2012 predictions

[Mike G]

Posey wasn't gonna add much anyway.

Code: Select all

Ind               2012 projections    2011 numbers
.   Pacers       e484    min    eW   e484   mpg   min   eW
David West       1.72   2800   9.9   1.79   35   2454   9.1
Danny Granger    1.60   2800   9.3   1.47   35   2764   8.4
Roy Hibbert      1.35   2400   6.7   1.35   28   2246   6.3
Darren Collison  1.22   2600   6.6   1.18   30   2357   5.7
Tyler Hansbrough 1.05   2000   4.3   1.06   22   1535   3.4
George Hill       .80   2000   3.3    .74   28   2149   3.3
Paul George       .75   1800   2.8    .68   21   1262   1.8
Dahntay Jones     .56    800    .9    .54   13    590    .7
Brandon Rush      .39   1000    .8    .37   26   1760   1.4
Jeff Foster       .47    800    .8    .57   17    939   1.1
A.J. Price        .35    600    .4    .42   16    799    .7
Lance Stephenson  .20    200    .1    .12   10    114    .0
James Posey       .10      0    .0    .13   17    839    .2
.   totals             19800  45.9              19808  42.0

This is still per 82 games.
Converting individual equivalent-Wins to eXpected wins:
xW = eW*2 - G/2 = 91.8-41 = 50.8

There's no big predicted improvement (or none) for several guys, just more minutes. These minutes are taken from less-effective players last year: TJ Ford 800 min @ .36 e484, Sol Jones 500 @ .06, and Posey 800 @ .13. Weak links remaining get fewer minutes.
Dunleavy and McRoberts were both good (total 3200 @ ~1.00), but West is lots better.

[Mike G]

Blazers are "not optimistic" about Oden. I'd just seen he had gotten a contract extension.
I gave Camby just 24 mpg over the whole season. Probably should downgrade.
Mills was about 1 win over replacement.
As I said, it's very preliminary.

Battier is 33 this year, and his already marginal outputs seemed to drop off the table last year. In playoffs, he looked almost out of gas.

[mystic]

Mike G, I used your minute predictions for the Pacers and the result would be:

Using my SPM values from last season only: +2.69 SRS
Using my SPM values from last season with last season minute numbers: +1.32 SRS
Using SPM from the previous 3 seasons (3x 2011, 2x 2010, 1x 2009): +0.94 SRS
Using SPM from the previous 3 seasons (3x 2011, 2x 2010, 1x 2009) with last season minute numbers: -0.20 SRS

For players who haven't played in 2010 or 2009 I assumed the average SPM value for players without minutes in the previous year of -1.52.

So, translating SRS into wins while assuming a SOS of 0.5 we get between 40 and 49 wins for the Pacers or between 33 and 39 respectively for a 66 game schedule.

Using the 6 year prior informed RAPM from Jerry, we would get -0.92 with the presented minutes distribution. That would translate to 38 wins in 82 games, 30 wins in 66 games.

[huevonkiller]

Anyone the Heat uses off the bench will be much better. No Bibby, Magloire, Dampier, Big Z.

Battier is better than Miller (who didn't even play) and will be used at the 4 sometimes. Chalmers and Cole are much better than Bibby. I think Miami will be fine. I could see Wade getting hurt and their bigs struggling a little, but Bosh has come back stronger. Z is an awful player Mike, especially defensively.

Also Joel Anthony is a horrible player. If the Heat use LeBron/Battier/Haslem/whatever next to Bosh, they'll have a much better team.

[Mike G]

Mike G, I used your minute predictions for the Pacers and the result would be:

Using my SPM values from last season only: +2.69 SRS
...
Using SPM from the previous 3 seasons (3x 2011, 2x 2010, 1x 2009): +0.94 SRS
.

The 2nd one is kind of a backward or reverse progression, and you can see that it gives a worse prediction.
In other words, maybe it shows these players are getting better, from 3 years ago, to 2 years ago, to last year. And then this year, they may well be better again.

The Pacers' minutes-weighted age would be 26.4 years, as I guessed them. On average, a 26 YO improves over his 25 YO play.
A +2.7 pt-diff should be 49 wins in 82, yes? And 51 is just a little overall improvement expected.

