Short version: if you read only the points table, Group H and Group G look like coin flips. Two teams on 2 points, one team on 1, everyone "still alive." Our model disagrees — and it disagrees loudly. In Group H it gives Cape Verde a 69% chance to reach the Round of 32 and Uruguay just 44%, a 25-point gap between two teams on identical points. In Group G it puts Belgium at 87% and Iran at 70%, again level on the table. That gap is the whole point of running a simulation instead of squinting at a standings column.
Why equal points don't mean equal futures
The standings answer one question: what has already happened. The model answers a different one: given everything we know, what happens next across the 10,000 ways the rest of the tournament could play out.
Here's how it works, in plain terms. Every team carries an Elo strength rating blended with public forecasting consensus. For each remaining match those ratings set an expected number of goals per side; a Poisson model turns that into a full grid of scoreline probabilities; then all the remaining group and knockout games are simulated 10,000 times. Advance and title percentages are just how often a team survives across those runs. The whole thing is recomputed every hour and after every result.
So when two teams sit on 2 points, the simulation isn't asking "who has 2 points?" It's asking three things the table can't: who is stronger on the underlying rating, who plays the easier remaining opponent, and who is better placed on the tie-breakers — head-to-head first, then goal difference, then goals scored, exactly as the official 2026 rules specify.
Group H: Spain runs away, then it gets interesting
Spain is the headliner and the model treats them accordingly: 100% to advance from the group, and an 18.9% chance to win the whole tournament — second only to Argentina's 21.5%. That title number is worth pausing on. Spain's model Elo is 2134 against Argentina's 2144. A ten-point rating edge is almost nothing, yet it maps to a 2.6-point title gap, because title probability compounds rating advantages across six or seven knockout rounds. Small input, amplified output. That is the same amplification that separates the teams below Spain.
Underneath the leader, the table reads: Cape Verde and Uruguay both on 2 points, Saudi Arabia on 1. The model's verdict:
| Team | Points | Model advance % | |---|---|---| | Spain | 4 | 100% | | Cape Verde Islands | 2 | 69% | | Uruguay | 2 | 44% | | Saudi Arabia | 1 | 32% |
Two teams, 2 points each, 25 percentage points apart. Why does Cape Verde — on paper the romantic underdog of the group — sit so far ahead of a Uruguay side most fans would rank higher by reputation?
The answer is the third-placed race. With the top two of many groups effectively settled, the realistic route for both Cape Verde and Uruguay runs through the best-third bracket: 8 of 12 third-placed teams advance. There, goal difference and remaining-fixture strength do the heavy lifting, and that is exactly where the simulation's view of each team's profile starts to diverge even from a shared 2-point line. Cape Verde already sits 4th in the live best-thirds ranking at 69%; that ranking, not the group column, is the table that actually decides their summer. Reputation doesn't get a vote in a Poisson grid — the rating and the run distribution do, and across 10,000 of them Cape Verde simply survives more often.
Group G: Belgium, Iran, and a quieter version of the same story
Group G rhymes with H. Egypt leads at 100% to advance. Then Belgium and Iran are both on 2 points — and the model splits them 87% to 70% in Belgium's favour.
| Team | Points | Model advance % | |---|---|---| | Egypt | 4 | 100% | | Belgium | 2 | 87% | | Iran | 2 | 70% | | New Zealand | 1 | 13% |
Belgium's 87% is not just ahead of Iran — it is one of the strongest third-place survival figures in the entire field. In the live best-thirds standings Belgium ranks only 5th on results (2 points, goal difference level), yet carries 87%, comfortably above teams sitting above them on the ladder. That is the cleanest illustration of the model's core claim: where you sit on the results table and how likely you are to advance are two different measurements. Belgium's underlying strength and the shape of its remaining draw lift it well past what 2 points would suggest, while Iran's identical 2 points come with a less favourable projection.
The editor's read: the table tells you the score, the model tells you the odds of survival
If there's one idea to take from these two groups, it's this. A points table is a receipt — a record of matches already paid for. It cannot price the matches still to come. The moment qualification depends on the best-third bracket, the decisive currencies become goal difference and the strength of who you still have to play, and those are precisely the inputs a standings column throws away.
That's why Cape Verde at 69% sits above Uruguay at 44% on equal points, and why Belgium's 87% towers over a results-table position that flatters Iran. None of these are predictions carved in stone — they're model estimates that move every hour as real results land, injuries and line-ups the model never sees can swing a single match, and one upset can flip a third-place cut line. But that's the value on offer: not a reprint of the standings, but an honest read of where the remaining matches actually point.
Want to see it update live? The full advance and title numbers, plus the best-third cut line that decides Cape Verde's and Belgium's summer, are on our forecast page and refresh after every result.