Monday, February 16, 2004
Can you Dig it?
By Joe Lunardi Special to ESPN.com
There's a certain analyst for a certain four-letter network who gets paid by the same people I do. His real name is Richard, and he won a whole bunch more college basketball games than I ever will.
But, in an attempt to separate contenders from pretenders this season, Richard got it wrong. By a lot.
I know about the attempt because close to 200 e-mails have come my way on the subject. And I know who's right -- and who's not -- because of something called Adjusted Scoring Margin (ASM).
We first started playing around with ASM prior to the 2000 NCAA Tournament. The premise is to do the same thing with a team's average margin of victory as RPI does to a team's won-loss record; namely, measure success (or failure) in the context of each team's level of competition.
In other words, if both the Duke Blue Devils and the Central Connecticut Blue Devils are 21-2, one might assume that the teams are not really equal due to a vastly different level of opponent. The RPI rankings do a nice job of confirming this.
Similarly, if both Duke and Central Connecticut are winning by an average of 20 points per game, it might be useful to know that Central Connecticut is doing so against below average teams while the Dookies are playing a national schedule. Adjusted Scoring Margin tells us that. It determines whether or not teams are winning (or losing) by the margins they are supposed to, given a set slate of opponents.
Bottom line: Last season, Duke played a slate of opponents who scored 71.9 ppg and allowed 69.0 ppg. Against that schedule, the Blue Devils scored 82.1 ppg (a positive Offensive Quotient of 13.1 ppg, or 82.1 minus 69.0) and allowed 70.2 ppg (a positive Defensive Quotient of 1.7 ppg, or 71.9 minus 70.2). Add the two quotients together, and you get an ASM of 14.8. In other words, Duke was 14.8 points per game better against its schedule than the average team would have been.
So, back in 2000, we ran ASMs for all 64 NCAA teams. The two best totals on the board were Michigan State and Florida, the latter a No. 5 seed in the East. Just for kicks, I publicly predicted the Spartans over the Gators for the national championship (including a Florida upset of Duke along the way). And darn if that isn't exactly what happened.
In 2001, the four best ASMs belonged to Duke, Arizona, Michigan State and Stanford. The first three made the Final Four and Stanford, a No. 1 seed, lost to the fifth-best ASM (Maryland) in the West Region final.
In 2002, the Final Four included Kansas, Oklahoma and eventual champion Maryland. These three teams had the second-, third- and fifth-best Adjusted Scoring Margins in the field. Even Indiana, the unlikely runner-up that year, had a healthy ASM of 15.2 (the very best teams in a given year are typically between 18 and 20).
Finally, last season, the selection committee broke our streak by placing the top two ASM teams -- Kansas and Arizona -- in the same region, with two of the remaining top five (Pitt and Kentucky) on the same side of the bracket overall. Thus, because of less than great placement, only two of those teams could reach the Final Four. The next team on the list, Texas, did advance to New Orleans.
To review, our four-year experiment ends like this: 10 of the last 16 Final Four teams have had Adjusted Scoring Margins in the top five of their respective seasons. So, when I hear my friend Richard calling out top five ASM teams this year, my ears perk up.
We will release ASMs for all NCAA teams as part of the ESPN.com Insider tournament guide. The current top five includes Duke, Saint Joseph's, North Carolina, Stanford and Gonzaga. And it just so happens both Saint Joseph's and North Carolina are on Richard's "pretender" list.
Me? I'm fairly convinced at least two of the five teams listed above are going to the Final Four, due to both their on court performance as well as the four-year history of ASM. But if Richard wants to discount North Carolina (because of six losses, two in overtime, against the No. 4 schedule) or Saint Joseph's (with both the No. 1 InsideRPI and NonConf RPI, not to mention a nation's best 14-0 record vs. Top 100 teams), he does so at his own risk.
And I can dig that.
Joe Lunardi is the resident Bracketologist for ESPN.com, ESPN Insider and ESPN Radio. He may be reached at firstname.lastname@example.org.