Skip to content
SimplePro
โ˜… New here? Start simple

THE SIMPLE VIEW

No spreadsheets, no jargon. Just the few things worth knowing about the 2026 World Cup right now โ€” who's favoured, where our model disagrees with the bookmakers, and which matches to watch. Hover any underlined wordLike this! Tap or hover any underlined term for a plain-English explanation. for what it means.

Who's going to win it?

Our model played the whole tournament out 10,000 timesWe simulate the entire 48-team World Cup ten thousand times to see how often each team lifts the trophy.. These are the favourites.

Where the model disagrees with the bookies

A model disagreementA spot where our model thinks something is MORE likely to happen than the bookmaker's odds suggest. Not a play and not an edge โ€” graded against the close, these have lost more often than they have won. Just a spot where the price looks generous. is where the bookmakers' price looks generous versus our model. Here are the biggest gaps right now โ€” explained in plain English.

Acadรฉmico de Viseu to winmodel disagreesOur model rates this more likely than the bookmakers do. We do not publish how much more: graded against the closing line, calls made this way beat it 26 times, tied 16 and lost 52.

The model makes Acadรฉmico de Viseu clear favourites to beat Vitรณria de Guimaraes.

Model: 48% likely  ยท  Best price 3.95 at Matchbook  ยท  2026-09-12
Real Madrid v Rayo Vallecano: a tight gamemodel disagreesOur model rates this more likely than the bookmakers do. We do not publish how much more: graded against the closing line, calls made this way beat it 26 times, tied 16 and lost 52.

The model expects a low-scoring game (under 2.5 goals total).

Model: 47% likely  ยท  Best price 3.75 at William Hill  ยท  2026-09-12
Sheffield United v Wolverhampton Wanderers: 3+ goalsmodel disagreesOur model rates this more likely than the bookmakers do. We do not publish how much more: graded against the closing line, calls made this way beat it 26 times, tied 16 and lost 52.

The model expects an open, high-scoring game (over 2.5 goals total).

Model: 78% likely  ยท  Best price 1.75 at William Hill  ยท  2026-09-13
RAAL La Louviรจre to winmodel disagreesOur model rates this more likely than the bookmakers do. We do not publish how much more: graded against the closing line, calls made this way beat it 26 times, tied 16 and lost 52.

The model makes RAAL La Louviรจre clear favourites to beat KV Kortrijk.

Model: 63% likely  ยท  Best price 2.48 at Grosvenor  ยท  2026-09-13
Neom SC v Al Fateh: a tight gamemodel disagreesOur model rates this more likely than the bookmakers do. We do not publish how much more: graded against the closing line, calls made this way beat it 26 times, tied 16 and lost 52.

The model expects a low-scoring game (under 2.5 goals total).

Model: 60% likely  ยท  Best price 2.30 at Tipico  ยท  2026-09-13

Matches to watch

What our model expects from the next batch of fixtures.

First up ยท Friday, September 11

Quick glossary

ModelOur maths engine. It estimates how likely each result is, based on team strength and recent form.
Probability / %How likely the model thinks something is. 70% means it expects it to happen about 7 times in 10.
Odds / priceWhat a bookmaker pays out. A price of 2.00 doubles your stake if it wins; bigger numbers mean less likely (in the bookmaker's eyes).
Model disagreementA spot where our model rates something more likely than the price implies. That is a difference of opinion, not an edge: graded against the closing line, our calls beat it 26 times, tied 16 and lost 52.
Edge (pp)How far apart the model and the market are, in percentage points. Bigger gap = stronger disagreement.
Over/Under 2.5A bet on the total goals in a match. "Over 2.5" wins on 3+ goals; "Under 2.5" wins on 0, 1 or 2.
For entertainment and information only. 21+. Model probabilities are estimates, not guarantees. Want the detail? Everything here is drawn from our full model pages. Ratings as of 2026-08-26.
For entertainment and informational purposes only. 21+. Please bet responsibly. Model probabilities are estimates, not guarantees. How the model works โ†’