Skip to main content
ABC News
What’s New In Our 2017-18 Club Soccer Predictions

This methodology article is for an old version of our club soccer forecasts. See how our latest club soccer predictions work.

We’ve launched a big expansion of our soccer predictions here at FiveThirtyEight. We’re now forecasting 24 club soccer leagues, with two more — Champions League and Europa League — to come in a few weeks. We’ve added leagues from South America (Brasileirão and the Argentine Superliga), along with 14 new European leagues, including five second-tier leagues, such as the English Championship.

For the most part, the methodology behind our forecasts is the same as last year’s. We’re still using four metrics from each match — goals scored, adjusted goals, shot-based expected goals and non-shot expected goals — to evaluate team performances. Those evaluations are expressed as offensive and defensive ratings for each team. And those ratings, in turn, let us calculate win/loss/draw probabilities for future matches and simulate the season thousands of times to estimate each team’s chances of winning the title.

All our significant changes are in how we assess the relative strength of domestic leagues. The goal was to improve our forecasts for the Champions League and Europa League and to better compare clubs in different countries — say, Juventus in Italy to Ajax in the Netherlands. We’re using recent matches played between teams from different leagues, supplemented with league market values (from Transfermarkt), to assign a strength rating to every league that we’re forecasting. Our new league ratings also give us the ability to a calculate a global Soccer Power Index (SPI) rating for each team — a number from 0 to 100 that represents the overall strength of each team.

We’ll get into more detail about that methodology below, but to start, here are our league strengths for most of the top-tier domestic leagues in Europe, North America and South America:

Relative strength of club soccer leagues from Europe, North America and South America, by country

Based on matches played in the past five years and the current market value of each league

Spain UEFA 2.01 2.10 2.01
Germany UEFA 1.89 1.88 1.89
England UEFA 1.58 2.28 1.60
Italy UEFA 1.48 1.89 1.50
Brazil CONMEBOL 1.48 1.64 1.49
France UEFA 1.51 1.01 1.49
Russia UEFA 1.49 1.27 1.48
Argentina CONMEBOL 1.48 0.71 1.45
Portugal UEFA 1.18 1.16 1.18
Turkey UEFA 1.11 1.16 1.12
Mexico CONCACAF 1.12 0.93 1.09
Ukraine UEFA 1.10 0.88 1.09
Switzerland UEFA 1.07 0.75 1.06
Colombia CONMEBOL 1.02 0.46 0.99
Belgium UEFA 0.97 1.00 0.97
Paraguay CONMEBOL 0.90 0.90
Austria UEFA 0.86 0.59 0.84
Greece UEFA 0.81 0.62 0.80
Ecuador CONMEBOL 0.77 0.46 0.75
Netherlands UEFA 0.71 1.05 0.73
Romania UEFA 0.75 0.24 0.71
Chile CONMEBOL 0.74 0.18 0.71
Sweden UEFA 0.73 0.29 0.70
Czech Republic UEFA 0.71 0.44 0.69
Poland UEFA 0.63 0.21 0.60
Uruguay CONMEBOL 0.60 0.15 0.58
Denmark UEFA 0.59 0.38 0.57
USA CONCACAF 0.54 0.66 0.56
Croatia UEFA 0.55 0.49 0.55
Bolivia CONMEBOL 0.53 0.53
Norway UEFA 0.50 0.19 0.48
Kazakhstan UEFA 0.40 0.27 0.39
Israel UEFA 0.22 0.23 0.22
Scotland UEFA 0.14 0.44 0.17
Slovakia UEFA 0.19 -0.16 0.15
Belarus UEFA 0.15 0.02 0.14
Panama CONCACAF 0.07 0.07
Cyprus UEFA 0.04 -0.30 0.02
Costa Rica CONCACAF 0.09 -0.30 0.02
Azerbaijan UEFA 0.01 -0.04 0.01
Peru CONMEBOL -0.06 -0.03 -0.06
Serbia UEFA -0.11 0.12 -0.09
Venezuela CONMEBOL -0.09 -0.11 -0.09
Bulgaria UEFA -0.15 0.27 -0.10
Slovenia UEFA -0.16 0.01 -0.14
Hungary UEFA -0.36 0.11 -0.29
Iceland UEFA -0.31 -0.69 -0.36
Ireland UEFA -0.28 -1.07 -0.40
Finland UEFA -0.45 -0.38 -0.44
El Salvador CONCACAF -0.46 -0.46
Bosnia UEFA -0.61 -0.03 -0.51
Honduras CONCACAF -0.61 -0.61
Guatemala CONCACAF -0.64 -0.64
Moldova UEFA -0.63 -0.88 -0.67
Albania UEFA -0.77 -0.30 -0.70
Georgia UEFA -0.79 -0.30 -0.71
Macedonia UEFA -0.74 -0.88 -0.76
Montenegro UEFA -0.99 -0.43 -0.87
Armenia UEFA -1.05 -0.42 -0.92
Latvia UEFA -1.46 -0.35 -1.27
Luxembourg UEFA -1.41 -0.84 -1.29
Malta UEFA -1.65 -0.63 -1.44
Wales UEFA -1.60 -0.88 -1.46
Estonia UEFA -1.67 -0.52 -1.47
Northern Ireland UEFA -1.63 -0.88 -1.48
Faroe Islands UEFA -1.70 -0.88 -1.54
Lithuania UEFA -2.08 -0.43 -1.69
Andorra UEFA -2.79 -0.88 -2.25

