The final piece of the Smackdown predictions is to predict a win total for each team in the league. As I noted in my first pick, I’m at a bit of a disadvantage here. Arturo has developed a sophisticated model to make his picks; other people use other measures to predict team performance. All I have is a blog and a dream. So what am I going to do? What everyone is supposed to do in tough times: fall back on my training. Mine is in psychology, so I’m going to lean on the wisdom of crowds. The idea is that the average of a group of people’s opinions or guesses will be closer to the truth than any particular person’s guess. On one hand this sounds like a cop-out because I’m not making my own picks; on the other hand, it’s scientifically justified and I know someone who’s done related research!
The main requirements for the wisdom of crowds to work is for the decisions to be based on at least somewhat separate information and for the decisions to be independent of each other in that one person’s decisions shouldn’t be influenced by someone else. In my case, I’m going to draw together picks from various sources and average them together. Those sources are going to be win shares and statistical plus/minus, the current guesses listed on the Smackdown page, Arturo’s previous choices (unadjusted for Mike Miller’s injury or strength of schedule) and the ESPN folks. The ESPN guys don’t have win numbers, just order in each division, so I made some executive decisions. I also had to make some executive decisions in terms of adjusting people’s projections to add up to 1230 wins; come on guys. But importantly each of these sources should be independent and draw on separate pieces or kinds of information, so I should be on solid wisdom-of-crowds ground. It’s possible that this assumption is false for the Smackdown folks, who are all using wins produced at some level (I assume), but hopefully they all have different feelings about who will get how many minutes and the like.
Without further ado, my picks (East then West by seed):
1: Miami (66) Portland (59)
2: Orlando (55) Lakers (57)
3: Boston (52) Spurs (52)
4: Chicago (51) Dallas (49)
5: Bucks (45) Thunder (48)
6: Atlanta (45) Denver (47)
7: Charlotte (37) Utah (44)
8: Cavs (37) Phoenix (44)
9: Knicks (36) New Orleans (44)
10: Nets (34) Warriors (43)
11: Indiana (28) Houston (41)
12: Detroit (27) Sacramento (37)
13: Philly (25) Memphis (36)
14: Toronto (25) Clippers (24)
15: Washington (19) Minnesota (23)
I think something else interesting I can do with the rankings is look at the standard deviation in wins by team to see which teams are the divisive ones; a big standard deviation means that there’s a wide range in the predicted wins for that team. The Wizards are the runaway winner, ranging anywhere from 37 wins (what I figured ESPN would put for a team in Washington’s position) to 4 (Arturo). The Warriors are also a bit above the rest of the crowd, ranging from 26 (ESPN) to 50 (Nerd Numbers). The most consistent teams are the Hawks (43 to 47 wins) and Knicks (34 to 40).
On top of the predictions above, here’s a reminder/update (with links) of my picks for individual awards:
Most Improved, media: Robin Lopez. WP: Kevin Love
MVP, media: Dwayne Wade. WP: LeBron
ROY, media: Blake Griffin. WP: Griffin
Sixth Man, media: Marcus Thornton. Evan tells me in the comments that DeJuan Blair will be starting, so I’m going to change my pick. WP: Lamar Odom. Everyone from my original post is hurt (Miller) or apparently a starter, so I’m going to cross my fingers and hope that Bynum holds up for at least 30 games and Odom qualifies for Sixth Man.