TennisEdge's AI model finds no betting value in A. Michelsen vs M. McDonald (Washington on 2026-07-28) — the market is efficient, with A. Michelsen favoured at 1.31.
═══ A. Michelsen (A) vs M. McDonald (B) — Washington, hard ═══
MODELS vs MARKET (Michelsen A)
ELO 0.752 | ML 0.763 | blended 0.738 | de-vigged market 0.734 (book 1.31)
→ All four estimates sit within ~3pts. Full model-market CONSENSUS on
Michelsen as a ~0.74-0.76 favourite. McDonald priced 3.62 (raw 0.276);
models give him ~0.24-0.26, i.e. the dog is if anything slightly
OVER-priced. No numerical gap on either side.
EXCHANGE / ALT MARKETS
betfair_ml returns 1.03 / 1.02 on BOTH sides — corrupt/stale, unusable, so
no executable exchange price exists (Pricing Discipline #4: ML then needs
>=8% edge vs book AND book-priced flag — not met, there is no edge at all).
alt_markets.handicap and totals are null — no margin market to price.
RECENT FORM / SHAPE (last 5 each)
Michelsen: just won Bloomfield Hills (hard) — 5-0 this week. Beat Harris
(1749) 6-3 6-3, Lajal (1705) 6-3 6-4 [DOM vs Strong], Suresh (1666) in 3,
Fearnley (1702) final 7-6 6-3. SHARPENING and MATCH-SHARP on the surface;
genuinely strong hard-court run against Strong-bracket fields.
McDonald: reached a hard final 07-26 (lost 4-6 4-6 to Svajda 1630), beat
Butvilas in 3 en route. Competent but streaky; older data labels garbled.
FATIGUE (Step 2)
Michelsen: 5 matches/5 days, back_to_back 4, 572min/7d — heavy load, but
last match 07-26 was a 2-SET win (7-6 6-3), 2 days ago, so NO 3-set
hangover haircut applies (toll is the extra-set-yesterday effect only).
Age 21 -> recovers fast, soften further. Context, not an edge.
McDonald: lighter load (226min), also last played 07-26. Roughly equal.
NARRATIVE
Michelsen LETDOWN — won a Final 2d ago (title). This is the one fade angle,
but the market is NOT fading him (still 1.31) and models agree at ~0.75, so
a let-down does not create a numerical edge; using it to back McDonald would
be a manufactured G1 edge. No surface transition (both on hard). Michelsen's
shape is SHARPENING, which cuts against the let-down fade anyway.
GATES
G1 edge-origin: FAIL — no model-market gap exists before any narrative
adjustment; a McDonald bet would be built purely on letdown/fatigue.
G2 consensus: FAIL — market AND all models sit together on Michelsen ~0.75.
G3 fatal-risk: N/A — no bet to protect.
VERDICT — PASS. Textbook consensus favourite: ELO, ML, blend and de-vigged
market all agree Michelsen ~0.74-0.76; the dog is fairly-to-over priced; the
exchange feed is corrupt and there are no alt markets to express a margin view.
The only contrarian angle (post-title letdown + court-load) is exactly the
over-priced fatigue narrative the framework warns against, and it is not even
being offered as value here. No articulable edge -> no stake.