On Sunday, November 26, 2023 at 12:40:50 PM UTC-5, JGibson wrote:
Apparently, the four computers used to determine the championship game were Anderson & Hester (the one formerly known as the Seattle Times ranking in the BCS days), Colley Matrix, Anderson, and Wolfe.
It came up with Boise State at UNLV.
How does that compare to different Dolphin rankings, which I've always appreciated because of the way he has different-style rankings depending on what you would want.
Dolphin Standard: Boise State at UNLV
Dolphin Median Likelihood: Boise State at UNLV
(these two are the ones I would expect to act most like that set of 4 rankings)
Dolphin Predictive: San Jose State at Boise State
(this one would act more like FPI or the College Basketball NET rankings) Dolphin Improved RPI: Boise State at UNLV
Dolphin Pseudopoll: Boise State at UNLV
(the one that will act most like humans)
Most of the ones that factor in W/L heavily and not just scoring margin agree with the call.
ESPN FPI and Sagarin would have gone with the Dolphin Predictive: SJSU at Boise State
So, all of them finished 6-2 in conference, but total, SJSU and Boise State were only 7-5 compared to UNLV's 9-3.
How did they fare non-conference?
UNLV: beat Bryant 44-14, lost to Michigan 35-7, beat Vanderbilt 40-37, beat UTEP 45-28
Boise State: lost to Washington 56-19, lost to UCF 18-16, beat North Dakota 42-18, lost to Memphis 35-32
San Jose State: lost to USC 56-28, lost to Oregon State 42-17, beat Cal Poly 59-3, lost to Toledo 21-17
I meant to add ESPN's SOR to this analysis as well - philosophy is very similar to Dolphin standard, although not necessarily with the same math. ESPN SOR actually would have gone with San Jose State at UNLV, although I should note that in both ESPN SOR
and Dolphin Standard, Boise State and San Jose State are very close (63rd vs. 66th for SOR and 56th vs. 58th for Dolphin Standard).
--- SoupGate-Win32 v1.05
* Origin: fsxNet Usenet Gateway (21:1/5)