John Hollinger recently came out with his NBA Hollinger Playoff Odds for 2007-08 (I take no responsibility for that title). Using the power rankings that he developed for ESPN, the remainder of the season is simulated 5,000 times to “reveal the most likely win-loss record for each team — and how likely it is for each team to make the playoffs, win the NBA title, win the lottery, and so on.”

Something about the Hollinger odds struck me as… odd, I guess. Four teams currently have a 100% chance of making the playoffs: the Celtics, Detroit, Orlando, and San Antonio. Boston also has 100% chance of winning the Atlantic Division, and a 61% of winning the NBA finals. Let me repeat that. A 61% chance of winning the FINALS. Crazy things can happen in the course of a season, never mind the playoffs (just ask Phoenix) so those odds didn’t pass the smell test for me.

But what good is complaining if you’re not going to do anything about it, I say. So I went and created a better, stronger, faster version of the NBA Odds Report. It’s based on a Monte Carlo simulation and takes into account a team’s expected record using the Pythagorean 16.5 Method (and provided by ESPN), a team’s actual record as the simulation progresses, strength of schedule on a game by game basis, home court advantage, and the effect of back-to-back games. Research by people smarter than me and research by Kevin Pelton, who is much smarter than me, suggest that home court advantage in the NBA is on the order of 10% to even 20%. I went with a more conservative 5% in the simulation. I also gave teams playing in the back end of back to back games a 5% discount in their odds of winning a game. That also seems a little conservative, but again I didn’t want to go overboard.

Anyway, here are the results of the simulation based on 10,000 runs. I didn’t try to predict the playoffs out of fear that I’d end up with something crazy like Boston winning the finals 60% of the time. The simulation predicts that Isiah will get fired after losing to the Celtics by 60 points on January 21, 2008.*

I like those odds. I still couldn’t get Boston not to make the playoffs in any of the sims, likely due to the horrible Eastern Conference, but at least they don’t win the division 100% of the time. Hollinger’s odds say that Boston goes 70-12 or better 6.88% of the time, while in this sim, they only win 70+ twice out of those 10,000 games (0.02%). Basically, the simulation is predicting that all those back-to-back games will eventually catch up with Boston and make it difficult to match the Bulls record for wins in a season. I agree.

Every single team made the playoffs at least once, so keep the faith Seattle and Minnesota fans. Seattle made the playoffs five times, but never won the division. Minnesota only went to the playoffs once off their only division win. Minnesota also had the worst of the worst records, going 7-75. I didn’t save all the individual runs, so I have no idea how in the hell the Wolves beat Denver, Utah, and Portland for the division that one time.

Hollinger’s odds like Miami and Chicago more than mine, so he’s clearly giving them credit above what they’ve shown in their expected winning percentages this season. I think that makes sense if you’re trying to predict the future, especially for Miami now that Wade is back. That’s something I’ll try to incorporate in V2 of the simulation program if I get the motivation.

* Not really.