Roman Safiullin vs Joao Fonseca: AI Prediction | Games, Spread, Aces & Double Faults

The model’s edge for Fonseca comes mainly from recent form and surface fit, supplemented by lower fatigue and a higher overall strength rating. Recent form contributes +5.6 percentage points toward Fonseca, and surface fit adds another +4.4; fatigue and overall strength add +3.7 and +3.3 respectively, with age providing an additional +3.2 in Fonseca’s favor. That pattern is reflected in the raw numbers: Fonseca carries a higher Elo (1872.16 vs 1774.13) and a much stronger tour ranking (No. 27 vs No. 132). Safiullin’s proprietary form index is actually higher (92.23) than Fonseca’s (72.22), but the model still leans to Fonseca because of the combination of Elo, freshness and surface alignment. Fatigue is a clear separator: Safiullin shows heavy minutes at this event (870 minutes) compared with Fonseca’s 348, which helps explain the workload factor pushing to Fonseca. Surface strength indices are similar (Safiullin 21.12, Fonseca 22.65), so neither player has a marked grass advantage by that metric. Serve and return profiles are closely matched — mean serve indices are almost identical (96.15 vs 96.32) and mean return indices are comparable (92.45 vs 87.61) — so there’s no single big serving mismatch to exploit. Recent match lines in the dataset show Fonseca arriving with straight-set wins in his last two Wimbledon matches, while Safiullin’s last three recorded matches at this event include two tight 3–2 scorelines and one longer encounter; the records indicate contrasting paths into this round.