20 Second Timeout is the place to find the best analysis and commentary about the NBA.

Tuesday, April 03, 2007

Why Danny Sheridan Gets Paid to Handicap Games--and You Don't

Have you ever wondered how well a handicapper like Danny Sheridan would do if he filled out an NCAA Tournament bracket? Every year before the tournament there is that list in USA Today that shows Sheridan's odds for each of the 64 teams. ESPN.com allows you to fill out up to five brackets, so this year I filled out one bracket based entirely on Sheridan's odds. The result? Sheridan's bracket finished at the 97.8 percentile, with 53 games correctly picked out of 63. Of course, since ESPN gets so many entries, that worked out to finishing in 63,232rd place. One big caveat: Sheridan had Florida and North Carolina each as 3-1 favorites to win the title; Florida was listed first and based on the order that other tied teams were listed in that did not seem to be because of alphabetical order, so I selected Florida in the "Sheridan bracket." Obviously, if I had gone the other way then the bracket would not have scored so well. Sheridan's Final Four was Florida, North Carolina, Ohio State and UCLA (tied at 6-1 with Kansas, but listed ahead of Kansas in USA Today's rendering of Sheridan's odds). So he got three out of four, missing Georgetown (tied with Memphis for 6-7 on his list). His Elite Eight were Florida, North Carolina, Ohio State, UCLA, Kansas, Memphis, Georgetown and Wisconsin, so he scored seven out of eight; Oregon, which got an Elite Eight spot instead of Wisconsin, was ninth on Sheridan's list at 12-1 (Wisconsin was 10-1).

So, if you went strictly by Sheridan's picks you would have gotten three of the Final Four teams and seven of the Elite Eight teams. You would have had one half of the championship game matchup. The main thing that he missed was ranking North Carolina so highly. Of course, the Tar Heels blew a big lead and lost to Georgetown in overtime.

In addition to filling out a "Sheridan bracket" I also did a "Jeff Sagarin" bracket. Sagarin is a 1970 MIT graduate who has been ranking teams in various sports seemingly forever. There were no ties in his rankings, though some teams were ranked within hundredths of a point of each other. The "Sagarin bracket" finished at the 87.4 percentile, with 49 games correctly picked out of 63. That works out to 368,944th place. Sagarin had North Carolina and Ohio State 1-2, but based on the seeding that would have meant that North Carolina would beat Ohio State in the Final Four and then win the championship over Florida. Sagarin's Final Four was North Carolina, Ohio State, Florida and Kansas. UCLA and Georgetown, which made it instead of North Carolina and Kansas, were sixth and eighth in his rankings. His Elite Eight picks were North Carolina, Ohio State, Florida, Kansas, Wisconsin, UCLA, Memphis and Georgetown, so he only got five of those right.

It is interesting that both Sheridan and Sagarin overrated North Carolina and Wisconsin, though I have no idea what that actually means regarding their methodologies. If it is not obvious I should clearly state that this is not meant as a scientific evaluation of their abilities as a handicapper or statistician respectively; one tournament bracket is far too small of a sample size from which to draw any sweeping conclusions. Still, both men did a more than reasonable overall job of assessing the tournament field.

What about my other three ESPN brackets? Before answering that, I should mention that I watch less college basketball now than I ever have, mainly because I watch so much NBA basketball--and I don't feel like I am missing anything, because I have always felt that the NBA game is a much better game than the college game, a subject that I wrote about last year; a recent Charley Rosen article also makes an excellent case for the vast superiority of the NBA game to the NCAA game. When I do watch NCAA basketball, I am much more likely to see the "name" teams than the "Cinderella" ones, so the early rounds are a bit of a crap shoot for me outside of the slam dunk 1-16 matchups. With those excuses out of the way, the best of my other three ESPN brackets finished at the 91.7 percentile (48-15), which is 244,355th place. I had the correct title game matchup but picked Ohio State to win. My only Final Four miss was picking North Carolina over Georgetown. I was 7/8 in the Elite Eight, missing only Oregon. I always like to fill out a couple "wacky" brackets that are a bit more loaded with upsets than what I really expect to happen. This year was pretty much a "chalk" year, so the "wacky" brackets did not do so well: 45-18 and 43-20. I stuck with Ohio State over Florida in both of those brackets because I felt pretty strongly that those two teams would meet in the championship game.

March Madness was fun, but the NBA playoffs, culminating in the NBA Finals--that is where it's really at!

posted by David Friedman @ 9:52 PM



At Wednesday, April 04, 2007 6:02:00 AM, Blogger vednam said...

Charley Rosen is a good example of what you have described as a "wack analyst".

I bought his recent book about the 1971-72 Lakers expecting a good read. I never got past the first few pages. While summarizing the Lakers' past playoff woes, he described Game 7 of the 1969 finals as follows.

