TennisEdge's AI model finds no betting value in . R. Pascual F. vs C. Smith (Los Cabos on 2026-07-26) — the market is efficient, with C. Smith favoured at 1.19.
MATCH: R. Pascual F. (rank 768, age 20) vs C. Smith (rank 145, age 23) — Los Cabos, hard
Book odds: 4.30 / 1.19 (implied A 21.7%). No exchange markets (alt_markets null — book-priced only).
MODELS
ELO prob A: 31.5% | ML prob A: 33.8% | Blend: 36.0% | Market: 21.7%
Nominal edge on A: ~12-14 pts. Value pick flagged A (+14.3%).
ELO/ranking sanity: FAILS. A is ranked 768 yet models give him 32-36% vs a rank-145
opponent. A's last ~16 matches: M15 Monastir events vs opponent_elo 1500 defaults
(unrated fields) plus one small challenger — textbook farmed rating. The ELO edge is
inflated by weak-field wins and must be discounted. B rank 145 with real challenger/
tour-level results (beat Mmoh 1739, pushed Bu 1733 to three) — market side is credible.
RECENT FORM (5 most recent shown; all fetched matches used in counts)
Pascual F. (A): 12-4 in window, but ALL opponents Weak bracket (<1550, mostly 1500 default)
07-23 Bloomfield Hills 1/8 L vs A. Martin (1522) 3-6 6-1 3-6
07-21 Bloomfield Hills 1/16 W vs M. Imamura (1527) 6-1 6-1 DOM (vs Weak)
07-20 Bloomfield Hills Final W vs S. Kozlov (1562) 4-6 4-6
07-19 Bloomfield Hills SF W vs M. Kuhar (1174)
06-28 M15 Monastir 24 Final W vs M. Picard (1500) 3 sets GRIND
Zero wins vs Avg/Strong/Elite in the entire window. Just lost to a 1522 player.
DATA-QUALITY FLAG: Bloomfield Hills rounds are chronologically inconsistent
(Final 07-20 dated before 1/16 on 07-21 and 1/8 on 07-23) — history unreliable.
Smith (B): quality of opposition far higher
07-21 Bloomfield Hills 1/16 L vs J. Wolf (1622) 6-7 3-6 GRIND (competitive)
07-18 Lincoln QF L vs Y. Bu (1733) 3 sets w/ TB, GRIND vs Strong
07-16 Lincoln 1/8 W vs H. Matsuoka (1540) 6-3 7-5
07-14 Lincoln 1/16 W vs A. Ilagan (1587) 3 sets TB GRIND
06-25 Wimbledon Final L vs M. Zheng (1636)
Also in window: DOM win over M. Mmoh (1739) at Little Rock. Losses are tight vs
Strong opponents — no form collapse.
FATIGUE
A: 4 matches / 7d, 12 sets, 351 court-min, back-to-back x2, last played 07-23 (2d rest).
Age 20 -> recovers fast, penalty softened.
B: 2 matches / 7d (4 in 14d), 501 court-min, no back-to-backs, last played 07-21 (4d rest).
Age 23 -> baseline. Raw court-time gap ~150min against B, but B has 4 days rest and
fewer recent matches; call it roughly neutral in practice. Fatigue cannot create an
edge regardless (G1).
NARRATIVE
Let-down: A won Bloomfield Hills Final 07-20 per data (5d ago -> WEAK letdown), but
round-order data is corrupt, low confidence in the signal. B lost Wimbledon Final 30d
ago — outside window. Venue: no Los Cabos history for either.
Surface transition: none — both on hard.
GATES
G1 edge-origin: [PASS] — edge is numeric (models 32-36% vs market 22%), not fatigue/narrative-built.
G2 consensus: [PASS] — ELO and ML both sit above market on A; not a lone-signal override.
G3 fatal-risk: [FAIL] — the numeric edge rests on a farmed ELO: A's rating was built
entirely on M15 1500-default fields (rank 768 confirms), his history data
is internally inconsistent, and he just lost to a 1522 player. If the
rating is inflated — and every independent check says it is — the thesis
is gone. No exchange price exists to validate the 4.30 book quote
(book-priced small-event segment = known backtest loser).
VERDICT: PASS. The 14% edge on Pascual F. is a model artifact of weak-field ELO farming,
not market error. Rank 768 vs 145, zero wins above the Weak bracket, corrupted match-history
data, and no exchange liquidity to confirm the price. Skeptic wins easily here.