Started By
Message
locked post

RPI Rant

Posted on 2/12/09 at 2:37 pm
Posted by lsumatt
Austin
Member since Feb 2005
12812 posts
Posted on 2/12/09 at 2:37 pm
As many of you know, I am big fan of computer polls in football. They aren't perfect, but they are objective and they average 6 of them together. They are pretty good about treating a #75 team almost equally with a #120 team, which I think is good.I never followed RPI much, but I always thought it gave screwy results. It seems you are waaaaay better off playing really bad teams than really awful teams for your SOS (I contend you ought to beat either one). Moreover, it seems to give too much influence to SOS.

Here's the kicker, and I might be wrong, but I just did some back-of-the-envelope math and I think if you replace LSU's win over Alcorn State for example with a LOSS to Duke, our RPI goes up. So how does LOSING to a really good team make you a better team? Me and 4 of my buddies could lose to Duke.

Now, I know LSU has played an easy schedule and they don't deserve to be put with other 4-loss teams that played tough ones. But I think LSU's RPI is a little deflated due to some oversimplified RPI formulas
This post was edited on 2/12/09 at 2:39 pm
Posted by LuckyLee
inside vaginas
Member since Jul 2008
9145 posts
Posted on 2/12/09 at 2:42 pm to
It depends on where the teams play wrt your Duke example.
Posted by I-59 Tiger
Vestavia Hills, AL
Member since Sep 2003
36703 posts
Posted on 2/12/09 at 2:44 pm to
quote:

but I just did some back-of-the-envelope math and I think if you replace LSU's win over Alcorn State for example with a LOSS to Duke, our RPI goes up.


I remember in 2000 when we were RPI watching, St Johns lost a mid-season game at Duke by 35 and their RPI went up the next week.
Posted by TigerNacho
Member since Jul 2005
274 posts
Posted on 2/12/09 at 2:45 pm to
One could make the argument that losing to Duke, rather than beating a cream puff shows some amount of effort and risk-taking (to play a tougher schedule) that should perhaps be rewarded.
By the way, always enjoy your math posts.
Posted by el tigre
your heart
Member since Sep 2003
49712 posts
Posted on 2/12/09 at 2:47 pm to
i have zero problem withe formulation. In a tourney that must select from 300+ DI teams this is a necessary evil to ensure that every team with a nice record against shithole competition doesn't make it to the dance. I also like that in theory it should reward fans with interesting OOC matchups.
Posted by The Easter Bunny
Minnesota
Member since Jan 2005
45570 posts
Posted on 2/12/09 at 2:50 pm to
quote:

Moreover, it seems to give too much influence to SOS.


Definitely. Football doesn't factor it in enough, basketball does too much I think. We should work on making our own computer poll that's an extension of the RPI
Posted by Nuts4LSU
Washington, DC
Member since Oct 2003
25468 posts
Posted on 2/12/09 at 3:50 pm to
quote:

Here's the kicker, and I might be wrong, but I just did some back-of-the-envelope math and I think if you replace LSU's win over Alcorn State for example with a LOSS to Duke, our RPI goes up.


I was seriously wondering the other day whether our loss at home to Xavier was helping our RPI more than our win at Tennessee. Since the SOS only counts winning percentages and doesn't count what the team's ranking is, we don't get much credit for beating Tennessee even though they are highly ranked in the RPI ratings because their record is mediocre.
first pageprev pagePage 1 of 1Next pagelast page
refresh

Back to top
logoFollow TigerDroppings for LSU Football News
Follow us on Twitter, Facebook and Instagram to get the latest updates on LSU Football and Recruiting.

FacebookTwitterInstagram