T. Sahtali vs K. Mabrouk — prediction
›Tour Elo: 1659 vs 1415 — favorite by rating
›ITF tier · 121 matches in the favorite's track record
›Elo estimate (not the ATP factor model): these are softer, less-analyzed markets
!Soft market: the value edge in Challenger/ITF is NOT proven live — treat it as an estimate, not an opportunity.
The 244-point Elo gap (1659 vs 1415) is the single clearest signal in this match, translating to an 80% model win probability for Sahtali. This is a Challenger/ITF Elo estimate, though, built on a softer, less-analyzed data pool than tour-level markets, so the edge should be read as directional rather than precise.
Neither player's recent form nor scheduling load offers a tiebreaker here. Sahtali is 6-4 in his last 10 with a 1-match streak, Mabrouk is also 6-4 with the same streak — recent results are a near mirror image.
Rest is identical too: both players are one day removed from their last match and have played three times in the past two weeks. Neither man carries a fatigue or freshness advantage into this one.
No surface, weather, altitude, serve/return, or ranking data is available for this match, so those mechanisms can't be assessed. The read here rests almost entirely on the Elo differential, with head-to-head history also absent.
At odds of 1.08, the market is pricing Sahtali at roughly 93% to win, well above the model's 80% estimate — producing a -13.3% expected value. Even though Sahtali is the clear favorite on rating, the market is asking you to pay more certainty than the model itself supports.
Being the favorite is not the same as being a value bet: here the gap between model and market points the other way, with no edge for backing the favorite at this price. Treat this as a soft-market estimate rather than a proven opportunity, and note the size of the discount before committing any stake.
Impact and analysis from real match data (Elo, form, head-to-head, rest, surface vs baseline, weather, altitude). Soft-market estimate: the value is unproven live. 18+ · gamble responsibly.