Monday, April 26, 2010

Brew Plop expanded week 3 roto standings

I think this may actually be the table instead of just a picture of the table. Now you too can download and sort if you feel like.

Rotating League Name Week 2

I think I figured out why Shane wanted a running total of the weekly high scores. Congrats to him (and condolences to Rod Beck) on the first 12 category sweep in league history.



[link]http://lh4.ggpht.com/_yL5evi7KW4k/S8xZDZVKgsI/AAAAAAAAE0s/5er_gF-PJuU/s912/bestthrough2.JPG[/link]

I think my favorite discovery of this process is that the Van Super's have the best and worst in runs, while the Brooklyn's have the best and worst in wins, in only two weeks.


I also added this:


[link]http://lh4.ggpht.com/_yL5evi7KW4k/S8xeWLXnfmI/AAAAAAAAE00/T_ayJD21DeA/week2expectedwins.JPG[/link]

To get the expected wins, I summed the probability that you would win any particular category this week given the rest of the scores this week. For instance, if you had the highest amount of HR this week (13), you would win that category in 11 of 11 possible matchups.

What I see are two major standouts and a personal gripe. Shane was going to do well against anyone this week, but did even better than expected, while the Handlebars had an average week that netted nothing. The four worst teams this week played each other, so the Kids and Obamers got luckly to win some games. On the other hand, Brownie and I played each other and both came out with not so much to show for it.

Brew Plop through 2

Well, we mostly survived the week where half the MLBPA hit 4 HRs a piece.

The goal of today is to answer two questions: First, how would I have done against other teams.

Bam, drop an apple on his

[link]http://lh4.ggpht.com/_yL5evi7KW4k/S8xPtilJxkI/AAAAAAAAE0k/GWxZpLmLCn0/week2expectedwins.JPG[/link]

First, the methodology - I sorted through the weekly score and ranked each team based on how many teams they would beat (not tie) that week with that score (i.e., if you're tops, you get a 9, if you're tied with the second best score, you get a 7, if you're the lowest you get a zero). Added those scores up for all 12 categories then divided by 9. In essence, created a probability chart for each category.

There's quite a bit of rounding error, but I vagued that away (that is ignored it) by not accounting for ties at all.

It appears that most teams did about as well as they could expect. The former Tots and I both got a little hosed because we had two relatively strong teams going against each other. The Unicorns picked up a little something extra because their strengths and weaknesses largely match the Lefties, who got out muscled in a few categories.

There's more there that is interesting, but I won't force too many explanations on it.

The second question of the week is who have been the best and worst performers thus far. I'll try to update this list semi-regularly throughout the season. Burt appears all over this list, evidently due to his all or nothing strategy (and extreme lifestyle).




[link]http://lh6.ggpht.com/_yL5evi7KW4k/S8xPuI-B4jI/AAAAAAAAE0o/awGGoOPRnn4/s720/bestthrough2weeks.JPG[/link]