TennisEdge's AI model finds no betting value in S. Wei vs H. Kadi (Vancouver on 2026-07-25) — the market is efficient, with S. Wei favoured at 1.04.
S. Wei vs H. Kadi — Vancouver (hard), Semi-finals
═══════════════════════════════════════════════
MODELS vs MARKET
ELO: Wei 54.5% | ML: Wei 55.8% | Blend: Wei 64.8%
Book: Wei 1.04 (87.4% implied) vs Kadi 7.2 (13.9%)
Betfair ML: 1.02 / 1.02 on BOTH sides — degenerate/empty market, no executable exchange price.
Nominal value_pick B, edge ~22.6% — but see data quality below.
DATA QUALITY (decisive)
H. Kadi: ZERO match history (90d/60d), no ranking, no age, no ELO evidence. The 54.5%
ELO prob for Wei is a default-1500-rating artifact, not information. The entire
model-market gap originates from the model knowing nothing about Kadi.
S. Wei: ranked 287 but visible form is 2-5 with losses to ~1500-ELO opposition at
W15/W35 level; yet she stands in this SF with wins not present in the feed — the
event's results are not fully ingested. Both sides of the model input are unreliable.
RECENT FORM (5 most recent)
Wei: L Y. Sun (1500, Avg) W15 Luan 07-08 (4-6 1-2, unfinished scoreline) | L S. Ye
(1500, Avg) W15 Maanshan 07-01, 3 sets | L Y. Ren (1508, Avg) W50 Wuning 06-02
(7-6 set, grind) | L J. Wang (1468, Weak) W35 Wuning 05-28 | W J. Zhang (1304,
Weak) W35 Wuning 05-27 GRIND. Zero DOM wins; only win vs Weak. 17-day gap
before this event.
Kadi: no matches on record; form_analysis_fallback empty.
FATIGUE
Precomputed: both 0 matches/14d, 0 court minutes, well_rested (a feed gap for this
event, not genuine rest). Roughly equal — no fatigue input used anywhere below.
Wei age 22 (fast recovery), Kadi age unknown — age-neutral.
NARRATIVE
Let-down: none (no Final on record). Venue momentum: no prior Vancouver matches for
either. Surface transition: Wei's last matches were hard-court, same as here — no
transition. Nothing at all available for Kadi.
ELO vs RANKING SANITY
Wei ranked 287 yet her ELO makes her only a 54.5% favourite against a completely
unrated opponent — the ELO is not informative here, it is a placeholder. Kadi has no
ranking to cross-check. ELO untrustworthy on both sides; discount, never lean on it.
GATES
G1 edge-origin: FAIL — the numeric edge on Kadi exists only because ELO/ML default an
unknown player to ~1500; a default rating is not model-market disagreement.
G2 consensus: FAIL — market and any sane read sit against B, and there is no hard
independent evidence (zero H2H, zero live-event results, zero history for Kadi).
G3 fatal-risk: FAIL — if Kadi is simply an unrated local-level entrant (the natural
reading of a 7.2 book price and a dead exchange), the thesis is negated outright.
PRICING
No executable exchange price: betfair_ml shows 1.02/1.02 on both sides, i.e. dead
liquidity, so ML would be book-priced and require an 8% edge on evidence that does not
exist. Backing Wei at 1.04 offers no edge at any plausible probability; backing Kadi
at 7.2 is a pure data-void punt. No handicap or totals lines offered.
VERDICT: PASS. All three gates fail. The apparent 22.6% value on Kadi is a missing-data
artifact, not an edge. Match-winner forecast: S. Wei.