TennisEdge's AI model finds no betting value in A. Michelsen vs A. Mannarino (Washington on 2026-07-29) — the market is efficient, with A. Michelsen favoured at 1.30.
═══ A. Michelsen (A) vs A. Mannarino (B) — Washington, hard ═══
MODELS vs MARKET
ELO A 73.4% | ML A 85.1% | SIM A 75.5%
Book A 1.30 / B 3.08 → normalized A 70.3%
BETFAIR (executable) A 1.42 / B 3.25 → implied A 70.4% / B 30.4%
Note: exchange prices A LONGER than the book (1.42 vs 1.30) — the 1.30 book quote is
the soft/stale side, so the brief's 5.2% "value_edge" (computed vs book) does not
survive the exchange test. Commission-adj A = 1+(0.42*0.95) = 1.399 → 71.5% implied.
Edge at executable price: SIM +4.1%, ELO +1.9%, ML +13.6% (ML is the lone outlier).
RECENT FORM (5 most recent each; classification used all fetched matches)
A. Michelsen (21, #42) — 6-match win streak, all hard:
07-28 Washington R32 W vs M. McDonald (1695) 6-4 7-5 DOM (borderline)
07-26 Bloomfield FINAL W vs J. Fearnley (1702) 7-6 6-3 GRIND (TB)
07-25 Bloomfield SF W vs D. Suresh (1666) 6-2 6-7 6-2 GRIND
07-24 Bloomfield QF W vs M. Lajal (1705) 6-3 6-4 DOM vs Strong
07-23 Bloomfield R16 W vs L. Harris (1749) 6-3 6-3 DOM vs Strong
Whole run vs 1513–1749 ELO (Challenger field). Earlier: 6-7 grass/clay stretch.
A. Mannarino (38, #45):
07-28 Washington R32 W vs L. Tien (1928) 7-5 6-4 DOM over ELITE
07-09 Newport R16 L vs A. Galarneau (1574) 3-6 6-3 6-7
07-08 Newport R32 W vs R. Bertola (1653) 6-2 6-1 DOM
07-02 Wimbledon R64 L vs A. De Minaur (1872) straight
06-30 Wimbledon R128 W vs T. Droguet (1639) 6-2 6-4 6-1 DOM
SHAPE / PATH / SHARPNESS (Step 1b)
Michelsen: STABLE — margins are NOT widening; two of the last four wins needed
tiebreaks (incl. a 7-6 in a Challenger final and 6-7 sets vs 1666/1513 opposition).
Winning, but winning ugly against a sub-1750 field. PATH: maximally match-sharp —
5 matches at Bloomfield Hills + 1 at Washington, US-hard conditions fully grooved.
Mannarino: SHARPENING — a 7-5 6-4 straight-sets DOM over a 1928-ELO opponent (Tien)
is the single best scoreline-vs-expectation result on either side of this card, and
it came yesterday on this surface at this venue. PATH: 1 match at venue, no rust
(played 07-28).
Read: the sharpening dog is on the OTHER side of the model consensus. That kills the
"confirm the favourite" story rather than supporting it.
FATIGUE (Step 2)
Michelsen: 6 matches / 18 sets / 671 min in 7d, back_to_back 4 (07-22→07-26 streak),
rest day 07-27, played 07-28. Extreme-spot test: yesterday's win was a 2-SET 6-4 7-5
(no 150min/3-set hangover → no -4.5pp haircut, per calibration a 2-set win yesterday
is ~neutral); today is only the 2nd match in 2 days, not the 3rd in 3.
→ No extreme spot. Cumulative 671 min is CONTEXT ONLY. Age 21 → recovers fast,
soften one tier → "roughly equal" for betting purposes.
Mannarino: 1 match / 99 min in 14d, fresh. Age 38 → slow recovery, but he has no load
to recover from.
Market check: A's exchange price (1.42) is already well off the book (1.30) — the
fatigue/heavy-schedule story is visibly PRICED. No unpriced fatigue edge either way.
NARRATIVE (Step 3)
TITLE: Bloomfield Hills 3d ago (hard) — Michelsen won it 07-26.
Let-down: POSSIBLE (Final won 3d ago), but partly discharged — he already backed it
up with a 07-28 win. Directional negative on his level, not a stakeable edge.
Venue momentum: no prior Washington history in window for either player.
Surface transition: NONE — both played hard yesterday.
ELO/RANKING SANITY
Michelsen #42 and ELO-favourite → broadly consistent, BUT his entire ELO/form gain in
this window was farmed on a 1513–1749 Challenger field, so both ELO and especially ML
are inflated by a streak against weak opposition. ML's 85.1% for a #42-vs-#45 matchup
is not credible against a 70.4% exchange and is discounted. Mannarino #45 with the
weaker ELO despite yesterday beating a 1928 — his ELO understates current hard-court
level; do NOT back A purely on the ELO gap.
MY NUMBER
Michelsen ~72% (between the exchange's 70.4% and ELO's 73.4%; ML discarded as
Challenger-contaminated; Mannarino's Elite-DOM shape argues against going higher).
A: 0.72 vs eff-implied 71.5% → +0.5% edge. min_odds = 1.05/0.72 = 1.458 > 1.42
executable → automatic PASS on the favourite.
B: ~28% vs eff 3.1375 → implied 31.9% → NEGATIVE edge. No dog bet.
Alt markets: handicap/totals/model all null — ML only, no margin expression available.
GATES
G1 edge-origin: FAIL — best pre-narrative model-vs-exchange gap is SIM +4.1% (ELO only
+1.9%), under the 5% bar at the executable 1.42; the 13.6% ML gap is a lone outlier
contaminated by a Challenger win streak, and the book's 1.30 is not the executable price.
G2 consensus: PASS (tier only) — ELO+SIM both with A → would be full-ladder tier, moot.
G3 fatal-risk: PASS (ML bet — never blocks) — would carry risk_flag: 671 min/7d load
plus 3-day-old title vs a dog who just DOM'd a 1928 ELO.
VERDICT — PASS
Exchange 70.4%, my 72%: the market has this right. The apparent "value" is an artifact
of a stale 1.30 book quote; on the price you can actually take, there is <1% edge on a
favourite whose form was built against 1500–1750 opposition, who has 671 minutes in his
legs, is 3 days off a title, and faces the sharpest dog scoreline on the card. Nothing
to stake on either side. No bet.