Adrian Mannarino vs Jacob Fearnley — predicción

Mannarino holds a significant ranking advantage (No. 47 vs No. 159), which typically anchors long-term expectancy. However, the Elo picture complicates this: Fearnley's Elo (1917) exceeds Mannarino's (1807) by 110 points — a substantial margin that reflects recent tournament performance and rating momentum. This contradiction — better ranked but lower-rated — suggests Fearnley has been performing above his traditional standing, while Mannarino's ranking has not yet fully adjusted downward.

The model's 72% probability for Mannarino is built largely on ranking inertia and structural baseline (41% vs 38%). Yet this same probability sits well above the market's 40%, implying the model trusts ranking stability over recent Elo divergence. The 110-point Elo gap is substantial enough that one cannot dismiss Fearnley's claims; it reveals genuine recent form improvement, not noise.

Fearnley's recent record is stark: 9 wins in his last 10 matches, with only one loss. Mannarino alternates wins and losses (LWLLWLWLWL), a pattern that signals inconsistency and low confidence. Fearnley's quality wins are not explicitly listed, but his 9/10 run at an ATP Montreal final (reached 3 days ago) demonstrates he is competing at a high level right now. Mannarino's quality wins — Mensik and Tien, both Elo ~1990 — are respectable but do not offset the volatility.

This form disparity is the single clearest challenge to the model's 72% probability. Mannarino has not earned sustained confidence through results; Fearnley has.

Mannarino comes in refreshed: 6 days since last match, only 2 matches in the previous 14 days. Fearnley, by contrast, has played 6 matches in 14 days and reached the Montreal final three days ago — a deep run that commonly leaves physical and mental residue, even in victory. While fatigue is not a guarantee of decline, the recency and intensity of Fearnley's run (final at the host venue) adds legitimate depletion risk.

This is the primary 'against Fearnley' flag in the data. It does not negate his form edge, but it introduces a brake on his expected performance in a back-to-back scenario (final to earlier round).

On hard court, Mannarino's 42% win rate is 1 point above his 41% baseline, suggesting he holds court reasonably well. Fearnley's 30% hard-court rate is 8 points below his 38% baseline — a clear surface penalty. This gap aligns with hard court favoring aggressive, high-velocity play; Mannarino appears more suited to it, or Fearnley's style (likely more tempo-dependent) loses edge on the faster bounce.

Serve/return data shows near parity: Fearnley serves 64% vs Mannarino's 63%, and returns 42% vs 37%. The 1% serve edge and 5% return edge for Fearnley are marginal; they do not swing the match substantially. On hard court with these metrics, Mannarino's relative comfort (surface-specific) outweighs Fearnley's serve-return marginal gains.

The model projects 72% for Mannarino; the market prices him at 40% (odds 2.5). This 32-percentage-point gap yields +78.9% expected value — a substantial edge. However, this does NOT mean Mannarino is a lock. The model's accuracy is stated as ~65% out-of-sample, and this market is an ATP draw where upsets occur. Crucially, Fearnley's recent form (9/10) and Elo rating (1917) are real, not noise. The market may be overweighting his Montreal final run as a 'hot hand' signal; the model may be overweighting his lower ranking as structural truth.

Value exists at 2.5 odds for Mannarino *if* you believe ranking and baseline stability trump a 10-match form run. If you believe Fearnley's momentum and Elo improvement are genuine and fatigue-resistant, the odds are not favorable. The honest read: Mannarino is the ranked favorite with a structural edge and better rest, but Fearnley enters with superior recent form, a head-to-head win, and a Elo rating that says he is the stronger player *right now*. The 72% vs 40% gap is large; it assumes form reverts and rank reasserts. On a hard court with near-identical serve/return profiles, that assumption is testable and not certain.