Monday, October 5, 2009

Rankings Explained

There are as many different ranking systems out there as there are opinions about what which team should be at the top. This has ceratinly come to light as my rankings have found there way onto various message boards. Without context, any rating system leads to questions. This post is attempts to describe my own sensiblity about what matters most in ranking teams relative to each other.

I based my system on the KRACH system that Ken Butler developed for college hockey. This is a retrodictive (backwards looking) system that does a great job of ranking teams based on their actual performance against each other, creating a network of opponents' opponents. From that web site: "The key to understanding KRACH is understanding that it's calculated recursively, so that the end result is self-evident by the results. In other words, if you took one team's schedule to date, and played a theoretical "game" for each game already actually played, using the KRACH ratings themselves in order to predict the winner, then the end result would be a theoretical won-loss percentage that matches the team's actual won-loss percentage."

What concerns me about the KRACH system is the absolute nature of the win or loss (or tie) inputs. If team A beats team B by a score of 30 to 3 and team A beats team C by a score of 10 to 7, KRACH would award teams A 100% of both wins. I feel that team C deserves more credit than that. They were able to score more points against A as well as keep A from scoring as many points against them. Therefore, my adaptation of the KRACH system considers the margin of victory. To paraphase the above quote, taking one team's schedule to date and playing a theoretical game against each team already encountered, using the Boggo ratings instead of scores in the calculation, the result would be a theoretical MOV equal to the the actual sum of MOVs.

After creating the ratings, I then rank order the teams to produce the rankings seen on this site. The whisker charts, add recently, use the ratings to demonstrate the relative strength of conferences. The tails show the extremes, the boxes represent the middle 50% withing the conference and the vertical line is the median rating.

Please, feel free to post any questions or comments you might have on this way of ranking teams.

No comments:

Post a Comment

Post a Comment