My favorite research tool,
Basketball-Reference.com, has posted
a set of predictions for the upcoming season based on expected win shares and a formula to estimate playing time and time lost to injury. After the jump, we'll take a look.
B-R.com has the Sixers at 38.7 wins, good for the 8th and final spot in the playoffs. They have Cleveland leading the East with 50.5, the Bucks bringing up the rear with 33.6. They used formulas exclusively to determine minutes played, so these numbers are really trivial, but here's a breakdown of their Sixers predictions:
DP stands for draft position, by the way. The thing that jumps out at me right away are the minutes played. They have Iguodala playing almost 400 fewer minutes than he did last season, Willie Green as the 2nd busiest reserve. So I figured we'd preserve their win shares/48 rates and then apply them to what I think the minute distribution will look like:
The interesting part of their projections, at least to me, is the variance in win shares/48 minutes. Basically, how they project players to perform relative to their performances last season (or over their careers). Here's a brief look:
So, B-R.com is projecting improvement from Lou, Dalembert, Thad, Kapono, Smith, Carney and Ivey. A regression by Iguodala, Brand, Speights, Green and Brezec.
By no means is this gospel, but I appreciate any attempt at projections using advanced stats.
Interesting that their 'finely honed projection' and my ideas work out to the same thing - about 500 :)
Just out of curiosity where do they have the rest of the atlantic?
Toronto: 43.4
New York: 37
New Jersey: 36.2
Interestingly, Miami: 35.9
Miami is more hype than substance I think - but we'll see.
So 3rd place is the prediction based on their numbers - sort of what i thought would happen...but like i've said repeatedly, i'm not looking at W-l this year as a big concern as i am progression of players
Miami as a whole may be all hype but Wade is a hell of a lot of substance on his own. Him healthy almost guarantees that team wins more than 35 games no problem.
At this point no one or no statistical measure is suggesting the Sixers are better than a .500 team. And many are saying we are lottery bound with Miller and either an unproductive or again hurt Brand.
I'm not sure there's a reason a statistical measure should assume Brand is going to be either unproductive or again injured. To play devil's advocate for a minute, the Sixers are essentially replacing two starters from last season. Andre Miller and Willie Green for Elton Brand and Lou Williams. Brand needs to provide about 75% of the production he did in his last full season to better Miller's production last season (speaking of win shares here). In limited minutes, Lou has topped Willie every year he's been in the league in win shares.
I think it's a conservative estimate to say the Sixers will be back where they were last season, especially considering they've addressed their biggest weakness as a team, half-court offense.
I'd have to agree -- if all the other teams stayed the same. Unfortunately, just about everyone else in the east got appreciably tougher to beat than they were last year.
The elite teams feast on the gimmie teams and take the majority of the tough games. Sixers not an elite team yet and can't expect to be this season. With fewer gimmie games and the tough games largely not any easier, I don't see the changes that overwhelming to make up the new degree of difficulty. Not yet. Not yet.
I'm not sure I agree that all the other teams got appreciably better.
Cleveland - Yes
Boston - Maybe, depends on health
Toronto - Debatable
Miami - No
Orlando - Debatable again
New Jersey - Debatable
Washington - Pretty much exactly the same as they were at this point last season (when they thought they'd have Arenas), with a new coach.
Knicks - Equally pitiful
Charlotte - Probably worse
Chicago - Worse
Detroit - Your guess is as good as mine
Hawks - A touch better, but Crawford is not a good player
Milwaukee - Probably worse this year
Indiana - Probably worse
Have any of these teams improved enough to leapfrog the Sixers? Toronto, New York, New Jersey, Chicago, Detroit, Milwaukee, Charlotte, Indiana, Washington? Maybe Toronto, depending on whether defense matters at all. No one else, IMO. There weren't really any earth-shattering improvements in the East, outside of Cleveland, potentially.
One of those statistical attempts where the numbers lack the precision to be very accurate. Sure if all of the variable even out it might be a decent prediction.
But given all of the variables they are dealing with it seems like they should say something like:
38.7 wins +/- 5 wins within 80% probability
Its sort of like saying I estimate there are 38,759 beans in a jar, +/- 5,000.
Completely agree. Everyone is sort of scrunched together as well when, no one w/ fewer than 33 wins in the East?
I think the minute distribution has a lot to do with it, assuming starters only play 60% of the minutes seriously hurts the estimations for teams w/ guys like LeBron who play way more than that.
I just read Tom Moore report this quote from Jordan about Lou:
"We want him to attack. We want Lou to be Lou. We're not going to walk the ball up with him. We're going to give him every opportunity to be a 38-minute-a-game player."
Intersting that they would just flat assume starters only play 60% of the minutes - that's a flawed methodology when there's evidence to the contrary (strong evidnece) over more than one season for a guy like Lebron (or Iguodala)