TennisEdge's AI model finds no betting value in J. Lu vs C. Dolehide (Vancouver on 2026-07-27) — the market is efficient, with C. Dolehide favoured at 1.09.
J. Lu vs C. Dolehide — Vancouver (hard), 1/16-finals
MODELS vs MARKET
Book: Lu 5.65 / Dolehide 1.09 (implied Lu 16.2%)
Betfair (executable): Lu 4.80 / Dolehide 1.03 -> Lu implied 20.8%; commission-adj eff 4.61 -> 21.7%
ELO 26.3% | ML 26.8% | Sim 29.0% for Lu -> nominal edge on Lu ~5-7% vs exchange
Backend value_pick A (+12.8%) is computed vs the stale book 5.65 — the exchange at 4.80 contradicts it.
ELO vs RANKING SANITY (decisive)
Lu: age 36, ranked 387, entire recent record at W35/W50 ITF level. Wins come only vs Weak
(<1500 ELO) opponents; losses to 1353/1500/1596-level players. A 27% model prob vs a top-111
title-winner is not credible — the rating context is weak-field ITF.
Dolehide: ranked 111 but nearly all her recent opponents carry the 1500 default ELO (sparse
rating data at W100/WTA events) — her ELO is data-starved and almost certainly UNDERSTATES her.
Classic strong-rank/low-ELO case: do not back the opponent on the ELO gap.
RECENT FORM (top 5 each)
Lu: 07-25 L SF Saskatoon vs Chavez (1353, Weak); 07-24 W QF vs Marton (1241) DOM-Weak;
07-23 W 1/8 vs Plekhanova (1417) DOM-Weak; 07-21 W 1/16 vs Frey (1282) GRIND-Weak (TB);
07-16 L Dallas 1/8 vs Shcherbinina (1500, Avg).
Pattern: beats Weak only, loses to Avg and up. Lost her last match to a 1353 opponent.
Dolehide: 07-26 W FINAL W100 Evansville vs Ku — two tiebreaks, GRIND; 07-25 W SF vs Zarazua
7-5 6-4 DOM; 07-24 W QF vs Kalieva GRIND (TB, 3 sets); 07-23 W 1/8 DOM; 07-22 W 1/16 DOM.
5-0 title run. Earlier: competitive 4-6 6-7 loss to Montgomery (1895, Elite) at Wimbledon.
TITLE: W100 Evansville 1d ago (Dolehide) — live form strongly with her.
FATIGUE
Lu: 6 matches/14d, 18 sets, 495 court-min, 3 back-to-backs, last 07-25. Age 36 -> slow recovery.
Dolehide: 5 matches/14d (title run through 07-26), ~352+ court-min, 3 back-to-backs. Age 27 baseline.
Court-time diff ~143 min -> Lu is the heavier side; age 36 hardens the penalty. Fatigue sits
AGAINST the only side the models like.
LETDOWN: Dolehide won a Final 1d ago -> LETDOWN flag. Narrative only — cannot create an edge (G1),
and her run is itself live evidence of level.
GATES
G1 edge-origin: FAIL — the nominal 5-7% model-market gap exists numerically, but it rests
entirely on an ELO pair the sanity check discredits (Lu rating inflated by weak fields,
Dolehide rating understated by sparse data). Stripped of the distrusted rating gap there is
no independent numeric edge; the pro-Lu residual is the letdown narrative, which cannot
originate an edge.
G2 consensus: FAIL — exchange prices Dolehide 1.03; the only hard live evidence available
is Dolehide going 5-0 for a W100 title this week — live evidence AGAINST the pick.
G3 fatal-risk: FAIL — Lu is the fatigue-disadvantaged side (495 min, 3 B2Bs, age 36) and has
lost every recent match vs 1500+ opposition; either fact alone negates a Lu ML thesis.
OTHER MARKETS
Dolehide ML 1.03 eff 1.029: no edge. alt_markets handicap/totals null, no model margin probs,
no independent margin read -> no alt-market bet.
VERDICT: PASS. The apparent value on Lu is a stale-book artifact plus an untrustworthy ELO gap;
the exchange, live form, fatigue and class all point the same way. Predicted winner: C. Dolehide.