richter@milano.UUCP (02/18/86)
A couple of weeks ago I computed and posted the average rating per team for
several NCAA division I conferences. The team ratings are Jeff Sagarin's
computer ratings listed each Tuesday in USA today. I simply total the
ratings for the teams in a conference and then divide by the number of
teams in that conference. The values computed from the 2/18 ratings
(along with the values from two weeks ago, listed in parentheses), are:
1. ACC - 32.4 (31.4) 5. Metro - 53.0 (56.6)
2. Big Ten - 41.6 (40.7) 6. Southeastern - 53.2 (56.7)
3. Big Eight - 46.8 (47.6) 7. Sun Belt - 65.9 (72.1)
4. Big East - 50.9 (48.6) 8. Pacific 10 - 91.9 (89.0)
Some comments (intended as observations, not as value judgements):
1. Assuming the computer ratings are reasonable (and I suspect that
they are), I still don't know how much the average ratings mean.
Because this scheme takes into account all teams in each conference,
one or two real dogs at the bottom of a conference can affect the
conference's rating significantly. Likewise, great strength at
the top could offset weakness at the bottom.
2. The ACC's biggest advantage is at the top of the ratings, where
North Carolina, Duke, and Georgia Tech are ranked first, second,
and third, respectively. The Big East, on the other hand, enjoys
an advantage at the bottom, where Seton Hall is ranked 116th
(compared to 143rd for Wake Forest, 164th for Northwestern, and
146th for Colorado).
3. The Big Ten has five teams in the top 17, hence it's advantage
over the Big Eight and Big East. Furthermore, eight of the Big
Ten's ten teams are in the top 47. (That fact more than offsets
Northwestern's 164th place.)
4. The Big East is weak in the middle: there is a relatively large
drop from Georgetown (15th) to Pittsburgh (37th) and from Pitts-
burgh to Villanova (58th). Whereas the first seven teams in the
Big Eight are sprinkled fairly evenly from 4th to 61st, the Big
East has four teams clustered between 58th and 80th.
After I posted my earlier figures, I received some suggestions that I
adjust my procedure by including equal numbers of teams from all
conferences, dropping bottom teams, dropping top teams, etc. However,
I have once again used ALL teams from each conference, as I was
primarily interested in a conference's OVERALL strength.
I welcome any additional computations, suggestions, or observations.
Enjoy.
--
Charlie Richter MCC Austin, Texas
uucp: {ihnp4, seismo, harvard, gatech, pyramid}!ut-sally!im4u!milano!richter
arpa: richter@mcc.ARPA
"The airplane stays up because it doesn't have the time to fall."
- Orville and Wilbur Wright