· locked in with the confirmed lineups
Mexico expects to create five times as many chances as South Africa, so a Mexico win comes out at 67 percent.
Tap any market for the model's reasoning, the numbers behind it and to track the pick. Fair price is the de overrounded model probability expressed as decimal odds, not a bookmaker quote.
The SharpCast Builder is a same-game multi built only from this fixture's own picks. Because the legs come from one match they are correlated, and the combined chance is priced over the projected scorelines rather than multiplied as if independent. The price shown is the model's fair price for that chance.
The same-game multi for this fixture, built from its own picks and priced for correlation, is a Pro feature.
The higher-odds build for this fixture, reaching for 20/1 or more from supported legs, is a Pro feature.
2.5 expected goals · most likely total 2 · a balanced game
Running on the league baseline, no team cards history yet. Treat as a guide, not a team specific read.
Appointment confirmed. SharpCast has no settled match history for this official yet, so the profile builds as they officiate. Card strictness is not yet part of the model.
This call rests on Dixon Coles strength, time decayed form, expected goals and the de overrounded market consensus, across 10,000 simulated matches. It was locked before kickoff and is kept on this site.
model probability · fair price
Each side's goals per game over recent matches (the bar) against the expected goals their chances were worth (the tick). Scoring past the tick means finishing well; conceding inside the tick means defending well.
Model probability reads against the norm for each position, not value bets: player card prices are not on the market here and the referee is not confirmed until close to kickoff. No price is shown. SharpCast Cards is tracked on its own separate record. Open a row for the reasoning.
Among the most booked defenders we rate in the 88th percentile; the model puts his card chance at 32 percent against Mofokeng and Appollis.
Largest sample of the South Africa defenders at 53 matches and rated well above norm; the model estimates 32 percent against Quiñones and Vega.
Sits 10 points above the defender baseline across 44 matches; the model makes it 32 percent.
Extended 11 points above his position baseline; the model rates his card at 31 percent.
Less established in the 67th percentile of defenders we rate; the model sets it at 29 percent.
Smallest sample among these defenders at 25 matches and modest percentile standing; the model puts it at 28 percent.