Not all leagues have market values on Transfermarkt.

Sources: ESPN, Transfermarkt

There aren’t too many surprises at the top; out of the five biggest European leagues, four are in the top five, with La Liga in Spain and the Bundesliga in Germany pretty far ahead of the pack. The Premier League in England, despite being the most valuable league in the world, has struggled to compete in the Champions League recently, and their match-based rating lags far behind their market value rating. Another league whose recent performance according to our ratings has lagged behind their market value is Major League Soccer in the U.S., whose last CONCACAF Champions League title was in 2000. MLS shows up in 28th place, between the Danish and Croatian leagues and a long way behind their regional rivals Liga MX in Mexico.

To generate these league strength ratings, we’ve set up a system where we first assume that all leagues are of equal strength and determine how far above or below expectation each league has performed over the past five years. In order, we:

  1. Run through all domestic matches in history and calculate domestic team SPI ratings throughout time.
  2. Look at each inter-league match from the past five years and calculate the expected score of the match based purely on each team’s domestic rating at the time.
  3. Take the difference between our expected score of the match and the actual score and run these results through Massey’s Method to find a rating for each league, expressed in how many goals better or worse than average that league is.
  4. Regress these calculated ratings toward market-value based ratings, weighted by how many inter-league matches we have for each league.
  5. Run through all matches in history again, incorporating league strengths into the predictions for any inter-league matches to improve the final team ratings.

Our league strengths can be interpreted as a bonus (in goals) given to each team in an inter-league match. So, for example, if Real Madrid (league strength of 2.01) were playing PSG (league strength of 1.48) and Real Madrid were a 0.2 goal favorite based only on their domestic SPI ratings, our model would give Real Madrid an extra 0.53 goal bonus because of the difference in the two teams’ league strengths.

A nice feature of these league strength ratings is that they let us generate global SPI ratings for any club team in the world. These global SPI ratings are a combination of each team’s domestic SPI rating and the strength of the league they play in. To generate them, we set up a mock match on neutral ground against a team with a domestic offensive and defensive SPI rating of 1.35 that plays in a league with strength of 1.25. (These values are just arbitrary ratings that we use for a baseline team against which we can compare any other team.) We calculate the number of goals we expect each team to score in the match as well as the chances of each team winning. So each team’s global SPI ratings can be interpreted as follows:

Offensive SPI: the number of goals the team is expected to score in such a match
Defensive SPI: the number of goals the team is expected to concede in such a match
Overall SPI: the percentage of points the team is expected to take in such a match

We’ll be adding a list of all SPI-rated teams to our club soccer predictions soon. For now, you can find forecasts from 24 leagues. They’re available in three languages (English, Spanish and Portuguese) and will be updated after every match.

Jay Boice was a computational journalist for FiveThirtyEight.