So how are the ratings calculated?
We have an entire page devoted to discussion of that topic.
What games are included in your data?
Any quiz bowl game featuring two teams competing against each other, provided that I know about it, is included. Of course, I don't have knowledge of every match that takes place in the entire country, so your assistance in passing this information on to me is always appreciated.
What types of quiz bowl games are not included?
Any of the following:
* While it is true that these games could be set up as a series of pairwise comparisons, there are many circumstances in which winning is all that matters, so the second place team may not actually be better than the third or fourth place team. Thus, it is easier just to ignore these results.
Where do you get your data?
Several tournaments post their full results on the internet, and for that, I am very grateful--this project would not be possible without such help. If you have results for your team or tournament that I have not included, please feel free to e-mail me (dbykowski@aiquizbowl.com) with the appropriate information.
What information do I need to send you regarding results?
Since this rating system is based on a series of pairwise comparisons, I need to know who won and who lost each match that occurred. Margin of victory is not currently used, so the scores of the individual games would be helpful for aesthetic purposes, but they are not necessary. Indications of which matches took place during which rounds would also be helpful, but again, they are not required.
How often are the ratings updated?
The ratings are automatically recalculated nightly around 1:00 AM based on the entirety of the data entered in the database of games.
Is your rating system perfect?
Absolutely not. I played with it as an idea early in the 2006-07 season and kept it to myself to see how it would turn out by the end of the season. After a full season, I decided to release it to the public for discussion. Certainly improvements can be made and tested, and I'm interested in the ideas people have to offer.
What weaknesses do you currently see in your system?
Probably the most glaring weakness is how teams who play very few games and who win single-elimination tournaments end up being overrated. For example, in 2006-07, Gladstone High School won WNMU's High School Bowl television tournament and, to my knowledge, did not compete anywhere else that season. They ended up being ranked in the top 40 nationally, which is probably a massive overstatement. Also in 2006-07, Decatur Lutheran, until I included a known loss at a tournament, they were the #1 team in the country for much of the season because they were 10-1 in the IHSA State Championship series. The real problem comes down to teams that are not as well-connected to the rest of the circuit, whether because of actual lack of competition or a lack of results on my end.
How well does this system predict future results?
It does a fairly good job. You'll see that the ratings include a retro prediction percentage that, to a degree, reflects this information. In many ways, these ratings may be good for setting up relatively balanced preliminary round brackets for tournaments. However, I most definitely do NOT encourage using thes for gambling--the ratings are for entertainment purposes only.
How can these ratings be fair? The rosters of teams change all the time.
The same can be said, perhaps to a lesser degree, of sports teams, yet that argument is rarely brought up. It's often helpful to know the roster composition of a school's team so that, if 3 starters on an A team are absent and are replaced with B team members, a judgment can be made if that team's results should be grouped with the B team rather than the A team. Again, it isn't a perfect system, but at least a reasonable approximation can be made.
What inspired you to do this?
I first got the idea while taking a detailed look at different rating systems for college football and thought about why the same couldn't be done for quiz bowl. So it's people like Kenneth Massey, David Rothman, Jeff Sagarin, and others who have inspired me to try this for myself.