D. Sweeny vs A. Martin — prediction

The 134-point Elo separation (1791 for Sweeny vs 1657 for Martin) is the single largest signal in this match, translating directly into the model's 68% probability for the favorite. Sweeny's ranking sits at 127 with a positive trend of 55 spots, though without a ranking figure for Martin the comparison stops at Elo, which already captures the bulk of the class difference between the two.

This gap is meaningful but not overwhelming — a 134-point Elo difference at Challenger level typically produces exactly the kind of 65-70% win probability seen here, not a lopsided mismatch.

The service and return numbers are tight. Sweeny holds serve at 64% against Martin's 61%, and also returns marginally better, 38% to 40% is actually Martin's edge on return — so the split is nearly even, with Sweeny slightly ahead on serve and Martin slightly ahead on return. Neither number suggests a clear stylistic mismatch; this is a close, evenly-contested skill set rather than a lopsided matchup.

The clearest non-rating edge favors Sweeny: Martin has played six matches in the last 14 days and enters on just three days of rest, a workload that typically saps legs and serve percentage by the latter stages of a match. Sweeny, by contrast, has had 19 days off with zero matches in the last two weeks — full recovery, though also a longer break that can occasionally cost rhythm.

Both players enter on a losing streak of one match and identical 5-5 records over their last ten, so the fatigue disparity is the more concrete differentiator here than recent form trends.

The model prices Sweeny at 68%, while the market (odds of 1.43) implies 70% — a gap that produces a -2.2% expected value. In practical terms, the market is already pricing in Sweeny's Elo advantage and Martin's heavy recent workload slightly more aggressively than the model does, leaving no backable edge on the favorite at this price.

This is also an Elo-based Challenger estimate, a softer and less-scrutinized market than ATP tour-level pricing, so any edge — positive or negative — should be treated as a rough estimate rather than a proven opportunity.