Before the NCAA tournament began, Erich Doerr – a WoW Journal reader (and occasional contributor) – posted the results of his Monte Carlo simulation of the tournament.
An Instant Analysis of the NCAA Tournament
The above column included the following paragraphs:
Now that the seeds have been given, what are the odds on a national championship? One way to assess probabilities is to simulate the tournament, using basketball statistics. Given 10,000 simulations, a Monte Carlo method, we can generate a plausible list of championship odds for each tournament entrant.
The two strongest public NCAA metrics are the Sagarin Ratings and Ken Pomeroy‘s Pythagorean Ratings. Statistics used by the Wages of Wins are parallel to Pomeroy’s approach, as both incorporate offensive and defensive efficiency. Individual game outcomes can be modeled via methods like log5 analysis and Sean Foreman’s Sagarin approach.
With this approach, we should expect to see favorites generally prevail, and given enough trials, we’ll see a low seed team go all the way.
Which #1 got the easiest bracket? This analysis suggests Kansas has the easiest path by either metric. Sagarin seems to show UCLA got the toughest draw while Pomeroy’s stats believe North Carolina received the toughest draw.
The following table reports my analysis:
Table One: Projecting the National Champ
As you can see, Erich’s simulation – conducted before the tournament began — indicated the top two teams were Kansas and Memphis. And of these two, Kansas had the best odds to be the champion.
Tonight – after Kansas defeated Memphis — Erich posted the following comment:
I certainly hope you played that KenPom bracket!
Keep your eyes here in May for an assessment on how these kids shape up for the next level in the NBA draft preview, and be sure to drop on by next year for another NCAA tourney preview.
For the record, I did not play the KenPom bracket. I picked UCLA to win the title (mostly on a hunch, I didn’t run my own simulation).
I do wish to thank Erich for posting this analysis (even if I failed to follow its advice) and I am very much looking forward to his NBA draft preview.
WoW Journal Approaches a Milestone
One last note before I close this post.
According to WordPress, the WoW Journal should hit 1,000,000 page views in the next day or so. The counter is on the right (look at Blog Stats, which is below the sitemeter and the list of Pages). As I post this column, the counter just passed 998,000. We get about 2,000 page views per day, so I would guess we will pass the magical 1,000,000 mark sometime tomorrow night.
There should be a prize for being the person who gets us to 1,000,000. But our WoW Journal budget (currently sitting at zero) has no funds for such a prize.
In all seriousness, though, we do with to thank everyone for stopping by and making this a small part of their day.
– DJ
Mike
April 8, 2008
I played a combination of KenPom and Sagarin ratings. I happened to be in Vegas the first weekend of the tournament, and did very well… and I destroyed folks in my work pool.
However, I’ve been playing Sagarin ratings for years now, and have never had nearly this much success. The success of the “sabermetric” type ratings this year is probably equally due to their predictive ability as it is pure luck.
goldshammgold - london edition
April 8, 2008
Unless the NCAA tournament selection team is out to lunch, is it really much of a surprise that a Monte Carlo simulation resulted in a Kansas-Memphis final? What I’m interested in, however, are the stats that explain why there were so few upsets this year — were the top four seeds really statistically superior than in prior years?
Rue Des Quatre Vents
April 8, 2008
I played the Sagarin ratings and it was still good enough to come out 2nd in one company pool, and 1st in a pool among family and friends. All good for just about $1000. Nice!!!
TheGreenMiles
April 8, 2008
Yup, I used the rankings to aid my picks and finished in the top five of a couple of pools. If I’d followed them more closely I might’ve won, but I didn’t want to pick all four #1 seeds to reach the Final Four, so Pitt did me in.
John Middleton
April 8, 2008
I used KenPom picks in one bracket and Sagarin in another and won them both. I dominated with the KenPom picks. I love using mathematical analysis to make the picks, because otherwise my biases get in the way (I’ve picked Gonzaga to win it all twice in the past 6 years). Yeah, I needed something else :-).
Will
April 8, 2008
I used the KenPom picks and won all three pools I was in – $1500 coming my way. Thank you Erich – if you’re ever in LA I owe you a drink or twelve.
Ben Guest
April 8, 2008
I did use the KenPom picks and picked Kansas to win, even though I hadn’t seen one game. I won my office pool. Another victory for Wages of Wins…
niles
April 8, 2008
The post is interesting but the relative strength of each bracket changes throughout the tournament with upsets. UCLA had a relatively easy road and got waxed by Memphis. As someone else said, the committee, and most fans basically got it right. There was not much to figure out besides Davidson and Western Kentucky
Erich
April 8, 2008
Yes, this was a banner year for the favorites and the KenPom bracket came up roses. Hearing of your successes is personally thrilling and I congratulate each of you.
I had some pep talk ready just in case the dice went sour, but I’ll save that for next year. Today is a day for celebration.
Thank you for your success stories, and thank DBerri for the forum!
k
April 9, 2008
The measure of predictive ability is not predicting that a low seed will win, it’s predicting when a low seed will lose. Notice that Duke is ranked 6th of 64 by this method.
Erich
April 11, 2008
Nice,
Thanks to the Numbers Guy column, the degrees of separation between my name and the WSJ is down to two. I never would have guessed it would have been NCAA basketball….
francesco
April 17, 2008
I tried to follow the example of calculation of Wins Produced presented in “TECHNICAL NOTES” (Pistons 77-78).
Here is the formula used for calculate DEFTM48;
DEFTM48=[(3688*-0.032 + 853*0.033 +
18*-0.032 + 437.7*0.033)/19,855]*48 = -0.1839
now, I can’t understand where 853, 18, 437.7 came from. These number should be
– opponent’s TO
– team TO
– team Reb
but the stats I took in http://www.basketball-reference.com don’t confirm this.
Plese, help me.
Thanx