Tuesday, June 01, 2010

Rotating League Name Week 7 recap

Since Shane wasn't the big winner this week, I decided to not keep the running tally of weekly bests and worsts. Ever. For this week. And last week.

What you see here is how a team can expect to finish a week given their past performance in each category. What I did was figure out how many weeks (out of 7) a team won, lost, or tied a particular category, then figured out your winning percentage for that category alone (using the same methods as ESPN, where T= .5W +.5L). What you see below is how each team did in each category. If a team won 60% or more of the time, they can expect to win that category each week. Less than 40% and they can expect to lose. Anywhere in between and it's a toss up from week to week.

I used this green color because it's very ugly.

Of note is that there are 4 categories that have been won 100% of the time. Hunt's Wins and Bartha's BB, SV, and K/9. Given this, it is a surprise that Bartha is not doing even better. Also, there's only 1 team that hasn't won a category and that's Corley's RBI, or rather, his lack there of. There's a whole lot of ugly for some people and it's never good when you have several "assumed loss" categories but few "assumed win" to go with it. The take away from this is that you can probably predict at the beginning of a week which categories a team will take and which are toss ups for that particular week.

Speaking of predictions, last week I used some method I don't remember (I think using YTD stats) to predict the outcome of each matchup. Here's how that went:

Well the prediction machine (notice how they are no longer my predictions, but the machine's) had a decent go of it. Only 1 upset (based on overall standings) was predicted. That matchup ended in a tie. Of the other 5 where the computer chalked, 3 were wins and 2 were upsets. As one of the upsets had 3 ties (and actually had the correct number of wins for Streeter's team), and the other upset was my team winning instead of losing, I'm fairly happy with the results. As expected, the results were only barely better than a coin flip.

Let me know if there's some kind of breakdown you want to do and I'll be happy to post it. Barring that, feel free to send along ideas for me to do.

No comments: