Quote Originally Posted by ursus arctos horribilis View Post
That's the reply booty.
Thanks! I didn't actually expect an answer. Ha ha!

I figured it would be something like that. I believe he calculates a score for the offense, the defense, and overall for the team. (I'm sure there are many layers to developing each of those scores.) My guess is that he has those values for each year and applied them to his match up formula as he indicated.

The piece that may get a little tricky is the normalization process for those values across multiple years. I'm not arguing with his methodology at all, but my understanding is that each year the ratings are in relation to the other teams that year. I'm not clear how much validity the methodology has in comparing across multiple years.

I think it's a fascinating concept either way. It's possible that the exercise was simply for the fun of doing it without contemplating that challenge too deeply, or it's also possible that he has addressed the issue of comparing relative strength across multiple years and has found it to be a non-issue. It just seems like it would be hard to "prove," given that the only objective results to compare are those from games in the respective years they are played.

In other words, is he normalizing to the strength of the champion each year, with the understanding that some champions may be stronger than others -- but ignoring that for the sake of the exercise? Or is he normalizing to some other metric that would consider the relative strength between years somehow?

If you've got an open dialogue with him, it would be fun if he could address any of those questions as well. But again, I'm not expecting answers. No need to press. I'm just curious more than anything.