Selection Sunday: What Were The Best Conferences in Mens College Hoops in 2011?

In the adjacent table, I attempt to find out how the conferences stack up against each other by measuring the parity of the conference and the strength of the conference.  The conference labeled in blue means that it was top 5 in that category and the red label means bottom 5. 
There are some interesting results in this table; the Pac-10 being the best overall conference came as a slight surprise to me.  It was in the top 5 in conference parity and conference strength, so I should not be too surprised.  If I had to pick the top 5 conferences just based on what I saw and perceived, I would have picked the same 5, but probably not in that order.  The Southland conference had the most parity with a 136.2 rating; by default that was a surprise because I haven’t been following the Southland all that closely.  The strongest conference, which was more important to me, was the Big Ten and I would tend to agree with that.  An earlier post of mine showed that the 2011 Big Ten and Big East conferences were two of the top three strongest conferences of the last decade.  Click to see my methodology
I see perfect parity in the conference meaning that all teams finish at .500 in conference play, as a result, the standard deviation of wins in conference should be zero.  Therefore, as the conference standard deviation of wins approaches zero, the parity rating is higher.  The converse is also true; as the standard deviation moves away from zero the parity rating would be lower.  A parity rating of 100 means average parity compared to the other conferences.  For example, the Southland Conference parity rating is 136.2, which means that conference has higher than average parity.
Conference strength was a much simpler measurement; I just used the conference’s non-conference win percentage and compared it to the rest of the conferences.  Just as with the parity measurement, 100 is average. 
To measure the overall quality of the conference I combined the strength and parity indices.  I believe that conference strength is more important than conference parity when comparing the quality of different conferences, so I counted strength twice and parity once in computation of overall conference quality.  However, I also believe that a conference having great strength and little parity should not be rewarded with a high overall quality rating.  As a result, I used the geometric mean of the strength and parity measures, rather than the simple average of the two.  Here’s an illustration showing why I chose geometric mean.  Look at the Southland Conference; it is in the top 5 in parity and the bottom 5 in strength.  If I just averaged parity and strength, the parity measure would have too much influence on the overall quality measure: the overall quality would be 103.9, an almost 2 and a half-point increase, a big difference on this scale.
Advertisements

2 Responses to Selection Sunday: What Were The Best Conferences in Mens College Hoops in 2011?

  1. Anonymous says:

    why would the mean be at all reasonable…strength is far far more important than parity…if a conference had two teams that finished 0-fer in the conference but were undefeated in non conference that would still mean the conference was strong — but according to you the conference would be penalized for that…that makes absolutely no sense at all…by the same token…if everyone in the conference was at .500 in conference but were sub .500 in non conference, you would reward them for that parity…strength is far more important…and should be weighted at a multiple of parity, like 2:1 or 3:1

  2. First, thanks for the feedback.I did not use the arithmetic mean, which is what most people use when they say average. I used the geometric mean; the geometric mean doesn't reward a conference for having an extremely high parity with a low strength rating, and vice versa. As for your comment about the conference with two teams finishing 0-fer; the conference strength would be increased, but the conference parity will be decreased.You are correct about your second example. What is perfect parity? It's everyone having the same record.You are right about strength being more important the parity; in fact, I did weight strength twice and parity once. Giving strength more than two times the weight of parity is matter of opinion, you would not be necessarily wrong if wanted to the ratio to be 3:1.Finally, when I say best overall conferences, that could be misleading. If you are looking for the best conference, strengh is all that matters. So maybe I should change the title, but I rather follow a conference that has slightly less strength and a lot more parity.And again; thanks for the criticism.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

%d bloggers like this: