Hello and greetings from the Packers State or Cheese State or Badger State or whatever Wisconsin’s nickname is. All jokes aside, this is a great football state and the Badgers were 30 minutes away from a major upset. They certainly did enough to drop the Buckeyes from the top perch to a date with Clemson. Last night, I predicted Ohio State, LSU, Clemson and Oklahoma and was only off by the top 2 teams’ placement. I could argue either side. Clemson ended up #1 in the GCR going into the Bowl season but that had more to do with winning by a few hundred points every game than playing strong competition. There are 33 teams that have 10 or more wins in Division I. Clemson played none of them. Virginia will have to beat Florida to get that number. Ohio State will be the first. LSU played 3 (Florida, Alabama, and Georgia) and if Auburn beats Minnesota that’s a 4th. Oklahoma adds to the list. Ohio State has 5 (potentially 6) with Florida Atlantic, Cincinnati, Wisconsin, Penn State, and Wisconsin again. The potential is Michigan if they beat Alabama. An interesting twist is Oklahoma is now the all time lowest GCR rank team that made the playoff. I have them at 20 mainly because they didn’t play a team that ended with a winning record until Oct 12. They lost to 8-4 Kansas State and had 2 games against 10-win teams (both of them were Baylor). LSU will be it’s first non-Baylor 10-win opponent.
Each year, I relook at the GCR formula to see what can be improved. I think this offseason, I may need to put a bigger governor on running up the score against mediocre teams. There’s already a descending value so that runaways don’t give too much credit, but I’m thinking there has to be a way to tie in the value to the quality of the team played. It has to be balanced: teams that are so dominant they get through the gauntlet unscathed should be rewarded. I’ll play with the math and see what improvements there are. Speaking of that: I’m always open for ideas to incorporate. The predictions (10-9 overall this week, 6-4 in the FBS, and while Baylor and Cincinnati lost their games, they both covered the spread) were from a suggestion about 3-4 years ago. Send in requests and I’ll do what I can.
Here is the GCR following Championship week. There are 5 games next week: the Army/Navy game, and the quarter finals for FCS (Northern Iowa at James Madison, Austin Peay at Montana State, Montana at Weber State, and Illinois State at North Dakota State).
Later this week, in G’s Exploration we’ll talk about the upcoming bowl schedule and the blog schedule during bowl season. On Thursday, I’ll predict this week’s games.
With no further ado, here is this week’s GCR. Comment, challenge, and please share with others – thanks, G
Robert,
Yay! We’re number 73! Although we could move up if Utah becomes national champs! (Oh, wait…)
For the final point spread, just cap it at 35. If a “classy” team is ahead 35 points at half-time, they would mainly play their second stringers in the second half. Same if they are ahead 28 points going into the fourth quarter, which is not much different. Without examining game script and partial scores, you can’t tell if a team should be playing their second stringers or if they did. I was thinking of a much more complicated way to do that based on the final score, the half-time score, and maybe even the third quarter score, but I think the cap simplification saves you a lot of data input and study, and actually is still accurate for classiness.
Now, if you wanted to give teams credit for their depth, then you might want to get into something more complicated. However, your current system actually does that. All teams suffer injuries to their first string players. If they don’t have depth, their subsequent game results (which you track) would be different.
A classy coach worries about his first-string players’ injuries for the rest of the season, for next year, and for their potential NFL draft position. He also recruited a lot of players knowing that many of them would be second-stringers, and he can give them playing time under the lights.
Carl
Carl, you are insightful as usual. A cap might be the best way to go. Hmmm. I think I can do a what-if to see what it does. Thanks!!!!