TennisEdge's AI model finds no betting value in F. Cerundolo vs T. Boyer (Los Cabos on 2026-07-29) — the market is efficient, with F. Cerundolo favoured at 1.33.
═══ F. CERUNDOLO (A, 27, #21) vs T. BOYER (B, 25, #257) — Los Cabos, hard, 1/8-finals ═══
MODELS vs MARKET
ELO A 76.1% | ML A 65.6% | blend A 76.1%
Book: A 1.33 / B 3.40 (impl A 71.9%) — context only
Betfair ML: A 1.13 / B 3.55 → raw 88.5% / 28.2%, overround 116.7% (WIDE/ILLIQUID book).
Normalized Betfair ≈ A 75.8% / B 24.1% — near-exact agreement with ELO 76.1%.
Executable test: A @1.13 comm-adj 1.1235 → needs 89.0%; every model is 10–23pp below → A dead.
B @3.55 comm-adj 3.4225 → needs 29.2%. ELO gives B 23.9% (against), ML gives 34.4% (for).
Models straddle the dog price; blend is below it. No clean numeric edge either side.
RECENT FORM (last 5 shown; all fetched matches used for classification)
A — F. Cerundolo
07-18 UTS Championship (EXHIBITION, hard) W vs C. Norrie
07-18 UTS Championship (EXHIBITION) L vs L. Miguel
07-17 UTS Championship (EXHIBITION) L vs B. Nakashima
06-30 Wimbledon (grass) L vs J. Munar (1586) — straight sets, UNDER-performing
06-21 London (grass) FINAL W vs T. Paul (1852) — TITLE, DOM-adjacent over Elite
B — T. Boyer
07-28 Los Cabos R32 W vs G. Diallo (1735) 7-5 6-2 — DOM over Strong, main-draw scalp
07-27 Los Cabos (quali final) W vs B. Tomic — walkover (0-0)
07-26 Los Cabos (quali) W vs L. Claverie (1256) 7-6 6-3 — GRIND vs Weak
07-23 Bloomfield Hills L vs J. Fearnley (1642)
07-21 Bloomfield Hills W vs B. Tomic (1626)
SHAPE / PATH (Step 1b)
A: UNKNOWN-to-FADING. Last competitive match 06-30 (Wimbledon straight-sets loss to a 1586).
UTS 07-17/18 is an exhibition format (placeholder 1500 ELOs, non-standard scorelines) — not
evidence of match sharpness. => 28-day competitive layoff + RUST RISK, first match at venue.
A: SURFACE TRANSITION grass/clay → hard. Depresses level and widens margin variance.
B: SHARPENING. 3 matches at this venue this week (quali SF, quali F walkover, R32), conditions
-adjusted and timing grooved; DOM over a 1735 in the main draw is real over-performance
against bracket. Match-sharp beats rusty — this is the live qualitative case for the dog.
FATIGUE (Step 2, brief-precomputed)
A: matches_14d 3 (all exhibition), 1173 court-min, last match 07-18, 10 days idle. Fresh, rusty.
B: matches_7d 5, matches_14d 5, back_to_back 2, 387 court-min (avg 77/match — light), last match
TODAY 07-28. Wins 07-26/27/28 = 3rd match in 3 consecutive days, playing a 4th.
=> EXTREME SPOT triggers on B (3-in-3, plus same-week 4th).
Calibrated hangover: B's 07-28 win was a 2-SET win → ~neutral, no −4.5pp 3-set haircut owed.
Age 25 = baseline, no modulation. Net: mild real disadvantage on B, offsetting his sharpness.
Market check: no fatigue-narrative drift to fade here — the exchange is on the fresh side already.
NARRATIVE
Let-down A: won London FINAL 06-21 = 37d ago → None (outside window).
Title recency A: TITLE London 37d ago (grass) — stale, superseded by the Munar loss + layoff.
Venue momentum: A first appearance in window; B 3-0 at this venue this week.
Boyer's "Wimbledon Final / Tunis Final" rows are qualifying/challenger draws — his level is
Challenger, his lone main-draw ATP scalp is Diallo yesterday. Class gap is genuine.
ELO/RANKING SANITY
A: ELO-fav and ranked #21 — agree, ELO credible. B: #257 with challenger-farmed wins; ELO does
not overstate him. No divergence discount either way.
ALT MARKETS
handicap null, totals null, model null → ML only. No margin market to price.
GATES
G1 edge-origin: FAIL — stripped of rust/sharpness narrative, the numbers give nothing: A @1.13
needs 89% vs models' 66–76%; B @3.55 needs 29.2% vs blend 23.9% (only lone ML 34.4% clears,
+5.2%). No pre-narrative edge at the min bar on either side.
G2 consensus: FAIL — ELO (76.1%) and the normalized exchange (75.8%) agree on A; only ML
dissents. A lone ML split is explicitly not hard independent evidence for the dog.
G3 fatal-risk: FAIL — the dog thesis rests on A's rust, but B is on his 4th match in 4 days at
#257 against a top-25 hard-courter; if the class gap holds for two service games the thesis
dies. Plus overround 116.7% = illiquid/wide exchange, the segment the backtest bleeds on.
VERDICT: PASS
Cerundolo @1.13 is a market over-shade (models say ~76%, price says 88.5%) but there is no
playable other side: Boyer at 3.55 comm-adj needs 29.2% and my read is ~30% — min_odds
1.05/0.30 = 3.50 raw, which 3.55 fails once commission is applied. Rust vs match-sharp is a
real qualitative pull toward Boyer, and B's own 3-in-3 schedule cancels most of it. Sub-1%
edge on a wide illiquid book is noise, not value. No bet.
Estimate: Cerundolo ~70% / Boyer ~30%.