Premier League Table Predictor: How a Simulated Final Table Is Built (and How to Read It)
How a Premier League table predictor turns match probabilities into a projected final table: simulating the remaining fixtures thousands of times, what the percentages mean, and where the method is weak.
Updated 2026-11-19 · PunditRival
Short answer
A table predictor does not guess one final table. It takes a win/draw/loss probability for every remaining match, plays the whole rest of the season out thousands of times with random results drawn from those probabilities, and then counts how often each club finishes in each position. The result is a range, such as a title chance, a top-four chance and a relegation chance, not a single order.
The steps
First, every unplayed fixture gets probabilities for home win, draw and away win from a goal model. Second, the points already won are the starting table. Third, each simulated season draws a result for every remaining match, adds the points, and applies the league tiebreakers. Repeat that many times: ours uses 10,000 simulations. Fourth, the share of simulations in which a club finishes first, in the top four or in the bottom three becomes its percentage.
The average finishing position and the average points total come from the same simulations. Because every run uses the same fixtures, strong teams with easy runs rise and teams with several tough away games fall, which a simple points-per-game extrapolation misses.
How to read the percentages
A 30% title chance means the club won the league in about three of every ten simulated seasons, not that it is expected to win. A club on 2% still wins in about one season in fifty. Close races give flat percentages across several clubs, and those numbers move a lot after a single result.
The table is a statement about uncertainty. Treat the average points as the centre of a range and the percentages as the width of it.
Where the method is weak
A predicted table is only as good as the match probabilities behind it. Injuries, managerial changes and transfers can change a team's strength mid-season, and a model that rates teams from past results reacts to them slowly. The simulation also assumes each match is independent, which ignores momentum and fixture congestion.
That is why the betting market is a useful cross-check: where the model and the margin-free market price disagree sharply, the disagreement is worth reading before trusting either. Our own testing shows the market was slightly more accurate than our model on individual matches, and the table inherits that limit.
See it for the current season
Our predicted Premier League table is rebuilt from the latest fixtures and results, with the title, top-four and relegation percentages for every club. It is a projection for information, not advice or a guarantee of any outcome.
Questions people ask
What is a Premier League table predictor?+
A tool that projects the final league table by simulating the remaining fixtures many times from match probabilities, then reporting how often each club finishes in each position.
How many simulations does PunditRival run?+
10,000 simulations of the rest of the season, rebuilt as fixtures and results change.
Does a high title percentage mean a club will win the league?+
No. It means the club won the title in that share of simulated seasons. Results in a real season can and do differ from the most likely path.
Why do the percentages change after one match?+
Every result changes the points base and the strength estimates for the remaining fixtures, and in close races a single result can shift many clubs' chances at once.
