TennisEdge's AI model finds no betting value in G. Diallo vs T. Boyer (Los Cabos on 2026-07-28) — the market is efficient, with G. Diallo favoured at 1.38.
G. Diallo (A, rank 40, age 24) vs T. Boyer (B, rank 257, age 25) — Los Cabos, hard, best-of-3
═══ MODELS vs MARKET ═══
ELO Diallo 65.7% | ML Diallo 62.5% | consensus 65.8%
Book Diallo 1.38 (72%) / Boyer 3.05 — SOFT, contradicted by exchange
Betfair (truth) Diallo 1.62 (comm-adj 62.9%) / Boyer 2.60 (comm-adj 39.7%)
ELO + ML + Betfair + ranking (40 vs 257) ALL converge: Diallo ~62-66% favourite. Efficient.
═══ SIMULATION — REJECTED (favourite backwards) ═══
resimulate baseline: Diallo win 41.3% / Boyer 58.7% — sim makes BOYER the favourite,
inverting every other signal. Cause = unadjusted serve/return inputs:
Diallo spw .582 rpw .401 — built vs ELITE grass fields (Draper 1848, De Minaur 1879,
Mannarino 1763, Bonzi 1722) on a low-return surface (rpw deflated).
Boyer spw .591 rpw .409 — FARMED vs weak Challenger/ITF (opp ELO 1250-1650, many <1500).
The divergence (sim 41% vs market/ELO/ML 62-65% for Diallo) is far beyond a ±0.05 input
clamp to correct honestly; forcing 4 maxed overrides to a predetermined answer = manufacturing.
=> the sim's handicap_lines and totals_lines are NOT a trustworthy margin anchor here.
Note: the sim even puts the "-" line on Boyer (Boyer -2.5 46.7%), while the MARKET lays
Diallo (-2.5) — complete inversion, confirming the sim is unusable for the real lines.
═══ RECENT FORM (last 5 each) ═══
Diallo: L Sonego 1643 (5 sets, grass); W Bonzi 1722 (grind/ret, grass); L Draper 1848 (6-1 6-4);
W Etcheverry 1567 (grind, grass); W Atmane 1499 (grind, grass). 3-5 recent, all grass, wins
are GRINDs vs Avg/Weak. Shape STABLE-to-FADING. Last match 2026-07-02 = 25 DAYS OFF (rust risk)
AND grass->hard TRANSITION (level depressed, margins wider — a RISK vs a lay, not support).
Boyer: W Tomic 1500 (Los Cabos F, data 0-0); W Claverie 1256 (7-6 6-3 — grind vs Weak =
under-performing); L Fearnley 1642 (6-4 6-4); W Tomic 1626 (ret 3-1); L Borges 1594.
Path: 4 matches in 7d at Los Cabos on hard = MATCH-SHARP, conditions-grooved. Grinder who
stays close (many 3-setters/tiebreaks). Strong clay pedigree (Tunis title) but rank 257 class.
═══ FATIGUE ═══
Diallo: well rested (0 matches 14d) — but that = rust, not freshness (see form).
Boyer: 4 matches/7d, 11 sets, back_to_back 1, last 07-27, 297 min TOTAL but avg only 74 min/match
(quick matches). Not an extreme spot (no 150min+ 3-setter yesterday; age 25 baseline recovery).
Per calibration: cumulative load = CONTEXT only, market over-prices visible fatigue —
cannot create an edge and cannot enter a margin thesis.
═══ NARRATIVE ═══
Diallo: grass->hard TRANSITION (last 2 matches grass) — depresses level, widens margins.
Boyer: on-surface momentum (hard win streak at Los Cabos); no let-down spot relevant to this side.
Book badly overprices Diallo (1.38) but exchange already corrects it (1.62) — no free ML value.
═══ CANDIDATE EDGES (all priced at the EXECUTABLE Betfair price) ═══
ML Diallo @1.62: edge = 65.8% - 62.9% = +2.9% (ML model -0.4%). Below 5%. min 1.60. NO.
ML Boyer @2.60: model 34-37.5% vs comm-adj implied 39.7% = NEGATIVE. NO.
AH Diallo -2.5 @1.98: laying favourite -2.5 = HIGH-VARIANCE tier (needs edge >=10%). Sim
unusable; brief's biased 56% cover is not credible; with the correct favourite ~market-fair.
Diallo rust+transition vs a match-sharp opponent = margins against a lay. NO (variance + G3).
AH Boyer +2.5 @1.93: independent cover est ~55% (Boyer wins ~38% + loses <=2 ~17%) vs comm-adj
implied 53.1% = ~+2%. Below 5%, and the lean rests on rust/transition narrative. NO.
Total under 23.0 @1.98: sim has large 3.0-game bias correction + unusable inputs; no clean edge. NO.
GATES (evaluated on the least-bad candidate, AH Boyer +2.5)
G1 edge-origin: FAIL — pre-narrative numerical edge ~+2% (< 5% bar); the only pull toward a
bet is Diallo rust/transition, i.e. narrative creating the edge. Not allowed.
G2 consensus: FAIL — market AND ML both sit with Diallo as favourite; no hard independent
evidence (no on-surface H2H, Boyer's live results are vs sub-1500 fields).
G3 fatal-risk: FAIL — Boyer +2.5 thesis needs Diallo to stay tight, but the same rust+
transition also raises blow-out variance (loose favourite runs a match-sharp dog off the
court just as easily); the negating risk is live. Units cannot buy it back.
═══ VERDICT: PASS ═══
Independent signals (ELO, ML, Betfair, ranking) converge on Diallo ~62-65% — the exchange is
efficient (ML edge ~3% < 5%). The point-by-point sim is rejected (favourite inverted by farmed/
surface-biased serve-return inputs), so no trustworthy handicap/total anchor exists. The soft
book price and Diallo's rust+transition are real observations but produce RISK, not a clean
numerical edge that clears the bar at the executable price. No bet.