TennisEdge's AI model finds no betting value in M. Kouame vs G. Dimitrov (Los Cabos on 2026-07-27) — the market is efficient, with G. Dimitrov favoured at 1.31.
M. Kouame (A, age 17, #833) vs G. Dimitrov (B, age 35, #47) — Los Cabos, HARD, 1/16-finals
═══ MODELS vs MARKET (Kouame / Dimitrov) ═══
Betfair ML (executable): 2.88 / 1.51 -> norm 34.4% / 65.6% (book 3.50/1.31 is stale)
ELO: 34.2% / 65.8% (agrees with market almost exactly)
ML: 24.0% / 76.0% (lone outlier — favours Dimitrov)
Sim baseline: 43.4% / 56.6% (OFF-SURFACE outlier, distrusted — see inputs)
Sim corrected: 30.6% / 69.4% (converges to consensus after clay->hard fix)
Sharp move: book 3.50 -> Betfair 2.88 on Kouame = informed money is ON Kouame,
not Dimitrov. Backing Dimitrov fights the direction the exchange has moved.
═══ SIM INPUT INSPECTION & RE-RUN (logged) ═══
Baseline sim = 43.4% Kouame, far above ELO/market (34%) and ML (24%) — the outlier.
Cause: inputs are opponent/surface-unadjusted. Kouame's 33-match window is ENTIRELY
clay (rpw 0.426 structurally inflated — returning is easier on slow clay); Dimitrov's
45-match window is grass-heavy (Wimbledon/Mallorca/Dublin, spw 0.603 grass-inflated).
Justified correction: shade Kouame rpw 0.426 -> 0.40 for the clay->hard transition
(clamped, within +/-0.05). Re-run -> Kouame 30.6% / Dimitrov 69.4%, which CONVERGES
with ELO (34%), market (34%) and ML (24%). Corrected distribution used as anchor.
═══ RECENT FORM (last 5) ═══
Kouame (A) — entirely CLAY, FADING post-RG:
07-14 Bastad clay L N. Borges (1766) 6-4 6-2
07-06 Braunschweig L Rehberg (1605) 6-4 6-4
05-30 RG clay L Tabilo (1851) 6-4 3-6 4-6 7-6 (close 4-setter vs Elite)
05-28 RG clay W Vallejo (1822) from 2 sets down, 5th-set breaker
05-26 RG clay W Cilic (1686)
Read: hot RG junior run (beat 1800+ ELO on clay) then FADING — lost last 2 on clay.
No hard-court sample. Now transitioning clay->hard, 13 days idle.
Dimitrov (B) — grass/clay, legit tour level:
07-15 Bastad clay L N. Borges (1771) 6-4 6-2
07-14 Bastad clay W Svrcina (1658) 7-6 6-1 4-6
07-06 Wimbledon L Fery (1735) 5-setter breaker
07-04 Wimbledon W Berrettini (1872) 5 sets — genuine ELITE scalp
07-02 Wimbledon W Mensik (1650)
Read: Wimbledon R16, beat Berrettini. Class edge on hard (his best surface).
Transitioning grass/clay->hard, 12 days idle.
═══ FATIGUE (Step 2) ═══
Both last played 12-13 days ago; matches_7d = 0 for both. Dimitrov's 269min cumulative
(14d) and back_to_back=1 were 12 days ago — fully recovered, NO extreme spot (nothing
yesterday, no 3-in-3, no live b2b). Kouame 97min, also rested. Fatigue = roughly equal,
non-factor. Age note: Kouame 17 (fast recovery), Dimitrov 35 (slow) — moot after 12+ days.
═══ NARRATIVE / SANITY ═══
- Surface transition: BOTH -> hard. Kouame clay->hard (widens his margins = cover RISK,
never support). Dimitrov grass/clay->hard. Symmetric, no differential edge.
- ELO/rank sanity: Kouame ELO 34% but ranked #833 = "high ELO, modest rank" — rating
FARMED on a hot RG clay run. Discount that ELO edge; live form does NOT back it
(fading, off-surface). True Kouame prob leans below market, toward ML/corrected-sim ~30%.
- Dimitrov #47, ELO ~66% fav — ELO ~ rank, credible.
- No Finals (no let-down), no venue history, no title recency.
═══ GATES ═══
GATES
G1 edge-origin: FAIL — only positive number is Dimitrov ML via the ML model alone
(76% vs 67.4% needed). ELO, sharp market and corrected sim all sit at ~66-69% =
no edge at 1.51. A lone-model split is not a numerical edge.
G2 consensus: FAIL — ELO AND the sharp exchange price both sit against the ML edge;
no hard independent evidence (no on-surface H2H, no live-event result) overrides.
G3 fatal-risk: FAIL — sharp money has moved Kouame IN (3.50->2.88); betting Dimitrov
trades against the sharpest estimate on a lone model.
═══ EDGES AT EXECUTABLE PRICES (corrected anchor) ═══
ML Kouame @2.88: -1.5% to -5.3% PASS
ML Dimitrov @1.51: -1.6% (ELO) / +2.0% (corr-sim) / +8.6% (ML only) — below 5% bar
ex-ML; lone-model, fights sharp move -> PASS
AH Kouame +3.5 @1.73: cover 48.2% -> edge -10.9% PASS (baseline looked close but
was the off-surface artifact — correction kills it)
AH Dimitrov -3.5 @1.94: cover 51.8% -> edge -1.0%, high-variance lay -> PASS
Total UNDER 22.0 @1.76: ~51% -> edge -7.1% PASS
Total OVER 22.0 @1.93: ~49% -> edge -4.1% PASS
═══ VERDICT: PASS ═══
Market (Betfair 34%/66%) agrees with ELO and, after the justified clay->hard input
correction, with the sim (31%/69%) too. The only edge is a lone ML split on Dimitrov,
which the sharp exchange move (money ON Kouame) directly contradicts. No market clears
its bar at the executable price. Efficient price — no bet.