How to read this. Voros sets the actual 3-way moneyline (it's the only model with a native draw). Elo and Fogle don't get blended in; they only vote on direction: a check means that lens agrees with Voros on the favorite, an X names the side it prefers instead. Because the three correlate 0.92 to 0.93, agreement is one shared signal, not three, so treat ALL 3 as "look harder" and SPLIT / VOROS ALONE as "stand down or dig." These are fair (no-vig) prices: to find real edge, de-vig the book on the match engine page and beat the closing line.
| Match | Grp | Voros favorite | Voros 3-way (H / D / A) | Elo | Fogle | Lens read | xG |
|---|
ALL 3 Elo & Fogle both agree with Voros
2 OF 3 one lens agrees
VOROS ALONE neither agrees (the distinct Voros call)
DRAW draw is the single most likely result
Method & cautions
Why independent lenses, not a blend
Voros, Elo, and Fogle are all trained on past results, so they correlate 0.92 to 0.93 on team strength (77 to 87% shared variance). Averaging them would mostly triple-count the same signal while hiding the disagreements that actually matter. So Voros prices the line and the other two only confirm or contest the favorite. When all three line up, it's one well-supported read, not three independent ones, and it may still be wrong about a lineup or injury the market already knows. When they split, the honest move is to wait.
What the numbers are (and are not)
- 90-minute result. Win / draw / loss for regulation, group stage. Not advancement, not knockout.
- Fair, no-vig. Voros prices are de-vigged by construction. They aren't edge until you compare them to a de-vigged book price (use the match engine page) and beat the close.
- Dixon-Coles ρ = -0.10 (the site default), which lifts the low-score draws. The DRAW-flagged games (draw is the modal result) are the natural draw-no-bet and Asian-handicap-0 spots.
- Team strength only. No lineups, injuries, or motivation. Voros's biggest solo calls (VOROS ALONE) often sit on CAF / AFC sides with the thinnest data; treat those as leads to dig into, not locks.