The core signal here is the Elo gap: 1708 for Svajda against 1585 for Petrovic, a 123-point difference that in Challenger/ITF terms usually points to a clear quality edge in baseline shot-making and match management. This is the single most defensible reason to lean favorite, since it's built from a broader sample of results than any single-match snapshot.

Still, this is a soft, less-liquid Elo market (per the model notes), so the gap should be read as a useful prior, not a guarantee — ITF results are noisier and single matches swing more on the day.

Recent form actually tilts toward Petrovic, who is 6-4 in his last 10 vs Svajda's 4-6 — a modest but real signal that the challenger is playing with more confidence right now. Neither player has listed quality wins, so this form read is about consistency, not signature victories.

Workload adds a further wrinkle: Svajda has played three matches in the last two weeks versus just one for Petrovic, even though both are on one day of rest. That heavier recent workload can show up in the legs and focus over a best-of-three format, partially offsetting his rating advantage.

Svajda's own numbers show a 63% rate on service points won, a solid figure that suggests he can hold serve comfortably when it's working. His 35% return rate is unremarkable, meaning his path to winning likely runs through his own serve rather than dismantling Petrovic's.

Because there's no serve or return data for Petrovic, this comparison is incomplete — Svajda's serve strength is real, but we can't say how it stacks up against what Petrovic does on his own delivery.

The model rates Svajda at 67% to win, but the market (via 1.40 odds) is pricing him at 71% implied probability — meaning the market is more confident in the favorite than this model is. That gap produces a -6.3% expected value, a mathematically unfavorable price on the favorite.

Being the favorite here does not equal value: Svajda's Elo edge and serve numbers make him the more likely winner, but at these odds the bet does not clear the model's own bar. Treat this as a case where the pick and the price disagree, and remember that Elo-based edges in ITF markets remain unproven in practice.