Skip to main content
Menu
What Would Happen If The English Premier League Had Playoffs?

The final weekend of the English Premier League is upon us and the league has scheduled all 10 matches simultaneously on Sunday in order to maximize excitement. The problem is that there’s not much excitement left — and there hasn’t been for awhile. Chelsea has been the runaway favorite to win the title for a few months now, and the team officially clinched a week ago. This isn’t uncommon. Even Leicester City’s shocking run to a league title last year, one of the most exciting outcomes in league history, was clinched two weeks before the end of the season. And in the 2014-15 season Chelsea clinched the title a full three weeks before their final match. Awarding the EPL trophy to the team atop the table at season’s end eliminates most of the debate about England’s best team, but it also can eliminate fun.

It doesn’t have to be this way. Imagine for a moment that after this weekend’s mostly meaningless matches,1 we were headed for the inaugural round of Premier League playoffs to decide the league champion. What would happen if more than a century of English soccer tradition was thrown out in favor of a shamelessly Americanized format?

As an experiment, I set up a mock six-team playoff for the Premier League and forecasted the results using our SPI ratings. There are a variety of playoff formats that could be used, but six teams feels right; it’s a small enough number that teams would have to compete to get in, but big enough that it would widen the pool of potential champions. Also, to reward teams that do well during the regular season, the top two teams in the table would be awarded byes. The next four teams would be seeded and would play one other team in a home-and-home aggregate, similar to the Champions League. From there the semifinals would follow the same format. Also similar to the Champions League, the final would be played as a single match at a neutral site such as Wembley.

If you applied that playoff set-up to this season’s standings, the playoffs would give us two more Manchester Derbies in the first-round and would pit Liverpool against Arsenal. Chelsea and Tottenham would be waiting to play the winners.

Chelsea would still be favored to win the playoffs — their first round bye and top seed helps quite a bit — but every team would have a reasonable shot at the title. Here’s how it would look:

CHANCES OF …
SEED TEAM MAKING SEMIS MAKING FINAL WINNING FINAL
1 Chelsea 70% 38%
2 Tottenham Hotspur 52 26
3 Manchester City 64% 34 18
4 Liverpool 58 19 8
6 Manchester United 36 14 6
5 Arsenal 42 11 4
Mock 2016-17 Premier League playoff forecast

Assumes the order of the top six teams will not change during the final weekend of matches. This playoff structure has six teams, with the top two seeds receiving byes.

Chelsea’s chances of winning the league would drop from 100 percent to 38 percent, and the five other teams that have been eliminated from the title race would still have hope.

I also set up playoff forecasts to see how our system might have affected the number of titles won by each Premier League club since 1992-93. Liverpool and Arsenal would likely have benefitted the most, while Manchester United likely would have won fewer than their league-leading 13 titles in that time span.

TITLES
TEAM ACTUAL EXPECTED UNDER PLAYOFF SYSTEM CHANGE
Liverpool 0 2.6 +2.6
Arsenal 3 4.8 +1.8
Newcastle 0 0.9 +0.9
Leeds United 0 0.6 +0.6
Tottenham Hotspur 0 0.6 +0.6
Aston Villa 0 0.4 +0.4
Everton 0 0.2 +0.2
Manchester City 2 2.1 +0.1
Southampton 0 0.1 +0.1
Nottingham Forest 0 0.1 +0.1
Wimbledon 0 0.0 0.0
Queens Park Rangers 0 0.0 0.0
West Ham United 0 0.0 0.0
Ipswich Town 0 0.0 0.0
Bolton 0 0.0 0.0
Norwich City 0 0.0 0.0
Blackburn 1 0.5 -0.5
Chelsea 5 4.4 -0.6
Leicester City 1 0.2 -0.8
Manchester United 13 7.5 -5.5
Impact of playoff system on league titles, 1992-93 through 2016-17

Includes all teams that would have qualified for the playoffs since 1992-93. This playoff structure has six teams, with the top two seeds receiving byes.

Sources: ESPN, James Curley

Playoffs, of course, inject luck into a championship but there are indications that a team still needs a significant amount of luck to win the league outright.

You can see how luck infiltrates league titles by looking at teams’ goal tallies. Goals scored and conceded will get a team to the top of the standings but they aren’t necessarily the best indicator of a team’s underlying talent, nor are they the most predictive of future performance. Expected goals, a metric we use in our SPI ratings, are a better gauge of each team’s quality of play. We can also use expected goals to approximate how likely each team was to win the league based on the shots it took over the course of the entire season.2

These expected goal simulations indicate that Manchester City have been unlucky not to win the title this year; if we rerun the season thousands of times, they win the league 61 percent of the time given the shots they took and conceded. Last year’s Leicester City team was fortunate to win the league according to these simulations; they only had a 9 percent chance of winning the league based on their expected goal numbers.

Finally, if we look at the projected number of titles won over the past seven seasons3 based on these expected goal simulations, the number of titles each team is projected to have won is quite similar to the number of titles we expect them to have won if the league had playoffs. In other words, this implies that there may be just as much luck involved in winning the league outright as there would be in a playoff system.

NUMBER OF TITLES
TEAM ACTUAL EXPECTED UNDER
PLAYOFF SYSTEM
EXPECTED UNDER
XG SIMULATIONS
Manchester City 2 2.0 2.6
Chelsea 2 1.5 1.3
Manchester United 2 1.2 0.7
Leicester City 1 0.2 0.1
Arsenal 0 1.1 1.4
Tottenham Hotspur 0 0.5 0.3
Liverpool 0 0.4 0.6
Everton 0 0.1 0.0
Southampton 0 0.1 0.1
Newcastle 0 0.0 0.0
Expected titles under playoff system vs. xG simulations, 2010-11 through 2016-17

Includes all teams that would have qualified for the playoffs since 2010-11.This playoff structure has six teams, with the top two seeds receiving byes.

Sources: ESPN, Opta

Players have incredibly busy schedules already with some juggling league and cup play, the Champions League and their international team responsibilities, so any playoff system probably shouldn’t add more matches to their already packed schedules. Perhaps the size of the league could be reduced by a team or two, or some regular season matchups could be a single match on neutral ground rather than home-and-away to allow for the extra playoff matches.

I don’t know if playoffs are the right thing for the Premier League. I do know that I haven’t watched many matches during the second half of the season, but I’d watch these playoffs.

Check out our latest soccer predictions.

 

Footnotes

  1. Fortunately, there’s still some uncertainty around the final Champions League spots.

  2. To summarize this process, every shot taken is assigned a probability of scoring based on the location it was taken from, the part of body used, the player taking the shot and a few other factors. We can roll up these individual shot probabilities into the number of goals each team was expected to score in a match based on their shots. Finally, we can use these match-by-match expected-goal totals to go back and replay each season thousands of times (a team’s simulated score in a match is drawn from a poisson distribution with a mean equal to their expected goal total for that match) to see how likely each team was to win the title

  3. We only have the play-by-play data needed to run these back to the 2010-11 season.

Jay Boice is a computational journalist for FiveThirtyEight.

Comments