MLS Power Ranking Averages - Week 27

14 of 17 sites used, rankings do not include Wed. night’s games.

I was out of town again last week so the wk 26 rankings are on the table below. 14 of 17 sites updated this week. The rankings do not include last night’s shocking results! Big D Soccer will be dropped next week if there is not update.

As always, if you know of a site I have not used please let me know in the comments and I’ll look into adding it!

Sites Used

• Jeff Sagarin
• Five Thirty Eight (Proj. points)
• Who Scored
• ESPN
• Massey
• Soccer America
• MLS
• MLS Fantasy Boss - No ranking Wk 27
• Reddit MLS
• Dallas News
• Atlanta News Now (Abbr. AJC)
• Big D Soccer (FC Dallas blog, analytic prediction) - No ranking Wk 21-27 (Drop warning)
• Pro Soccer USA
• Cincinnati Soccer Talk
• Freeborg Power Index
• In On It
• US Soccer Players (Monthly only) - No ranking Wk 27

You may notice the list includes several ‘analytic’ sites; they serve to function as a stable counterbalance or ‘control group’ to the rest which are based more on personal opinions, and those done by ‘team’ blogs. There are also sites included that some will argue are ‘prediction rankings’ - but that also is a form of Power Ranking is (see the quote at the top).

All the averages use the Olympic method (I.E. Truncated Mean value) of dropping the highest and lowest scores, and then averaging the rest to reduce bias.

The 7 year weighted average is the averages that decrease in value over time. Thus 2019 will be worth 100%, 2018 = 90%, 2017 = 80% and so on down. Thus a bad year 4 years ago won’t drag down positive recent years. For expansion teams that don’t have long histories, they are given the overall league average for that year.

• Highest Standard Deviation: San Jose Earthquakes at 4.05
• Lowest Standard Deviation: Los Angeles FC 0.0
• Biggest Rise: Portland Timbers - Up 2
• Biggest Fall: New England Revolution, Seattle Sounders - Down 2

Site Bias

All sites are going to have bias, whether conscious (blogs) or due to analytic calculation (based on past performance). This section lets you see the extent that sites are biased for/against teams.

Most / Least Site Bias

These are the teams that each site is most biased for/against, as well as what team it has the least bias for/against.