The All-Knowing Derbytron is now on

This site will no longer be updated. Go to for all the fresh Derbytron content.


March 2 Analysis - The Bot is Back Edition

Is it finally roller derby season? The 24 WFTDA Sanctioned games played this weekend seem to think so.

I'm not going to get too much into the games played this weekend as DNN already did a better job than I would do, so just read that. I'll stick mostly to what's different this year and clear up some confusion some people seem to be having about the latest rankings.

During the offseason, I took some time and tried to figure out why the rankings screwed up the teams from the West region so bad last year. It was close in pretty much every aspect but the West. The first thing I changed was weighting recently played games heavier. A team like Rocky Mountain came on strong at the end of the season but that wasn't really reflected in the rankings because the games played earlier in the year (before picking up some standout skaters) were counting just as much. You won't see this factor come into play until later in the season as no one has played more than three games yet.

I also made the home track advantage a little wider. Last year, the home team won by an average of 16 points so the addition of points for the visiting team and subtraction of points from the home team's score got a little wider.

That's pretty much it. Not too many changes but I don't really feel like too many changes needed to be made.

Now, let's get to a couple questions that the Derbytron has received.

1. Where have you been?

Unlike some other rankings systems, the Derbytron isn't a constant stream of updates from one season to the next. I think everyone can agree that there is a massive drop-off in games played after nationals. There's also a large number of skaters that retire and new skaters that join teams. What that basically means is that teams could look very different on 1/1 than they did on 11/1 of the previous year. If the teams are very different, why should the Derbytron be using games from the year prior when the team had a completely different makeup? So, if you read the methodology, you would know that games are only counted from the last 6 months of the previous year and they're weighted much lower than the current year. So, taking time off gives teams a couple of months to get a couple games under their belts before these rankings are posted.

2. What's the deal with this latest update? This isn't even close. Oly in 7th?

So, since you've read the methodology and the rest of this post, you know that there aren't many games being factored into the rankings right now and any games this year are effectively weighted extra heavy because so few have been played. This is why these rankings are "unofficial." They're not going to ever make sense at the beginning of the season. Computer rankings systems need data and since the Derbytron views data from June 2008 and earlier as essentially incorrect, there isn't much to go on right now. This is how it is going to work so don't start freaking out about how Oly is ranked too low until the rankings are "official" which won't happen until the teams making up the rankings have played an average of 3 games in 2010.

No comments:

Post a Comment