Bucks Diary

Thursday, June 05, 2008

Likeliest Outcomes for the 2008 NBA Finals


Whom the odds favor in this NBA Finals depends on which numbers you trust. If you go by the playoff numbers, the likeliest outcome is the Lakers winning in 5 games. If you go by the regular season numbers, the likeliest outcome is the Celtics in 7 games, just slightly ahead of the Celtics in 6 games.


Used Efficiency Numbers to Calculate the Odds

How did I determine the likely outcomes? I used the team's offensive and defensive efficiency numbers for the regular season, and for the playoffs, and then plugged them into this formula, and based upon the outcome, I just used the same calculations you would use playing craps (eg, the likelihood of two events occuring is derived by multiplying the likelihood of each event seperately occuring).

Did I factor in the Celtics home court advantage?

Yes, I calculated each set of numbers Game-by-Game, and according to that game's home team's "Home" efficiencies and away team's "Away" efficiencies. I also factored in the unusual 2-3-2 format.

Do the playoff numbers account for strength of opponent?

Yes, I used my "point value over average" calculation method to level the strength of each team's postseason opponents. All PVOA calculations likewise take into account every opponent's home and away efficiencies, and I also break the Finalists teams PVOAs down according to home and away results as well.


What is PVOA?

I would ask you to read this brief description. Basically I take the team's opponent's efficiency numbers and use them to calculate what the score of the game would have been if those numbers held. I then check the actual score. If, for instance, the Celtics opponents would have yielded 95 points in the given number of possessions in their normal game, then whatever the Celtics actual point total is above or below that is their "Point Value over Average". Its a useful tool for comparing two team's performances when they have not been facing a similar schedule.

Why do your playoff numbers make the Lakers prohibitive favorites?

Basically for this reason. The Lakers have been invincible at home in these playoffs, and pretty formidable on the road. Meanwhile, the Celtics have been fairly good at home, and frankly, below average on the road, even counting the Detroit series. In fact, when plugging my playoff numbers into Professor Berri's efficiency formula, the Celtics have been so bad on the road, and the Lakers so good at home, that the formula predicts the Celtics have virtually no chance of beating the Lakers in Los Angeles. Meanwhile, the Lakers have a 28.4% chance of beating the Celtics in Boston. That doesn't sound like much, but in fact it means the Lakers have about a 50/50 chance of splitting in Boston. And if they do, the odds would favor them winning the remaining 3 in LA.

Which numbers are more trustworthy?

I frankly do not know. Why have the Celtics been so mediocre on the road? And did they revert back to their old form in the Detroit series? Those are the key questions. My gut tells me the playoff numbers are closer to right, and that the Lakers will prevail. But, if you go by my formula, the Celtics probably should not have won any games at Detroit, and they won twice. So you never know.

0 Comments:

Post a Comment

<< Home