RebKell's Junkie Boards

[Shades]

_________________
Nnekalonians 1:14 - Thou shalt not accept that which is not earned

[Shades]

_________________
Nnekalonians 1:14 - Thou shalt not accept that which is not earned

[Shades]

_________________
Nnekalonians 1:14 - Thou shalt not accept that which is not earned

[linkster]

On 12-11 Tennessee, 10-0, was ranked 9th in RPI (.6742) and 54th in SOS(.5657), having moved up 17 spots in the rankings, likely due to their win over Texas.

On 12-28, Tennessee, 12-0, had slid to 17th in RPI(.6443) and 102 in SOS(.5258). This slide came despite the Vols wins at Long beach St and at then No 18 Stanford.

Green Bay is currently ranked 5th at 9-0 and yet in the event they loss but one more game this season but because of that loss they don't win their conference, they will have to sweat out an invite to the tournament. I understand the need for some sort of tool to rank 350 teams who for the most part never play each other but it's time for committee to reduce it's importance while adding a ranking based on performance in games. A 2 pt loss in OT shouldn't be equivalent to a 30 pt loss. And a 2 pt win in OT shouldn't be much better than a 2 pt loss in OT. It's time for the NCAA to move into the 21st century and utilize the advanced metrics that are available.

[GlennMacGrady]

[Phil]

I am sympathetic to the reasoning behind the exclusion of point margin in the calculation, but there are multiple ways to include some measure of margin without creating the concern about running up the score. I'm surprised that no one at the NCAA has figured out how to do it.

One week argument for not making the change (other than the most likely reason — inertia) is that the goal is to help see the teams that tournament time not provide a midseason ranking. By the time the tournament is here at the flaws in the RPI have asymptotically gone away and it's merely bad as opposed to atrocious. However, a simple adjustment would help make the metric more useful during the season and would be a small additional contribution to fan interest.

Or who knows, maybe they did a study and found that fans get a lot of satisfaction about whining about how bad the RPI is and that exceeds the value of making it useful.

[calbearman76]

I have come to believe that the RPI is actually worse than no rating at all. Its fatal flaw is the overweighting of games against awful teams so long as they are Division 1. Baylor played Coppin St (1-11), McNeese St (1-7), and Lamar (1-5) so their strength of schedule looks really weak. To show the stupidity of the RPI, look at Lamar and Baylor. Lamar is 1-5 vs Division 1 schools but has also played 4 non D1 schools. According to the RPI they have the #7 toughest schedule strength. Their other game are against Purdue, North Dakota, Northern Illinois and New Mexico. Baylor has played UCLA, Georgia Tech, Stanford, Kentucky, Central Arkansas and North Dakota and hasn't played any non D1 schools. It has the #169 schedule strength. Schedule strength counts for 75% of a teams RPI. So even though Baylor has played a much tougher schedule than Lamar, the RPI says that Lamar is 162 slots better on that metric. Overall Baylor is 35, Lamar is 121.

We have to follow RPI because the NCAA uses it. But as a rating system it is worse than nothing.

[pilight]

[calbearman76]

[pilight]

[calbearman76]

[pasteurize]

I understand that RPI is comprised of 25% of the win percentage of a team, 50% of their opponents' win percentage, and 25% of their opponents' opponents' win percentage. And I also know that the win percentages are calculated using only games played against other D-I teams.

Is there anything else that factors in? Are home vs away games weighted differently when the NCAA performs their calculations?

[purduefanatic]

[pasteurize]

[Shades]

_________________
Nnekalonians 1:14 - Thou shalt not accept that which is not earned

[Shades]

_________________
Nnekalonians 1:14 - Thou shalt not accept that which is not earned