Menu
Log in


Week 11 MLS Meta Power Rankings - Cascadia Surge Edition

05/14/2013 11:19 AM | 107ist Admin (Administrator)

—by Mike Coleman

Note: you can click on any graphic to make it larger / more readable

Welcome to Week 11's MLS Meta Power Rankings, MMPL (tm). I take rankings from six different sources (Soccer By Ives, ESPN, Bleacher Report, Sporting News, MLS, and The Oregonian) and average them out. Additionally, I also look at the standard deviation between the movement for the week and the overall rankings (this to me shows how in agreement the 'experts' are in their rankings) - finally as another really basic measure of "power" I look points in the past 5 games.


Dallas continues to hold down the #1 spot universally across all six sources despite being gifted a PK to save 2 points against Portland (you realize you're reading this on the Timbers Army website, right?) and a not-so-convincing win on the road against the consensus worst team in the league, D.C. United.  The 'experts' also agree that the LA Galaxy are universally the 7th best team in the league (note their Std Deviation of 0.00).  The Timbers, according to these averages, are the 2nd best team in the league at this moment, their highest ranking of the season.

If we look at movement for the week clearly Cascadia's domination over the teams from California this week paid dividends with the top 3 teams in term of spots gained were Seattle (5.3), Portland (3.0), and Vancouver (2.8). On the other end of the spectrum Chivas (3.3), Columbus (3.2), and L.A. (3.0) fell the furthest.

In looking at the standard deviation in rank, it's apparent that the 'experts' are more in agreement this week than in the past. The outliers (New York, Montreal, and Philadelphia) aren't that far off from the rest of the league. For New York, their ranked #2 by ESPN and #6 by Soccer By Ives (they also have a two #5's, one #4,  and one #3) - clearly there is wide variety of opinions. Montreal is similar to New York in that they have a range from #2 to #6. Philadelphia benefits from a #8 ranking from Soccer by Ives when the rest of the group has them lower (one #9, three #11, and one #12).

In movement Seattle has the highest standard deviation in their group. Clearly everyone agreed they deserved a big jump (I begrudgingly agree, and admit to almost enjoying watching them kick the crap out of San Jose), experts just weren't sure exactly how much (the movement ranged from 3 spots to 7). To some extent this is a case of some folks not having them nearly as far down the rankings as others in the previous weeks (for instance Oregonian had them at #12 in week 9 and everyone else had them ranked #14 - #17).

Finally, I've complained the last two weeks about San Jose having a nearly league low points out of their past 5 games, but hanging around the middle of the rankings, so I'd be remiss if I didn't comment on Portland being ranked 6th as far as points earned in the past 5 games, but 2nd over all in the rankings. I think the 'experts' realize that Portland could (should) have had 2 more points out of Dallas (you do realize you're reading this on the Timbers Army website, right?). Of course it's not nearly as bad as SKC being near the top of the rankings with a whopping 6 points out of their past 5 games (I guess you get bonus rank points for snapping Houston's streak, and the Seattle loss could have easily been 3 points). And to avoid any (more) screams of homerism, I'll point out the huge disparity between Seattle's points in the past 5 vs. their ranking.

Below is a historical graph and table showing average power rankings from week 1 through week 11.


 


Comments

  • 08/08/2016 8:03 PM | 107ist Admin (Administrator)
    Ross says:
    Tuesday, May 14, 2013 at 11:10 am

    Is there any chance your historic graph could use patterned lines in addition to colors to help differentiate? Maybe have a Eastern and Western conference versions that would help us sort out the teams?
    Link  •  Reply
  • 08/08/2016 8:04 PM | 107ist Admin (Administrator)
    Mike Coleman says:
    Tuesday, May 14, 2013 at 11:16 am

    Ross –
    Thanks for reading. I have thought of conference based graphs, and working on the legend.
    I’m actually looking to see if there is an interactive plug-in for word press that might make the whole thing a lot more dynamic and readable.
    I’ll see what I can do to make it better.
    Link  •  Reply
  • 08/08/2016 8:04 PM | 107ist Admin (Administrator)
    Nels Hesseldahl says:
    Tuesday, May 14, 2013 at 12:24 pm

    I love that there’s now a connection between soccer and standard deviation. Good stuff Mike.
    Link  •  Reply
  • 08/08/2016 8:05 PM | 107ist Admin (Administrator)
    Matt Ryan says:
    Tuesday, May 14, 2013 at 12:30 pm

    Have you considered using any computer rankings or even the table (or points per game) as additional polls to include in your data? I was interested to see the Timbers #1 in both methods of Sagarin’s ratings.
    Link  •  Reply
  • 08/08/2016 8:05 PM | 107ist Admin (Administrator)
    Mike Coleman says:
    Tuesday, May 14, 2013 at 12:45 pm

    I hadn’t seen Sagarin’s stuff before.
    I’m not really trying to create anything too complicated. I basically read these six power rankings and wondered how they compared, so starting back in week 9 I began this exercise.
    The Sagarin rating puts the Timbers up top, and has SSFC at 3rd – pretty interesting. Thanks for pointing them out to me.
    Link  •  Reply
  • 08/08/2016 8:06 PM | 107ist Admin (Administrator)
    Matt Ryan says:
    Tuesday, May 14, 2013 at 12:51 pm

    That’s completely understandable. I’ve really enjoyed looking at the statistical breakdowns you’ve created here. And it does create a similarly-sourced set of data points to stick to subjective human observers that tells us something unique that we wouldn’t get by including the things I flippantly suggested. I already check the table, after all.
    Link  •  Reply


Powered by Wild Apricot Membership Software