C. O'Connell vs A. Martin — prediction

The Elo gap (1819 vs 1703) is the clearest structural advantage in this match, translating into a 66% win probability for O'Connell. This is reinforced by the serve numbers: he wins 72% of service points compared to Martin's 63%, a 9-point gap that should let him hold more comfortably over the course of the match.

Martin does have a return advantage, winning 39% of return points versus O'Connell's 35%, but a 4-point return edge is unlikely to offset a 9-point serve deficit. The mechanism favors O'Connell holding serve more often while also finding occasional break chances against a slightly weaker service game.

Martin arrives with just 1 day of rest after playing 9 matches in the last 14 days, including a quarterfinal run at Bloomfield Hills that ended only a day ago. This kind of workload compounds over best-of-three tennis, particularly affecting movement and serve power late in matches.

His recent form also shows some cracks: a LLWLWWWWWL run over his last 10 matches, with a current 1-match losing streak and no quality wins recorded. Combined with the physical load, this points to a player less likely to sustain a high level for the full match.

O'Connell is the favorite on both class (Elo) and in-match mechanics (serve edge, opponent fatigue), and the model puts him at 66% to win. However, the market is considerably more confident, pricing him at an implied 84% (odds of 1.19).

That gap produces an expected value of -21.4%, meaning the price does not compensate for the risk even though O'Connell is likely to win more often than not. Being the favorite is not the same as being a value bet here — on these numbers, the market has already priced in more certainty than the model supports.