"Indeed, the Lakers were in firm control of the game when Chamberlain committed his fifth personal foul late in the third quarter. Coach Butch van Breda Kolff immediately sent Mel Counts, a lanky seven foot jump shooter, in for Chamberlain. Unfortunately, one of Chamberlain's most cherished personal records was his never having fouled out of a game. So when van Breda Kolff called for Chamberlain to reenter the fray midway through the fourth quarter, the big man refused, mumbling something about an aching knee. Infuriated by Wilt's monumental selfishness, van Breda Kolff vowed to keep Chamberlain on the bench and win the game, and the championship, with Counts.
As the game raced toward the wire with the Celtics relentlessly eating into the Lakers' lead, Chamberlain approached his coach and asked to return to the action. But the always stubborn van Breda Kolff refused, and Chamberlain sat on the bench for the duration."

There's so many lies in that paragraph that they couldn't have been a few innocent errors. Rosen must have deliberately made up the story to put down Chamberlain, against whom he apparently has a vendetta. My immediate reaction was "how can this guy blatantly write such crap and get away with it?". It's even more sad because Phil Jackson referred to Rosen as one of the NBA's "foremost historians" in the foreward.

The facts (which can be verified in every non-Rosen account of the game, as well as the film of the 4th quarter, which I've seen) are that the Lakers were down big (far from leading) in the 3rd quarter when Wilt picked up the foul. Wilt continued to play until about 5 minutes left in the 4th with the Lakers rallying, when he hurt his knee and took himself out of the game. Wilt tried to re-enter a minute later but was denied by the coach, who had feuded with Wilt all season. It was not the Celtics coming back from behind and cutting a Laker lead. The Lakers in fact trailed the entire 4th quarter.

I wonder how this guy is going to re-write Kobe's (who he seems to hate almost as much as Wilt) career in 20 years.

At Wednesday, April 04, 2007 3:45:00 PM, Blogger David Friedman said...

You are correct about those passages in Rosen's book about the 1972 Lakers. Sometimes I agree with Rosen and sometimes I disagree with him. On the issue of NBA versus NCAA basketball I think that he is 100% on target. From what I've seen, his take on Kobe seems to be a mixed bag, sometimes positive and sometimes negative; perhaps I missed an article where he took what you seem to think are gratuitous shots at Kobe.

I've rarely found any writer or commentator who I agree with on all subjects, so I quote them/praise them when I agree and criticize/correct them when I think that they are wrong.

For instance, overall, I think that Wilbon does a pretty good job but on the Patrick podcast he had his facts wrong, which led to some wrong conclusions (as I discussed in the post about wack analysis).

At Friday, April 13, 2007 5:58:00 AM, Blogger vednam said...

Making a few factual errors here and there doesn't mean someone's analysis should be disregarded. However, Rosen's account of game under discussion is so far from the truth and so twisted to make Wilt look as bad as possible, that it's obvious that he made it up to try to tear Wilt apart.

Rosen almost always has something negative to say about Wilt whenever he comes up (and the negativity is usually more character assasination than a criticism of Wilt's skills). This is just the cherry on the top. When someone obviously strays so far from the truth, it comes of as calculated, cheap, and pathetic. It also makes me question Rosen's integrity as a writer and value his analysis much less than I otherwise would.

At Friday, April 13, 2007 6:47:00 AM, Blogger David Friedman said...

You make some valid points about Rosen and Wilt. I do wonder why Rosen is so negative about Wilt and I have noticed that he does not always have his facts straight in this regard.

I tend to consider everything I read with a critical eye--as you can probably tell by some of my posts--so when Rosen says something I double check the facts, but I do that with everything I read. Still, Rosen has played college ball and coached in a variety of settings, so I find his commentary interesting, even though I don't always agree with it. On the specific issue of NBA versus NCAA, I agree completely with Rosen and I think that me makes his case well, with no factual errors or obvious bias.

You are right that a minor factual error does not necessarily discredit an author completely. What I am speaking of is not a simple typo or minor mistake but something at a level that makes the whole article (or podcast) faulty. When Wilbon says that he looked up Jordan's stats compared to Kobe's and then gets everything wrong that is hard to take seriously. When someone tells me that a 32/16 game or a 28/22 game is tougher to do than a 65 point game, I prove that either of those is in fact much more common than the 65 point game and then instead of acknowledging that he is wrong he simply switches the subject, that is bogus.

I understand that you feel that Rosen's statements about Wilt rise to that level and I am not disagreeing with you. Everyone seems to have one or two personal blind spots and whatever Rosen's problem with Wilt is does not seem to extend to other basketball subjects, about which I think he writes very well.


Post a Comment

<< Home