TennisEdge's AI model finds no betting value in F. Ferreira Silva vs L. van Assche (Estoril on 2026-07-20) — the market is efficient, with L. van Assche favoured at 1.20.
═══ F. Ferreira Silva vs L. van Assche — Estoril (clay), 1/16-finals ═══
MODELS vs MARKET
Book: A 4.50 (21.1%) | B 1.20 (83.3%)
Betfair: A 5.10 (19.6%) | B 1.23 (81.3%)
ELO: A 24.7% | ML: A 22.7% | Blend: A 23.5% → value_pick A, edge +2.4% (book) / +3.9% (exchange)
Alt (Betfair): AH A+4.5 @2.04 / B-4.5 @1.88 | Total 21.0 O1.96/U1.98
Sim: A covers +4.5 = 60.1% vs 49.0% implied | O20.5 = 52.4%
RECENT FORM (last 5 shown; all fetched matches used)
Ferreira Silva (5-12 overall, 90d):
07-12 Bastad SF L vs Krumich (1600, Avg) 6-4 6-2 (straight)
07-07 Iasi 1/16 L vs Pavlovic (1561, Avg) 3 sets
07-02 Brasov 1/8 L vs Wallin (1518, Avg) straight
06-30 Brasov 1/16 W vs Butvilas (1612, Avg) 6-3 6-3 DOM
06-22 Wimb QF L vs Dodig (1537, Avg) GRIND (2 TBs)
Pattern: 5-12, repeatedly losing to Avg/Weak opposition, several 5-6 game blowout margins.
van Assche (8-7 overall, 90d):
07-15 Umag 1/8 L vs Droguet (1798, Strong) tight 3 sets
07-13 Umag 1/16 W vs Lajovic (1652, Strong) GRIND
06-29 Wimb 1/64 L vs Fucsovics (1703, Strong) 3-6 0-4 (likely ret.)
06-20 Parma FINAL W vs Ofner (1755, Strong) TITLE
06-19 Parma SF W vs Rincon (1604, Avg) 6-3 6-0 DOM
Pattern: beats Strong opponents on clay, Parma title 30d ago, capable of 6-0/6-1 sets vs Avg opposition.
FATIGUE
A: 2 matches/14d, 252 min court time, last match 07-12 (8d rest)
B: 2 matches/14d, 322 min court time, last match 07-15 (5d rest)
Diff 70 min on B, but both well rested by matchday — negligible.
NARRATIVE
Let-down: none live (A title Francavilla 77d ago — stale; B title Parma 30d ago).
Venue Estoril: 0-0 both — no signal.
Title recency favors B: Parma 30d ago on clay with wins over Strong opposition.
VERDICT: PASS (all markets)
Every priced edge sits on the Ferreira Silva side (ML +2-4%, A+4.5 cover 60.1% vs 49.0% implied = +11.1% on paper).
But the qualitative read kills it: A is 5-12 vs mostly Avg/Weak opponents with frequent 5-6 game losing margins
right at the AH line; B is a Strong-bracket clay player fresh off a title run who regularly posts 6-0/6-1 sets
vs this opponent tier. Historical base rate of A covering +4.5 in his own losses is ~4/12 (~53% incl. wins) —
well short of the sim's 60%, and the same model overrates A on ML vs the exchange (23.5% vs 19.6%), suggesting
a systematic A-side bias bleeding into the cover estimate. Model-market disagreement exists, qualitative
validation does not → no bet in any market. Totals edge <2 pts — noise.
Predicted winner: L. van Assche (~78%).