Live World Cup Forecast
Tournament-win and advancement probabilities, updated automatically after every match. 5,000 Monte Carlo simulations per cycle.
Biggest movers since last match
Projected knockout bracket, round by round. Each slot shows the team most likely to fill it. Deeper rounds are inherently less certain — a projected finalist might appear in only ~15% of simulations. For "who reaches the final," the Likely Finals tab is a better guide. Once group stage ends, the early rounds become the actual bracket.
The 10 most common Final pairings across 5,000 simulations, plus how often each team wins when they meet.
Each team's chance of reaching every stage, by any route. Search or scroll, then tap a team to expand.
5,000 simulations · FIFA Elo (April 2026) · Official 2026 bracket · @worldcupscorigami.bsky.social
How the forecaster works
Our probability numbers (the same numbers used in our Bluesky World Cup Scorigami bot) come from a Monte Carlo simulation that plays out the rest of the tournament 5,000 times after every result.
The foundation is FIFA's official Elo rankings from April 2026. Argentina enters the tournament at 1875, France at 1877, Spain at 1876, down to New Zealand at 1281. These ratings convert to match-by-match win probabilities through a standard Elo formula with a draw probability that peaks around 22% when teams are evenly matched and shrinks when there's a clear favorite. Knockout matches collapse the draw probability into the two sides proportionally, since extra time and penalties decide tied games.
Our simulator plays out the rest of the group stage match-by-match, sampling realistic scorelines and accumulating points and goal differential per FIFA's actual tiebreaker chain. The top two teams from each group advance automatically, and the eight best third-place teams across all 12 groups round out the Round of 32. The third-place pool is ranked by the same tiebreakers used within each group.
The Round of 32 is where most generic Monte Carlo models go wrong. The naive approach pairs the 32 qualifiers randomly, which lets strong teams accidentally avoid each other and inflates favorite probabilities by 30 to 50 percent. The bot instead uses FIFA's actual 2026 bracket, hardcoded from the December 2025 draw. Match 73 is always Group A runner-up vs Group B runner-up. Match 86 is always Group J winner vs Group H runner-up. From there through the final at MetLife, the bracket is fixed.
Each simulation runs the full sequence — group stage matches, group sort with tiebreakers, third-place pool, Round of 32 through Final — and produces one tournament outcome. It runs that 5,000 times, and the proportion of simulations in which Argentina wins is their tournament-win probability. It's the same process for each team's chance of advancing, and for each potential final pairing.
When the bot posts after a match, it runs the simulation twice: once with the pre-match standings and once with the new result included. The difference between the two is what produces the "↑3" or "↓21" deltas you see in the probability post. A scoreline like 4-0 will move advancement probabilities more than 1-0 because the tiebreaker logic correctly captures that goal difference and goals scored matter when group teams finish on equal points. Tournament-win probabilities barely move from individual scorelines, though, because a team's Elo doesn't update mid-tournament. That matches how bookmakers handle futures pricing.
The live forecast page reads the most recent snapshot from the bot's storage and refreshes once a minute. Every team's current probability is the same data the bot uses for its post-match threads.