Speaking as a fan, Montreal should probably finish ahead of Dallas, and potentially ahead of Vancouver
Man, so close. If Monreal had pulled off a draw they would have been ahead of Dallas. Vancouver rawing with RSL sealed them to finish ahead of the Impact and man, Colorado finishing the season with a statement... Actually, that little blurb right there was more or less my entire analysis this week. Amazing that San Jose draws and takes a big knock in SHFR points. They're more or less expected to win all the time... Crazy. No real movement in the top 10, just a bit of the bottom of the ladder figuring things out. Here's the final power ranking table Here's your last flavor of the week candidates And the final chart for the season. I thought about increasing the Y axis so you could see the early season fluctuations, but it makes an already difficult to read chart almost unusable... if you'd want to see it, let me know. I'm about to go through and pull out some team stats that I think are interesting, so expect this to be bumped a couple times more this week, but there you go. All in all this was more or less an experiment to see if it was possible to come up with Power Rankings that had some actual numbers behind them instead of the sort of subjective "XX is better than YY becauseI think ZZ" that other "experts" seem to use, and I think it worked out pretty well. Like I said, if you've got any questions, comments, thoughts, flames, or whatever, just shoot em my way, and thanks for letting me bother ya'll with this.
One pragmatic application I've used this system on was the Euro 2012 tournement this summer. I won't say it did exceedingly well... I had Spain over Germany in the final, but that's not too surprising. The system did predict the Dutch collapse though, so, yay? Anyway, if the results work out in line with the power rankings, your 2012 MLS Playoffs will look like this - Maybe calling a San Jose/Kansas City final isn't all that bold, but it might be a safe bet. DC over NY might be a gutsy call too, and am I actually predicting Seattle to win a playoff series?!?!?! Against RSL no less? Yes. Yes I am. Anyway, there. I said it. San Jose beats Kansas City at Buck Shaw (please don't move it MLS) for the 2012 MLS Cup title.
Here's some quick stats from the spreadsheet to kick out. Final rating isn't that interesting. What's sort of interesting is the ranking standard deviation. I actually expected Toronto to be the most consistent team in this category, but believe it or not, they only landed at #6. That little winning streak they put together kind of skewed them from being godawful terrible to a glimmer of hope before they went crashing down again, but it was enough to add some variability to their rankings. San Jose, not surprisingly, was the most consistent team in MLS, followed by New England, New York, Chicago and Houston. Toronto drops in at 6th. What it's interesting is the bottom four teams or so. Salt Lake City was the most inconsistent team this year, by sort of a lot. Montreal follows them with a Jekyl-and-Hyde type of season, and LA's terrible start and great finish land them honors as the third most inconsistent team of 2012. It's also interesting to look at the best weekly form numbers. 12 out of 19 teams, at some point in the season, turned in the best single week perforamnce. That's actually a little misleading. Single week performance counts for 50% of weekly rating, the other 50% being the previous four games (or I guess it could be three if a team plays two matches in a week). This might point towards the parity present in the league. Even poor Toronto clocked in a #4 performance over the course of the season on July 18th, after three wins in a row. Portland proved to be the most impotent team in the league having only ever risen as high as #5 (May 20th, after two draws and a win against Chicago). On the flipside, 9 out of 19 teams turned in dead last single week performances. The team that had the "Best" worst week? San Jose, dropping only as low as #14 on the season on May 19th after going L-D-D. Here's the full "season end stats" table. I dunno if there's all that much more in here to be honest, but I'll try doing some tweaking during the offseason and see where that gets me...