'dawg's Bracketology by the numbers

[99topdawg]

Below is a compilation of 5 different computer rankings plus the RPI top 50 wins and 150+ losses. My explanation is given in each column’s description. I am by no means saying that this is the way it should be, but how the numbers ran. I came up with my format before I did this so I didn’t give any favor to any team I reeeeealy want to get in. My calculations are 100% accurate, unless you find a mistake.

Computers—I averaged each team’s ranking from the RPI, BPI, Massey, Sagarin, and KenPom (did not break ties)

Computer Adjusted—I took those 5 rankings and threw out the top and bottom score and averaged the 3 middle scores

SOS—I averaged the SOS from RPI, BPI, KenPom, and Sagarin (Massey doesn’t have a SOS ranking)

Top 50 wins – 105+ losses—I subtracted the number of 150+ losses from the number of top 50 RPI wins. I broke ties with the number of top 50 wins. If still tied, I averaged the top 50 win opponent RPI. If no RPI top 50 wins, I used 150+ loss RPI rank. I struggled with how to rank the top 50 and 150+. Wasn’t sure that I could compare 21-1 150+ with 5-1. The team that was 21-1 was already punished with SOS and 5-1 team was rewarded in that category as well.

Composite—I averaged rankings of the 4 columns and ranked the averages. Ties were broken by using top 50 wins. My Bracketology is below.

	Computers	Comp. Adj.	SOS	Top 50 wins – 150+ losses	Composite
Villanova	9	11	1	1	2
Wichita St.	1	3	11	8	3
Iowa State	4	4	7	7	1
Temple	12	12	8	3	11
LaSalle	10	10	12	11	12
Boise State	6	8	5	5	5
Kentucky	5	5	4	9	4
Virginia	3	2	10	12	6
Tennessee	14	14	2	2	9
Mid. Tenn. St	1	1	16	15	10
Ole Miss	8	7	15	13	13
Baylor	7	9	3	10	8
Southern Miss	11	13	14	16	16
Alabama	15	15	6	14	14
Iowa	9	6	9	4	7
Arizona St.	16	16	13	6	15

Last 4 byes
Iowa State
Villanova
Wichita State
Kentucky

Last 4 in
Boise State
Virginia
Iowa
Baylor

First 4 out
Tennessee
Middle Tennessee
Temple
La Salle

Next 4 out
Ole Miss
Alabama
Arizona State
Southern Miss

 

[volhawk50]

The computer adjusted may be a bad idea. In our case, it throws out our RPI numbers, thus lessening the effect of non conference strength of schedule and eliminating the factor of performance away from home. Those are important areas that are keeping us out of the field at this point.

 

[99topdawg]

The computer adjusted may be a bad idea. In our case, it throws out our RPI numbers, thus lessening the effect of non conference strength of schedule and eliminating the factor of performance away from home. Those are important areas that are keeping us out of the field at this point.

What I was trying to do with this is not predict the teams that get in, but rather, look at the teams as a whole. I didn't give weight to any of the computer rankings, even though I realize the committee does. My idea came from the fact that our other 4 rankings were very good, but the RPI was our outlier. I really don't see why the committee is stuck on using the RPI only.

 

[HawksMN]

I don't think the committee is stuck on using only one metric, are they?

 

[99topdawg]

I don't think the committee is stuck on using only one metric, are they?

I'm not positive of it, but according to this site, the criteria they use is the RPI.

Click "study guide" then "the stats"

 

[H8IAST8]

Their data displays RPI, however they are encouraged to visit other sites like Ken Pom, Sagarin, BPI, etc.

Different people look at different things. Some people looked at margin of victory while that number was irrelevant for most participants. Head-to-head and road wins were a major factor for some while others relied more heavily on top-50 wins and avoidance of bad losses. Some committee members bring the eye test or advanced metrics into the equation; some feel as those shouldn't be part of the process. Though we're relying primarily on facts, it's still a very subjective process.
Three things learned at the NCAA mock selection exercise - CBSSports.com

The NCAA allows (but from what I interpreted, does not heartily endorse) any Selection Committee member to use Sagarin, KenPom, LRMC, Massey or any type of ratings system (including — WHAT — the Coaches’ Poll? It’s true, unfortunately). Those systems are not brought up on the big screen, unless by request, which never happened at our mock.
RPI still hovers over, cloaks selection process - CBSSports.com

Committee members are able to use other metrics systems (ex. KenPom, Sagarin, etc.) to evaluate teams on their own, but RPI is all over every piece of material the NCAA provides committee members. (This means Cincinnati might have a tough time making the field -- RPI is 93 per CollegeRPI.com. A weak RPI for a major-conference team that could make the tourney is generally around 60-70.)
NCAA mock selection gives tips for evaluating bubble teams