The headline gap here is structural: Cobolli sits at #10 with a 2043 Elo rating, while Fery is ranked #118 with a 1941 Elo. A 102-point Elo gap is significant and is the single biggest input pushing the model toward Cobolli at 77%. This is a case where the ranking differential is not just cosmetic — it reflects a real, sustained difference in level across a broad set of matches, not just recent form.

Both players arrive on 4-match win streaks, so the raw streak numbers look similar. The difference shows up in opposition quality: Cobolli's recent wins came against Auger-Aliassime (Elo 2069) and De Minaur (2044), both rated above his own 2043 mark — a sign he's beating players at or above his level. Fery's best recent wins, over Dimitrov (1924) and Bergs (1912), came against players below his own 1941 Elo, which is a weaker signal of current form even though the streak length matches.

The surface numbers alone slightly favor Cobolli (60% vs 57% on grass), but the relative jump tells a different story: Fery's grass performance is 11 points above his own baseline (46%→57%), while Cobolli's is only 2 points above his (58%→60%). That means grass is disproportionately helping Fery's game relative to his general level, which softens Cobolli's edge on this specific surface.

This is reinforced by the serve/return split: Fery's own serve (65%) and return (39%) rates are both higher than Cobolli's (60%/36%). Combined with hot, dry conditions (31°C, 34% humidity, negligible wind) that speed up the ball and typically reward the better server, this creates a real mechanical risk for Cobolli even though he is the stronger player by ranking and Elo.

The only previous meeting between these two went to Fery in 2026, though with just one match played this carries limited statistical weight against the much larger ranking and Elo gap. Rest is essentially even — both players are two days removed from their last match — with Cobolli having played one fewer match over the last two weeks (4 vs 5), a marginal freshness edge rather than a decisive one.

The model puts Cobolli at 77% versus a market-implied 71% at odds of 1.40, producing a modeled edge of 8.4%. That gap is real but not enormous — the market is already pricing Cobolli as a clear favorite, and the model's disagreement is a matter of degree, not direction.

Given the serve/return and surface-relative numbers that lean toward Fery, and the H2H loss, this is not a case of the market missing something obvious. Being the favorite here does not guarantee a win, and the edge should be treated as modest rather than a strong mispricing.