Oh, and I haven't assigned any minutes (or rates) to rookies yet.
Last year's rookies are expected to be .10 eW/484 better this year. This is subject to further evaluation.

[Mike G]

I have the Heat at 46-20, equivalent to 58 wins in 82 games, same as last year.
Ilgauskas was productive for most of the year, and he's gone.
Battier and a healthy Haslem should help, offsetting expected declines by Miller, Jones, Howard, and the loss of Bibby.

General note: In this century, looking at 288 rookies who had 500+ minutes and also 500 minutes in year 2, avg increase in WS/48 from yr 1 to yr 2 is .013 .
So I'm expecting improvements in eW/484 to be +.13 for last year's rookies.

[Mike G]

Average value of rookies, equivalent wins contributed, by draft position, over the last 4 years.
The values are severely smoothed over, averaging in values from above and below each draft position.
1.00 eWins per 484 minutes is average production.

Code: Select all

dp    mpg   eW/484   eW82    eW66
1     29.1   1.35    6.63    5.34
2     27.2   1.15    5.30    4.26
3     25.2   1.00    4.26    3.43
4     23.3    .84    3.31    2.66
5     21.5    .75    2.71    2.18

6     20.3    .71    2.45    1.97
7     18.8    .70    2.23    1.79
8     18.2    .69    2.13    1.71
9     17.5    .68    2.02    1.62
10    16.5    .67    1.87    1.50

11    15.4    .67    1.75    1.41
12    14.3    .65    1.58    1.27
13    13.3    .63    1.43    1.15
14    12.5    .62    1.31    1.05
15    11.4    .59    1.14     .92

1625  10.0    .52     .89     .72
2640   7.5    .47     .60     .48
4160   3.9    .36     .24     .19

The bottom 3 lines are for draft picks 16 thru 25, 26-40, and 41-60. There just seemed to be little difference within these tiers.
Minutes include zeroes for those who never played.
MPG are season minutes divided by 82 . eW66 are eWins expected this season.

[Mike G]

Having put in the rookies and a few recent updates, the standings now look like this:

Code: Select all

West   W   L     W%       East   W   L     W%
SAS   42   24   .641      Chi   50   16   .752
Dal   41   25   .618      Mia   47   19   .712
LAC   41   25   .615      Bos   44   22   .663
LAL   40   26   .599      Orl   42   24   .641
Okl   39   27   .584      Ind   38   28   .575

Mem   38   28   .576      Phl   33   33   .507
Por   38   28   .572      Atl   33   33   .495
Den   33   33   .497      NYK   33   33   .494
Min   32   34   .489      Mil   30   36   .452
GSW   31   35   .471      Cha   24   42   .361

Sac   31   35   .470      Det   23   43   .347
Uta   31   35   .470      Tor   22   44   .338
Hou   31   35   .466      Was   22   44   .338
NOH   22   44   .338      Cle   21   45   .311
Phx   21   45   .324      NJN   19   47   .284

Is there an abnormally low player attrition rate this year? -- because it seems like a lot of rosters are jam-packed, and I'm guessing almost everyone gets fewer minutes (per game) than they got last year.
There will be 990 games won this year, and even without forcing the best players to take on more minutes, teams were predicted to sum to 1110 wins. So I had to multiply all by .892 to end up with a .500 season.

I rounded to integral numbers of wins and losses, since no one wins a fraction of a game; but the W% retain the actual expected wins/66.

[Crow]

Currently you have 7 teams projected to win .600+. That would be lower than in the last 4 years and tied with the 06-07 season. The 05-06 season only had 6 and was the lowest since the mid-80s. Last season there were 9.

[J.E.]

Those projections look better, in my opinion. Definitely agree that CHA/DET/TOR/WAS/CLE/NJN will be the bottom of the east. Pretty much agree with all EAST projections.

Miami may be stronger than Chicago on paper, but potential injuries will hit them harder because of less depth.

Orlando will be hard to predict because of the D12-drama, but if they deal him they probably won't go into rebuild mode (as in "deal him for draft picks") but instead will try to contend. Thus, an above average projection for Orlando makes sense even if D12 is playing somewhere else.

