Mini app · The big five leagues, match by match

What if football were played with Elo?

Today, in every league, a win is worth 3 points, a draw 1 and a defeat 0. It does not matter who you play or by how much: beating the leader is worth the same as beating the bottom side, and a 1-0 the same as a 5-0. Elo, the chess system, works differently: every club has a score that says how good it is, and after each match it goes up or down depending on how surprising the result was. Beating a big club earns a lot; beating a small one, little. Here you replay a real season with Elo and compare it with the real table.

Your rules

League
Season
Which table

The Elo table ranks clubs by how good they are now. The other ranks them by how much they improved or got worse since the season started.

What is this?

With Elo, a club that was already very good earns very little for beating the sides below it, because that was expected. If the table were only what each club gained this season, the top would be whoever surprised most, not the best: a big club winning everything it should would stay near zero. That is why the real table is the Elo one, and the other shows who was the surprise of the year.

Does last season count?

Decides whether each club starts the season with what it showed last year, or everyone starts from zero, as happens today.

What is this?

With Elo, what you earned is not wiped in August: the champion starts high and the club that barely survived starts low. That makes sense, because squads change little from one year to the next, but it also makes it hard for big clubs to fall even when they start badly. So at the start of each season part of that lead is trimmed and everyone moves towards the average. The more last season counts, the less is trimmed. If it does not count, everyone starts level, as in today’s table.

Do chances count?

Decides whether the side that played better and created more danger deserves something even without winning.

What is this?

Sometimes the side that did less wins: a deflection, a soft penalty, an inspired keeper. Chances are measured with expected goals (xG): how many goals each side “deserved” given the quality of its shots. If they count, the table rewards the dominant side a little even when luck is against it. They weigh less than goals, because in the end goals win matches.

Does playing at home count?

Decides whether winning at home deserves the same credit as winning away.

What is this?

In real football the home side wins quite a lot more often: its own pitch, its own crowd, no travel. If home advantage counts, a home win is treated as expected and earns less, and an away win earns more. “The usual” is the advantage real matches show. With “No”, a win anywhere is worth the same.

Do big wins count?

Decides whether a 5-0 proves more than a 1-0, or a win is a win whatever the score.

What is this?

Today a 1-0 and a 5-0 both earn 3 points. If big wins count, winning by a lot lifts a club more, because it shows more superiority. “Up to a point” stops a club that thrashes small sides from running away: each extra goal adds less than the one before, and a big win is never worth more than double a one-goal win.

What is this?

Decides whether what a club does in Europe says something about how good it is. European matches are the only time clubs from different leagues meet, so they are what shows which league is stronger. If they count, beating a big club from another country in the Champions League also lifts you in your league table, and losing in Europe drops you.

With these rules, Sunderland goes down. The real table had it 7th.

17 of 20 clubs change position. Arsenal is champion in both tables. Last season counts a lot. Also: West Ham United stays up (it was 18th).

The table · Premier League 2025-26

The top line marks the first 4; the bottom one, the last 3.

# Club Change Real table Rating Started
1Arsenal=185 pts17991690
2Aston Villa▲ 2465 pts16901617
3Manchester City▼ 1278 pts16801608
4Manchester United▼ 1371 pts16561571
5Liverpool=560 pts16271687
6AFC Bournemouth=657 pts16101569
7Newcastle United▲ 51249 pts15891589
8Nottingham Forest▲ 81644 pts15891560
9Brighton & Hove Albion▼ 1853 pts15821581
10Crystal Palace▲ 51545 pts15711575
11Chelsea▼ 11052 pts15681666
12Brentford▼ 3953 pts15631556
13Fulham▼ 21152 pts15501532
14Tottenham Hotspur▲ 31741 pts15411544
15Everton▼ 21349 pts15331537
16Leeds United▼ 21447 pts15001365
17West Ham United▲ 11839 pts14991494
18Sunderland▼ 11754 pts14891365
19Wolverhampton Wanderers▲ 12020 pts14111497
20Burnley▼ 11922 pts13481365

Tap a club to see its curve and the matches that moved it most.

How it is calculated

Every club has a rating. Before each match, the rating gap (plus home advantage) says how much of a favourite each side is. Afterwards, the winner takes points from the loser: many if nobody expected it, few if it was the expected result. A draw costs the favourite points.

The result is not just win, draw or loss: a big win multiplies the points, each extra goal is worth less than the one before and, with the cap, never more than double. xG, if it counts, weighs less than goals.

The starting rating comes from 2024-25. At the start of each season, part of it goes back to the league average, like the soft reset of ranked video games; promoted clubs start with the average of those who went down.

Leagues are linked through the Champions League, Europa League and Conference League: that is why a European match can move the league table. A home advantage of 50 points is the one that best predicts the 2024-25 matches.

“Only what it gained this season” is the sum of what each club won and lost in the season. It rewards whoever beat expectations most, not the best team: it is the table of this year’s Leicester.

Frequently asked questions

What would happen to the Premier League 2025-26 with Elo?
With these rules, Sunderland goes down. The real table had it 7th. With the starting rules, 17 of 20 clubs change place compared with the real table, with 3 points for a win, 1 for a draw and 0 for a defeat.

How would Elo work in a football league?
Every club has a rating. Before each match the system works out the favourite, home advantage included, and afterwards the winner takes points from the loser: many if it was a surprise, few if it was expected. A big win multiplies the points up to a cap, and xG can add with less weight. The table is the rating.

Why can the table not be the sum of Elo points from zero?
Because it would reward whoever beats expectations most, not the best team: a big club winning as expected would stay near zero and every year’s champion would be that season’s Leicester. That is why the table is the rating, with part of it pulled back to the average at the start of each season.

What is the soft reset?
At the start of the season, part of each club’s rating goes back to its league average, as in ranked video games. That way the champion does not start with a huge lead and the promoted side is not hopeless. How much goes back is a choice, and in the mini app it is yours.

Data: league and European cup matches from 2024-25 onwards. Calculation: La Lente by Sir Balone, edited by Juan Carlos Vasquez. An exercise: no league uses this table.