2008 MLB Iterative Power Rankings, Week 8

Huh? While they're 2-3 since the last power rankings, they've outscored the opposition 18-11 over that span, and last night they destroyed the #4 ranked Braves 11-1. Dominating wins like that over highly-ranked teams will give your IPR a big boost.

Last 3 games, not 5 games. So, we're both wrong. :P

Over the last 3 games, they've outscored their opponents 12-8 (11+0+1)-(4+3+1).

I just thought that they got killed last night... not the Braves.

The Chiefs don't like you either.

Thanks, KD. I just put it up as a lockerroom because I generate it automatically... Since it doesn't really require a lot of work on my part (once I set up the spreadsheet at the beginning of the season), I'd feel guilty getting opinion credit for it.

It's worthy of votes. Even if it is just a formula for an automatic article, it creates endless debate over the exact degree of utter needlessness that is statistical data. Huzzah!

Technically, wins and losses are statistical data, Manny.

Technically, nothing is technical.

The SOS adjustment is made based on what's called the log5 formula, which tells you the odds of one team beating another team. Since we have the projected W% for each team as well as their strength of schedule, we can just use simple algebra to rearrange the log5 and tell us what the team's pW% would be against a .500 team.

So Tampa, Toronto, and Colorado are all in the top 5 in SOS (Minnesota and Pittsburgh are the other 2), and they get a boost in their ratings for playing tougher competition. For instance, Tampa has played like a .533 team but they've played a .533 schedule, so their real strength is .566. However, SOS only can take you so far... According to the log5 formula (which is just a sports-specific application of an odds ratio), a team like Atlanta (.640 pW% against a weak .471 SOS) would still be expected to defeat a true .500 opponent more often than somebody like Toronto (.548 pW% against a strong .530 SOS).

Whoops, clicked "post" too soon...

Anyway, it makes sense for Atlanta to rank ahead of Toronto because if you look at the difference between their pW%s (.640 vs. .548), that's a bigger difference than the disparity between their schedules (.471 vs. .530). If it ever got to a point where the SOS difference was bigger than the pW% difference, then Toronto would certainly rank higher. Of course, as the season goes on you're going to see the SOS numbers slowly move toward .500, because the sample is getting bigger and bigger (SOS not only includes your opponents, but also their opponents, and so on and so forth).

Toronto and Colorado suck, but you're right about Tampa being so awesome, KD...

Thanks for the response Davis, but next time could you do it in English?

What's nerdier? Davis' statistics or the fact that you can quote Xena; Princess Warrior?