins above 1662; lost to a 1555.
FATIGUE
Borges: 3 matches/14d, 318 min, last played 07-17 (2 days rest)
Luz: 4 matches/14d, 376 min, back-to-back x2, played YESTERDAY
→ Luz clearly heavier (+58 min, b2b, zero rest). Disadvantage: Luz.
NARRATIVE
LETDOWN: Luz won a Final 1 day ago → peak-then-letdown spot (0-2d, strongest tier).
Venue: Luz 2-0 at this Estoril event — but vs Avg opposition; venue never creates an edge.
VERDICT — PASS
The 18.4% "value" on Luz is a quality-blind model artifact: his streak is
challenger-grade opposition, he is the most fatigued player on the slate, and
he sits in a textbook post-title letdown. Borges' true prob is plausibly
88-90%, not 72-83%. Pricing discipline seals it: the exchange snapshot is
corrupt (B at 1.02), so no executable price exists for the Luz side — book
10.0 cannot be confirmed and book quotes never validate value. Borges side:
Betfair 1.03 → eff 1.0285, min odds at even 90% prob = 1.17 → auto-PASS.
No stakeable market (handicap/totals null). PASS.