TennisEdge's AI model finds no betting value in C. Norrie vs J. Duckworth (Los Cabos on 2026-07-29) — the market is efficient, with C. Norrie favoured at 1.33.
C. Norrie (A, age 30, rank 28) vs J. Duckworth (B, age 34, rank 86) - Los Cabos, hard, 1/8-finals
MODELS vs MARKET
ELO A 71.4pct | ML A 72.4pct | SIM A 60.2pct
Book 1.33/3.10 | Betfair ML 1.35/3.70 (A implied ~74pct)
Value flag = B +9.8pct - comes ENTIRELY from the lone dissenting sim
(SIM B 39.8pct vs Betfair-implied B 28pct). ELO and ML both side with the market.
RECENT FORM (last 5)
Norrie: 07-28 Los Cabos hard WON 6-4 6-1 vs Kovacevic (1687 Strong) = DOM over Strong on THIS surface.
07-17/18 UTS Championship = exhibition point-format (unusable). 06-29 Wimbledon grass LOST 5-set
thriller vs Zheng (1678). Then grass/clay losses.
Duckworth: 07-29 Los Cabos hard WON 6-4 3-6 4-6 vs Mayo (1529 Weak) = 3-set GRIND over Weak, won ugly.
Prior: 27-day layoff back to 07-02 Wimbledon. Grass/clay body of work.
SHAPE / PATH (Step 1b)
Norrie SHARPENING - DOM 6-4 6-1 over a Strong opponent on the match surface yesterday.
Duckworth FADING - needed 3 sets to beat a 1529 today, shaking off a 27-day layoff; won ugly.
Path: both 1 match deep at venue; Norrie clean, Duckworth a grind.
FATIGUE (Step 2)
Duckworth: 3-set match TODAY (07-29) with essentially no rest, age 34 (over 30, hardened) = extreme
spot, disadvantaged. This disadvantages the already-underdog: no value to back him, cannot feed a cover.
Norrie: 2-set win yesterday (no extra-set toll), UTS load exhibition, roughly neutral.
NARRATIVE
No Final let-down. Both arriving grass/clay onto hard - Norrie reset with a clean hard DOM, Duckworth
with an ugly 3-setter. No meaningful Los Cabos history for either.
SIM RELIABILITY
Norrie sim serve/return inputs polluted by UTS exhibition point-scoring plus off-surface grass/clay;
Duckworth built off grass/clay too. The lone 60pct sim is the outlier vs ELO 71 / ML 72 / market 74,
and is contradicted by form (Duckworth faded and fatigued, Norrie sharpening). Trust converging models.
MARKET EDGES (commission-adjusted, executable prices)
Norrie ML: win ~0.72 vs break-even 0.75, NEGATIVE. Pass.
Duckworth ML: sim +11.8pct but lone contaminated signal vs consensus, qualitatively contradicted. Pass.
Under 22.5 at 1.88: sim 56.3pct vs break-even 54.5pct = +1.8pct, below 5pct bar. Pass.
Over 22.5 at 2.04: negative. Pass.
Handicap A minus3.5 / B plus3.5: no model handicap prob; wide-lay high-variance, no valid margin thesis;
B plus3.5 also barred by self-contradiction guard (Duckworth is the fatigue-haircut side). Pass.
GATES (candidate = value-flag Duckworth ML)
G1 edge-origin: PASS numeric sim-vs-market gap exists pre-narrative, but it is a LONE input-
contaminated sim outlier; ELO/ML/market converge against it.
G2 consensus: lone-signal (sim only), stake capped, and no independent qualitative validation.
G3 fatal-risk: ML, non-blocking; noted regardless.
VERDICT
The only edge on the board is a lone, contaminated sim pointing at Duckworth, unsupported by ELO, ML,
the exchange, or form, and directly contradicted by the shape/fatigue read. No market clears its bar
with independent validation. PASS.