{"id":26105,"date":"2003-12-09T23:58:34","date_gmt":"2003-12-10T05:58:34","guid":{"rendered":"http:\/\/www.uscho.com\/2003\/12\/09\/between-the-lines-dec-9-2003\/"},"modified":"2010-08-17T19:55:34","modified_gmt":"2010-08-18T00:55:34","slug":"between-the-lines-dec-9-2003","status":"publish","type":"post","link":"https:\/\/wwwproxy.uscho.com\/2003\/12\/09\/between-the-lines-dec-9-2003\/","title":{"rendered":"Between the Lines, Dec. 9, 2003"},"content":{"rendered":"
Once again I’ll use football to make a hockey point.<\/p>\n
Have you followed the BCS controversy? USC is No. 1 in both major college football polls, but because of their computer rankings, falls below the top two spots in the BCS standings and therefore will not play in the national championship game. However, because the AP Poll winner will also be recognized as a national champion — if USC wins the Rose Bowl over Michigan there will be a split national champion, something college football was trying to prevent with the institution of the BCS.<\/p>\n
What a mess.<\/p>\n
\n<\/div>\n
The problem is, people are up in arms about the wrong thing. The fact that USC is No. 1 in both polls is not a good reason for saying they are being “screwed.” Why are the polls correct and the computers wrong? We don’t know which one is right and which one is wrong. Both systems are flawed.<\/p>\n
The BCS Standings are derived from an amalgam of the two polls, and seven different computer rankings. Having polls involved in the rankings is ridiculuos to begin with, if you really want to make it scientific. Pollsters are influenced by trends and what happened lately. Or maybe that’s a good thing. A mathematical algorithm takes the season as a whole. But if you wanted to build in a “late-season quotient” into the mathematical rankings, you could.<\/p>\n
Unfortunately, you have seven different computer rankings doing seven different things, and none of the methods are made public. This leads to the type of thing I heard this weekend, where a radio commentator was recommending that, in order to avoid USC getting shut out of the title game, the NCAA should “fudge the computer.” He suggested that, since the methodology isn’t public, no one would know the difference.<\/p>\n
Therein lies the whole reason why these things should remain public. I don’t believe the NCAA would do that — but just the fact that many people think they would, or should, is reason enough to keep the process transparent. Unfortunately, with two “tweaks” in the past two years to the RPI calculation, college hockey has decided not to do that. And that’s a shame in and of itself — never mind that most of us believe that college hockey should be using a better computation than RPI anyway.<\/p>\n
\nIt’s possible that Sean Fields will start to play better. And I’m sure Frantisek Skladany will finally get a few goals. But when I look at BU, I see a team with a lot of hard-working pluggers, and not really any scorers.\n<\/div>\n
Meanwhile, once again applying KRACH’s methodology to college football, we see that USC ranks seventh<\/em>. It is not even close, and has nothing to be complaining about. Their strength of schedule is far inferior to the other Top 10 teams, and the only Top 20 team that’s worse is Utah. (KRACH, for the unitiated, is a statistical model that does a far superior job of determining relative strength of schedule than any other system — without needing to resort to tweaks and tricks. Further explanation<\/a>)<\/p>\n