Anything Is Possibuhhhhhl: Boston Celtics Will Repeat As NBA Champs

According to Kevin Garnett, "Anything is possibuuuuuuuuuuhllll!!!" Which is why the Boston Celtics will repeat as NBA champions.

Some people like to make predictions based on what they think will happen. They watch the injuries, check the depth charts, call their sources, in order to find as much information as possible to make the most accurate predictions.

Others pick their favorite team to win the title.

I, on the other hand, am using two things to predict what'll happen this year: Last year's win total, and three different win estimators—Pythagorean wins using the pace exponent, wins based off the Correlated Gaussian Method, and a schedule—and home/road-adjusted CGM, which is difficult to explain.

I found the weighted mean of those four (40% adjusted CGM, 30% regular CGM, 20% Pyth, and 10% real wins) to find average expected wins. Then, and this is hard to put in words, I found one-fourth of the difference in expected wins and real wins, and added that to expected wins; regress that number 10 percent to the mean, and you have your 2008-2009 projected wins.

(Here's a quick example: Toronto had 41 wins last year, and the weighted mean win total was 47.7, a difference of +6.7. One-fourth of that is 1.675; add that to the expected wins from last year and you have 49.375, which we'll call xWins. Regress 10 percent to the mean by taking 90 percent of xWins and adding that to 4.1, and that's the projected wins for 2008-2009. I also subtracted .0117 to each team's projected wins in order to get the total amount of league-wide wins to 1,230, but that made little difference.)

Okay? Okay.

Without further ado, here's the projected wins of all teams in the NBA. Teams in bold are projected to make the playoffs. I didn't base the seeds off of the division winners like the NBA does, if you were wondering.

Seed	Team	Division	Proj. Wins	Proj. Loss
East 1	Boston Celtics	Atlantic	63.4	18.6
West 1	Utah Jazz	Northwest	57.4	24.6
East 2	Detroit Pistons	Central	56.3	25.7
West 2	Los Angeles Lakers	Pacific	56.0	26.0
East 3	Orlando Magic	Southeast	53.9	28.1
West 3	Dallas Mavericks	Southwest	52.2	29.8
West 4	San Antonio Spurs	Southwest	51.4	30.6
West 5	Phoenix Suns	Pacific	51.3	30.7
West 6	Houston Rockets	Southwest	50.7	31.3
West 7	New Orleans Hornets	Southwest	50.5	31.5
West 8	Denver Nuggets	Northwest	48.6	33.4
East 4	Toronto Raptors	Atlantic	48.5	33.5
West 9	Golden State Warriors	Pacific	45.2	36.8
East 5	Philadelphia 76ers	Atlantic	43.0	39.0
East 6	Washington Wizards	Southeast	40.2	41.8
East 7	Cleveland Cavaliers	Central	39.4	42.6
East 8	Indiana Pacers	Central	38.7	43.3
West 10	Portland Trail Blazers	Northwest	36.9	45.1
East 9	Atlanta Hawks	Southeast	36.2	45.8
East 10	Chicago Bulls	Central	35.1	46.9
West 11	Sacramento Kings	Pacific	35.1	46.9
East 11	Charlotte Bobcats	Southeast	30.9	51.1
West 12	Memphis Grizzlies	Southwest	28.7	53.3
East 13	New Jersey Nets	Atlantic	28.6	53.4
East 12	New York Knicks	Atlantic	28.6	53.4
West 13	Minnesota Timberwolves	Northwest	26.4	55.6
East 14	Milwaukee Bucks	Central	25.8	56.2
West 14	Los Angeles Clippers	Pacific	24.7	57.3
East 15	Miami Heat	Southeast	23.6	58.4
West 15	Oklahoma City Thunder	Northwest	22.7	59.3

So you understand everything so far? Here's where it gets interesting.
Using the projections above, I created a playoff bracket, and using Bill James' Log5 formula , found the odds of each team winning their series. Of course, the teams with the highest winning percentage will win the Log5 playoffs, so the odds make it easier to tell the difference between a Celtics-Pacers series and a Spurs-Suns series.

Here's the bracket (cut off on the right side is "NBA Champions: Boston Celtics"):

Incidentally, I actually think the Jazz will make the finals against the Celtics (but the Jazz will win), if I were to predict it without the numbers. The Hornets were severely underrated if you look at their seed, but check out the No. 4 through 7 seeds in the Western Conference—less than one win separates them. You can change the order of those four, and the odds of each team winning or losing wouldn't change too much (if at all).