Home winning pct of top 100 RPI - cougarguard.com — unofficial BYU Cougars / LDS sports, football, basketball forum and message board

jay santos · 01-14-2008, 06:24 PM

Indy, you may be factual, but you're misleading. The statistics challenged portion of your readership don't understand what you're saying.

http://www.cougarboard.com/nologin/m...tml?id=3372154

Maybe top 100 RPI win 88% at home. (Using my computer model top 100 I got 85% winning pct at home.) But that doesn't mean you can take any given game for a top 100 RPI team at home and blanket say they have 88% chance to win.

Ranking 90-100 at home against top 10 was only 1/6.

In games with teams with equal ratings, you can expect home team to win 66% of the time (fairly close to college football).

In order to even the home field advantage, the visiting team would have to have a power ranking of 5.5 points. That's about the difference of #40 and #80 in a typical year.

My model has BYU #58, Utah #60, and UNLV #53, so for the games this week it would predict:

BYU 31% chance beat UNLV away with power rating diff of -0.5
BYU 34% chance beat Utah away with power rating diff of +.2

Since the computer models are still not very accurate, it depends on what you think of these teams and what the point differential would be on neutral court.

01-14-2008, 06:24 PM	#1
jay santos Senior Member Join Date: Jan 2006 Posts: 6,177	Home winning pct of top 100 RPI Indy, you may be factual, but you're misleading. The statistics challenged portion of your readership don't understand what you're saying. http://www.cougarboard.com/nologin/m...tml?id=3372154 Maybe top 100 RPI win 88% at home. (Using my computer model top 100 I got 85% winning pct at home.) But that doesn't mean you can take any given game for a top 100 RPI team at home and blanket say they have 88% chance to win. Ranking 90-100 at home against top 10 was only 1/6. In games with teams with equal ratings, you can expect home team to win 66% of the time (fairly close to college football). In order to even the home field advantage, the visiting team would have to have a power ranking of 5.5 points. That's about the difference of #40 and #80 in a typical year. My model has BYU #58, Utah #60, and UNLV #53, so for the games this week it would predict: BYU 31% chance beat UNLV away with power rating diff of -0.5 BYU 34% chance beat Utah away with power rating diff of +.2 Since the computer models are still not very accurate, it depends on what you think of these teams and what the point differential would be on neutral court.