Hollinger's 2010 Playoff Odds

[nickagneta]

Faf, you're a statistician. Can you honestly tell me you looked at those numbers and a bell didn't go off in your head saying, "THERE"S SOMETHING WRONG HERE!!!"?

You instantly went to defending him and even when it's been shown that there was something major wrong, you're still trying to defend an absurd point. How can you, who knows and works with numbers for a living, sit there and say that given the data he was claiming to be inputting, that we could know whether a normal random number generation or normal regression to the mean was impossible to produce the numbers he was coming out with in ANY METHODOLOGY?

Come on Faf. Think about that. Look at the input. Take a look at the numbers that was being output. It was impossible. I don't care what the normal regression to the mean or number generation methodology he was using.

One thing I can say about owning a business. When it's your money on the line and the numbers aren't jiving, 99.9% of the time, if you have any mathematical background, you are right. After a while, seeing numbers and knowing if they are right becomes second nature.

[Drucci]

I don't give any credit or value to Hollinger's stats but I've followed your debates a little bit so maybe you will be glad to learn that he has fixed his system according to his twitter :    @johnhollinger OK, we've fixed the hamster. Champs odds as of today: Celtics 28.3%, Hawks 17.1, Lakers 16.0, Magic 9.3, Cavs 8.8.

http://sports.espn.go.com/nba/hollinger/playoffodds

[MBunge]

I don't give any credit or value to Hollinger's stats but I've followed your debates a little bit so maybe you will be glad to learn that he has fixed his system according to his twitter :    @johnhollinger OK, we've fixed the hamster. Champs odds as of today: Celtics 28.3%, Hawks 17.1, Lakers 16.0, Magic 9.3, Cavs 8.8.

http://sports.espn.go.com/nba/hollinger/playoffodds

If Hollinger has a bias against the Celts, I suspect it prevented him from recognizing that his numbers were screwy, rather than screwing up the numbers in the first place.

Of course, when your stats tell you the Hawks are almost twice as likely to win the title as the Magic...you do have screwy numbers in the first place.

Mike

[Celtics17]

I dont know how to paste but this is to the guy with the moderator thing next to his name. Ok, so, you are right we are 25% of the way through the season, why dont you enlighten me with some of your statistical analysis. For your information, I happen to know a pretty fair amount about it myself having had a few statistics couses, an MBA with an emphasis in corporate finance, as well as B.S. in accounting and the ability to multiply 3 digit numbers in my head. But, since you so enlightened me that you can actually tell a lot about the season by analyzing it maybe you would be so kind as to share some of  your findings so far. 

[JHTruth]

Hollinger's system is really an advanced application of Monte Carlo analysis. He applies a probability distribution to different variables he has deemed important to a team's performance and then uses a program like Crystal Ball or @Risk to simulate the results of a season by running a "season" around 5000 times. Then he takes the mean of those 5000 iterations and those are the "expected" results. The "best" season is the best result received in 5000 simulations, the "worst" the lowest.

In practical terms, the C's are immensely efficient on both sides of the ball and it makes a lot of sense that a statistically driven system loves them...

[Roy Hobbs]

In practical terms, the C's are immensely efficient on both sides of the ball and it makes a lot of sense that a statistically driven system loves them...

Right, which is why his results as of yesterday (when the Celts were ranked 1st, but had the 6th highest probability of the title) should have raised giant red flags, for both fans and Hollinger.

[nickagneta]

In practical terms, the C's are immensely efficient on both sides of the ball and it makes a lot of sense that a statistically driven system loves them...

Right, which is why his results as of yesterday (when the Celts were ranked 1st, but had the 6th highest probability of the title) should have raised giant red flags, for both fans and Hollinger.

Exactly. TP.

[indeedproceed]

Currently: 40.6% Finals, 28.3% Champs

Talk about a turnaround.

Funny how I pay a lot more attention to Hollinger when we're winning. Just plain weird.