A little thought experiment

twoclubs · Sat, 17 Feb 2001 19:57:15 -0800

I had some time on my hands and Samer's stats
were up and available, so I crunched some numbers
based on the arguments developing concerning NAQT and
the selection process. Keep in mind I didn't know
what the outcome would be beforehand...If you don't
like stats, I would skip this message.

First, a
table of each of the regions based on average P/20TH
for all teams in attendance.

Team P/20TH
FS
ME 237.6 100%
MA 233.2 98.1%
W 222.3
93.6%
SW 218.6 92.0%
MW 218.3 91.9%
NE 211.9
89.2%
SE 207.7 87.4%
Can 172.5 72.6%
(NW is hereby
neglected based on their DII competition)

I have
argued in other posts that W-L is more important than
total pts. scored. Let us assume that each are of equal
importance for cross-region comparisons, especially when
taking into account field strength. Thus, for each team,
P/20TH divided by the number of losses (normalized based
on a 13 game schedule and adjusted for field
strength) should be an accurate method for ranking the
"bubble" teams.

21 teams received automatic bids,
leaving 15 for at-large status. Here are the
corresponding #s for those schools selected and waitlisted, and
the #s for a few random schools:

 Team
P/20TH/losses P20L*FS
 1.Mich A 187.8 187.8
 2.Wisc 169.6
155.9
 3.FLA 154.5 135.0
 4.CalTech 127.9 119.7

5.UVA 107.7 105.7
 6.Pitt 92.7 90.9
 7.TX A&M
95.9 88.2
 8.UGA 97.0 84.8

Everything's
kosher up until this point...

 *9.FLA ATL 91.3
79.8
 10.Yale 89.2 79.6
 11.Penn St. 69.0 67.7

12.Illinois 59.1 59.1

And of course, our usual
suspects from the MA...

*13.Maryland 59.5
58.4
*14.GW 59.3 58.1
 15.UPenn 58.8
57.7
--------------------------------------------------
*16.Williams 56.0 50.0
 17.MIT 48.6 43.4
 18.BGSU 57.5
41.7
*19.Duke 46.3 40.5
 20.Oklahoma 43.4
39.9
**21.Queens 52.3 37.9
*22.Princeton B 36.6
35.9
**23.Minnesota 37.3 34.3
*24.Stanford 36.1
33.8
**25.Michigan C 45.8 33.3
**26.S.Carolina 36.5
31.9
*27.Fla St. 32.0 28.0
*28.UCLA 28.7
26.9
*29.Harvard B 24.6 21.9

*-alternate to NAQT
nationals
**-did not receive alternate bid to NAQT
nationals

This method does not take into account IFT performance
or any other factors that NAQT deems important, but
I believe that it does correct for differences
across regions and provides a simple way to correlate
both stats and W-L. Of course NAQT can do what it
wants, but if I was a member of the Florida Atlantic or
GW teams, I would be a little curious about the
selection process as well. Feel free to debate as you
please.

Cheers,
Shaun