About
Prob11 calculates and visually shows the probability of each team reaching its goals in the league — title, continental spots, relegation and more.
Where the data comes from
The entire league structure — countries, competitions, divisions, editions, rounds, fixtures and each edition's qualification rules — is entered and maintained manually by us. Scores are entered as matches finish, and every result immediately triggers a new calculation. We do not republish third-party tables: the model, the qualification zones and the tiebreakers are built edition by edition.
How the simulation works
For every remaining fixture we estimate win, draw and loss chances from both teams' strength: each side's record in the competition, combined with its pre-season ranking, which counts as ten extra games — early on the ranking decides almost everything, after ten rounds it carries half the weight and then fades gradually. We add the edition's home advantage. These weights were chosen by testing the model on thousands of matches from finished editions. With those chances we draw the rest of the season 10,000 times. In each simulation we build the full final table, apply the competition's tiebreakers — including head-to-head and the FIFA fair-play index where they apply — and record which zone each team finished in. The probability shown is the share of the 10,000 simulations in which the team reached that goal.
What each match is worth
Because each of the 10,000 simulations draws every remaining match, we can split them by the result of one specific match: the simulations where the team won, drew or lost it. Counting where the team finished in each group gives its chance of each objective depending on that result. That is what “What the next match is worth” shows on each team page and “Matches that matter most” on the edition page: each team’s next match, and only the zones where the result moves the chance by at least 1 percentage point.
About 100% and 0%
We only show 100% or 0% when the outcome is mathematically settled, proven by a deterministic check rather than by the simulations. A team that finished first in all 10,000 simulations but can still be caught is shown as “greater than 99.9%” — never as champion. That is the difference between very likely and certain, and it matters.
Known limitations
The model sees results, not football. It knows nothing about injuries, suspensions, managerial changes, fixture congestion from other competitions, end-of-season motivation or pitch conditions. Early in a season the sample is small and probabilities swing widely. Because they come from 10,000 random draws, differences of about one percentage point between two runs are statistical noise rather than information — which is why our written analysis only comments on swings above three percentage points.
How often we update
Probabilities are recomputed on every result entered. Each edition page shows when it was last updated, and the “what changed” block states explicitly which point of the round the analysis refers to.
The figures are statistical estimates, for informational and recreational purposes only. They are not advice and must not be used for betting decisions. We are not liable for any use of this information.
Are the forecasts right? See the calibrationContact
Visitor country detection uses the IP to Country database by DB-IP, under the CC BY 4.0 licence. db-ip.com