TennisEdge's AI model finds no betting value in M. Giron vs C. Hewitt (Washington on 2026-07-28) — the market is efficient, with M. Giron favoured at 1.07.
M. Giron (A, 33, ATP #77) vs C. Hewitt (B, 17, ATP #653) — Washington, hard.
MODELS vs MARKET
ELO Giron 81.1% ML Giron 54.2% Blend Giron 67.6%
Market (Betfair) Giron 1.09 = 91.7% ; book 1.07. value_pick=B edge +21%.
All three models sit BELOW market on Giron — the blend's "value on B" is a
manufactured edge. Cause: Hewitt is a 17yo junior (rank 653) whose spw/rpw
and ELO are farmed on ITF M15/M25 futures fields (opp ELO 1200-1500) PLUS
two retirement-affected "wins" this week (Svajda 5-2/0-2 incomplete, Lajovic
5-2 incomplete). Giron's raw stats are DEFLATED — built vs elite grass/clay
fields (Zverev 1880, Shelton 1748, Draper 1818). Classic opponent-unadjusted
pollution: ML/ELO/sim all overstate the junior. Market's 92% is the sharp read.
SIM (validate + bounded re-run)
Baseline: makes Hewitt the 54.5% FAVOURITE, exp margin -0.5, exp total 22.2 —
absurd for a tour pro vs a junior on hard; inputs corrupted (see above).
Bounded correction (all four clamped ±0.05: spw_a .588->.638, rpw_a .404->.454,
spw_b .579->.529, rpw_b .420->.370, evidence = Hewitt futures-inflated /
Giron elite-field-deflated) -> overshoots to Giron 99.4%, margin +7.8, total 14.3.
The two sims bracket a huge range (Hewitt-fav <-> certain blowout); NEITHER is a
trustworthy margin/total anchor. Cannot pin a calibrated cover/total probability.
RECENT FORM / PATH (last 5 each)
Giron: W Washington F vs Ilagan(1580) 6-3 7-6; W Washington SF vs Shah(1408)
7-5 6-4; L Wimbledon vs Zverev(1880) 2-6 6-7 4-6; W Wim vs Halys(1617);
W Wim vs Moutet(1639). Match-sharp on hard, deep grass season behind him.
Hewitt: "W" Washington F vs Svajda(1778) retd; "W" SF vs Lajovic(1569) retd;
Boys Wimbledon junior run; before that ITF M15/M25 with LOSSES to sub-1500s.
True level = ITF futures / junior. Class gap is enormous on Giron's best surface.
FATIGUE / NARRATIVE
Both played back-to-back into 07-26. Giron 33 (recovers slowly) + won a
title/final YESTERDAY -> real LET-DOWN risk; a flat Giron still wins but may
grind (anti-cover, pushes total up). Hewitt 17, house-money -> can swing free
and force competitive games (over risk / anti -5.5). Fatigue is context only,
not an edge (calibration).
GATES
G1 edge-origin: FAIL — the only numerical model-vs-market gap points to backing
the JUNIOR (broken-model artifact), not a real edge; any alt edge exists only
after my hand-correction, which overshoots. No stable pre-narrative edge.
G2 consensus: FAIL — the default-trustworthy baseline sim leans OVER (59% over
19.5) and makes Hewitt favourite; only a hand-override flips it. No hard
independent evidence (H2H / live) supports an alt play.
G3 fatal-risk: FAIL — Giron let-down (title yesterday) + 33yo back-to-back +
junior free-swinging variance directly negate a tight -5.5 cover or a short
total. Units can't buy back a negating risk.
PRICING
Giron ML @ Betfair 1.09: win ~0.90 -> min_odds 1.05/0.90 = 1.167 > 1.09 =>
automatic PASS, no ML value at such a short price.
Handicap Giron -5.5 @ 2.14 is high-variance (|line|>4.5, needs edge >=10%,
min 1.10/p); no reliable anchor + let-down/variance => no confident 58% cover.
Total 20.0: no trustworthy anchor, genuine over-risk from flat-favourite grind.
VERDICT: PASS. Giron is the clear winner (~90%) but the market already prices him
correctly (1.09, no ML value), and the ONLY model disagreement is a broken-sim /
farmed-rating artifact favouring the junior. No calibrated margin/total anchor and
active let-down + junior-variance negating risks. Nothing to stake.