T. Legout vs J. Roberts — prediction

The 286-point Elo gap (1714 vs 1428) is the dominant signal in this match. At the Challenger/ITF level, gaps of this size typically reflect a meaningful quality difference, and it aligns with the model's 84% win probability for Legout. This isn't a marginal edge — it's a structural one, even accounting for the noisier nature of ITF-level rating data.

Legout's tracked numbers — 68% service points won and 44% return points won — describe a player who controls his own service games comfortably and also creates pressure on return. No equivalent serve or return figures exist for Roberts, so this factor can only be read as a positive marker for Legout rather than a direct comparative edge, but it reinforces the Elo-based favoritism.

The rest split is the one factor that complicates an otherwise clean picture. Roberts hasn't played in 34 days and enters with zero recent match load, meaning fresh legs but no recent competitive rhythm. Legout, by contrast, has played 5 matches in the last 14 days on just 2 days' rest — a workload that can accumulate physically over a tournament, though there's no data here confirming fatigue has yet become a factor.

This is a real countervailing signal worth weighing, even if the Elo and serve numbers still point clearly toward Legout.

Both players are on a 1-match losing streak, but the broader sample favors Legout: 6 wins in his last 10 versus 4 for Roberts. Neither player shows a listed quality win, so this is a modest, not decisive, tilt toward the favorite.

At odds of 1.06, the market prices Legout at a 94% implied probability, well above the model's 84% estimate — producing a -11.1% expected value. This is a case where being the clear favorite does not translate into betting value; the market is pricing in more certainty than the model supports.

Given that this is a soft Challenger/ITF Elo estimate rather than a full factor model, the edge should be treated as unproven rather than actionable. The honest read is that Legout is very likely to win, but the price offers no value at 1.06.