Updated 10/23/2015
Belo are the MLS Elo Ratings devised by Tom Worville. The Elo Rating system itself was originally invented by Arpad Elo to rate chess players more effectively. It's not a perfect system for soccer, as the game is more fluid and random than chess, but it's a better indicator of under and over performance than basic stats alone provide. See below for a more detailed explanation.
Rank | Team | Week 34 ELO |
---|---|---|
1 | FC Dallas | 2081 |
2 | Los Angeles Galaxy | 2075 |
3 | New York Red Bulls | 2061 |
4 | Portland Timbers | 2056 |
5 | Columbus Crew | 2044 |
6 | Seattle Sounders | 2029 |
7 | Sporting Kansas City | 2016 |
8 | DC United | 2014 |
9 | Vancouver Whitecaps | 2014 |
10 | Montreal Impact | 2006 |
11 | New England Revolution | 2004 |
12 | Toronto FC | 2002 |
13 | San Jose Earthquakes | 1998 |
14 | Orlando City | 1997 |
15 | Real Salt Lake | 1994 |
16 | Houston Dynamo | 1971 |
17 | Philadelphia Union | 1944 |
18 | Colorado Rapids | 1937 |
19 | New York City | 1928 |
20 | Chicago Fire | 1882 |
Elo is a zero sum system, meaning that sides are allocated equal points depending on the outcome of a game. It rewards more points to teams who cause upsets - so when the Chicago Fire beat the Seattle Sounders they were awarded a large number of points, and the same number of points were taken away from Seattle.
Elo only takes into account a few factors. For a start, the initial Elo rating was created for each team by looking at their goal difference at the end of the 2014 season. These were then weighted depending on what the goal difference was. Philadelphia Union were used as the base for the Elo Ratings (allocated 2000 Elo points, which is average in this model) because they had a goal difference of 0 last season. It then only takes into account the final result of the match between two teams, and using some fancy math that is explained here the Elo rating is calculated.
Note that this is version one of the rating model, and does not take into account any factors such as team strength (probably part of the reason for that Fire win over the Sounders) randomness due to PDO, underlying performance using TSR or xG or even a home field advantage. These are all additions that are likely to be added and tweaked in the future - so stay tuned.
If you have any feedback or ideas about the model please reach out on Twitter.