F. Meligeni Alves vs P. A. Saraiva Dos Santos — prediction

The core of this projection is the Elo differential: 1774 for Meligeni Alves against 1612 for Saraiva Dos Santos, a 162-point gap that at ITF level typically separates a mid-200s ATP-ranked player from an unranked one — consistent with Meligeni's No. 194 ranking against an opponent with no listed ranking. This spread alone explains most of the model's 72% probability for the favorite; there is no surface, altitude, or head-to-head data to adjust it further.

With no head-to-head history and no surface percentages available, the rating gap is the single most load-bearing number in this projection, and it should be read as the base case rather than a near-certainty.

The service and return numbers cut in opposite directions and largely cancel out. Meligeni's 66% serve-points-won rate is 4 points above Saraiva's 62%, a real but modest edge on his own delivery. However, Saraiva's 44% return-points-won is 5 points above Meligeni's 39%, meaning Saraiva has historically been the more effective returner of the two.

Mechanically, this suggests Saraiva may generate more break chances against Meligeni's serve than the reverse is true, which tempers the rating-based favoritism — the shot-quality numbers are closer to a coin flip than the Elo gap implies.

Both players arrive on 4-match win streaks, and both won 7 or more of their last 10 matches (Meligeni 8-2, Saraiva 7-3) — a marginal recency edge for the favorite but not a decisive one. Fatigue is a wash: both players are one day removed from a semifinal at this same M25 Sao Paulo event and have played 4 matches in the past 14 days, so neither carries a rest advantage into this match.

At odds of 1.12, the market implies an 89% win probability for Meligeni Alves, well above the model's 72% Elo-based estimate — a -19.5% expected value. Even granting that this is a soft, less-liquid ITF market where any 'edge' is unproven, the gap here runs against the bet rather than for it.

Meligeni remains the more probable winner on rating and marginal form, but being the favorite is not the same as being the value play: on these numbers, backing him at this price is a bet against the model's own estimate, not with it.