Again, I debated whether or not it was worth doing power rankings after midweek games but ultimately decided to do it since I wanted to introduce a new rating metric, which I'm calling the Super Happy Fun rating. It's actually basically the same rating number that I was using before, except now I take the inverse of that rating and mulitply it by a million, because it's fun to say 'Whoa, Seattle lost 3,000 points this week!' and so on and so on.
For a bit more explanation, what my model does is assign a rating each week (or each increment really as lately I've been doing midweek ratings) based on each teams form. Form is calculated half on results and half on goal differential, with the most recent game counting for 50% of the rating, the game after that 20%, the game after that 15%, the game after that 10% and the game after that 5%
The actual rating is each teams average weekly (or incrementally) form ranking. In the past, I just used that average and said lower was better (since 1 is better than 19), but now I decided to invert it and make it big. It does help with at a glance analysis and gives me a slightly better insight into what's happening. For example, I thought Kansas City's win last night would propel them higher, but it actually just evened out their form (from among the worst in the league to 8th) to their season long average, so not a whole lot of change there. In effect, we just "stopped the bleeding". Same deal with RSL and Seattle. They've both been playing poorly as of late relative to how they've played over the course of the season, so even their draw wasn't enough to stop them from dropping more points.
Here's how the table looks after the midweek games.
|Team||Rank||Weekly Δ||Super Happy Fun Rating||Weekly Δ|
|Salt Lake City||4||-1||125,000||-12,097|
As you can see, RSL dropped 12,097 points!!! Isn't that more fun than just saying they dropped a spot in the rankings?
Anyway, DC hops San Jose as the new #1 team, even though they're tied on points.