TennisEdge's AI model finds no betting value in B. Nakashima vs J. Mensik (Washington on 2026-07-30) — the market is efficient, with J. Mensik favoured at 1.69.
B. Nakashima (A, #33, age24) vs J. Mensik (B, #18, age20) — Washington, hard, 1/8-finals
MODELS vs MARKET
ELO A 0.368 | ML A 0.318 | blended A 0.382 -> all three favour B (ML gives B ~0.68).
Market: odds_a 2.17 / Betfair 2.10 ; odds_b 1.69 / Betfair 1.71 -> implied A ~0.44-0.48.
Models and market AGREE on the winner (Mensik). Nominal value_pick B, value_edge 0.056 vs book.
Executable test (Betfair B): eff odds = 1+(1.71-1)*0.95 = 1.6745 -> 1/1.6745 = 0.597.
Model B prob 0.618 -> edge ~2.1% at the exchange. BELOW the 5% bar. No exchange edge on the fav side.
RECENT FORM (most-recent 5)
Nakashima: Won v Etcheverry (1780,Strong) 6-3 6-4 DOM; then UTS exhibition block (ignore, novelty format);
Wimbledon L v Struff 5-set, W v Pinnington Jones; Queen's SF incl. QF W over De Minaur (1898,Elite) grass.
Read: SHARPENING on hard - clean DOM over a Strong opp in match one, live top-tier grass wins in June.
Mensik: Won v Svajda (1654,Avg) 2-6 7-6 3-6 - a 3-set GRIND over an Avg opp in his first hard match;
prior: Wimbledon L v Dimitrov (grass, ~26d ago); French Open SF run on CLAY (beat Rublev/De Minaur/Fonseca).
Read: UNDER-performing shape on this surface - needed 3 sets to beat a 1654.
PATH / SHARPNESS
Nakashima match-sharp on hard (won here yesterday). Mensik carries RUST RISK: last match ~26d ago,
grass->hard TRANSITION (widens margins, depresses level), and his hard-court reintro was a scrappy 3-setter.
FATIGUE
Nakashima: 2-set win yesterday (~90min) -> roughly neutral (~-2pp), age24 baseline. 14d minutes inflated by
UTS exhibition -> context only, no edge.
Mensik: WON a 3-set match yesterday -> -4.5pp hangover haircut, softened for age20 (recovers fast) ~-2/-3pp.
This trims B win prob toward ~0.59, thinning the already-sub-bar B edge further. Not an extreme spot either side.
NARRATIVE
Surface transition on B (grass->hard) - a RISK on the model-favourite side, not support for a bet on B.
No first-Final let-down. No title recency. ELO/ranking roughly agree for both (credible ratings).
GATES
G1 edge-origin: FAIL - only the fav side (B) has a nominal edge, and commission-adjusted at the exchange
it is ~2%, below the 5% bar. Strip narrative and there is no model-vs-market gap to stake; my qualitative
read (B rusty/under-performing) points AGAINST the only side that carries an edge.
G2 consensus: n/a (stake-tier only) - full model+market consensus on the winner = efficient.
G3 fatal-risk: n/a (no margin bet; would-be ML has no G1 edge anyway).
VERDICT: PASS. Models and market agree Mensik is the favourite; the only nominal value (B ~2% at Betfair) is
below bar and sits on the side my form/path read questions (rust, grass->hard transition, 3-set grind over an
Avg opp + hangover haircut). Backing A has no numeric model edge (models 0.38 vs market ~0.48). No stakeable edge.