Curious RAPM story: The Nets waived their best player last March, and so far nobody has picked him up. He is, of course, Quinton Ross, second best defender in the league. He played a whopping 350 minutes last season, during which the Nets held their opponents to 93 points per 100 poss. The Nets, over the entire season, held opponents to 110 points per 100p. Quinton's defensive On/Off numbers make Kevin Garnett jealous.

Now to the West:
Has Billups talked to the media again after the Clippers picked him off waivers? He seemed super-pissed about being amnestied and warned teams to pick him up. CP3 will probably help his decision to play. Even so, RAPM thinks Billups's defense last year was horrible, so much that he was barely positive overall. Jordan is seen as negative, Griffin as barely positive. RAPM also hates Mo Williams and Foye and their low post depth is really bad. Even with Paul fitting perfectly with Griffin and Jordan, are they good for 3rd in the west? I think not

Minnesota is still too high, I think. They have the same roster that sucked last year plus two rookies. Rookies generally aren't positive contributors. I agree they should improve from last year, but not to almost 0.500 in the west

Sacramento was at -5 last year, so far haven't resigned Dalembert, just signed Travis Outlaw (ugh), and, at the end of last year, traded for JJ Hickson (uuuuuugnh). I think they'll be at the bottom of the west.

Utah without Deron Williams and Kirilenko. I don't like it. They went 7-23 without Williams last year. If they could find someone to take over Raja Bell's minutes and if Harris can somehow stay healthy, they could be not horrible, but 31Ws seems high either way

Phoenix will also be hard to predict, because so much of it depends on the health of Steve Nash. If Nash can stay somewhat healthy and if Morris can take over Warrick's minutes I think they could almost be at +/- 0, so I think 21W is a little low

Currently you have 7 teams projected to win .600+. That would be lower than in the last 4 years and tied with the 06-07 season. The 05-06 season only had 6 and was the lowest since the mid-80s. Last season there were 9.

Since these are projections I'm sure it's fine to put such a low number of teams at .600+. Yes, there will probably be more teams to win over .600 but we don't know which ones, yet.

[EvanZ]

I'm curious what's your take on the Warriors? Both RAPM and ezPM suggest about 26-27 wins (40%).

[J.E.]

I'm curious what's your take on the Warriors? Both RAPM and ezPM suggest about 26-27 wins (40%).

I don't see a reason why they should be much different from last year. Less Radmanovic is bad, but more Udoh is good. Everyone else that is now gone from last year can probably be replaced rather easily. In my opinion anything between 27 and 31W is probably fine
Btw, RAPM probably hates Brandon Rush for his impact on team rebounding http://stats-for-the-nba.appspot.com/of ... rebounding. How a guard can such a large negative impact on offensive rebounding, I don't know. But it is what it is

[EvanZ]

Wow, that seems odd. Somebody has to be last on the list, I guess, but you wouldn't expect it to be a SG.

[Mike G]

There are but 4 wins separating the top 6 teams in the West, in this 2nd pass. The Clipps could be one. In my projection, I figure Paul for 32 mpg, Billups 23, and Mo Williams 21: total 76 minutes from these point guards (all of whom are also scorers).
The Clipps also added Caron Butler (27 mpg), and I suppose Jordan, Griffin, Foye, and Bledsoe will be as good as or better than last year. Yet most or all will be getting fewer mpg, due to the talent.

The Wolves add a #2 and a #5 pick. These tend to get 27 and 21 mpg respectively; I cut them to 26 and 20. JJ Barea and Brad Miller can only get about 2/3 their 2011 minutes (with better teams), Love drops to 33 mpg. Anthony Randolph was good for 23 games, and if they have him all year, his mpg drop from 20 to 15.

The Kings are also loaded with young players who should be good. Two or more of Thornton, Evans, Cousins and Thompson should improve. They add Chuck Hayes, got Salmons back, Hickson seems to be productive. They sure could use a point guard, and it might be Tyreke?

The Jazz were 7-9 with Devin Harris, so they were 0-14 with neither him nor D-Will.
They add 2 rookies I guess are worth 40 mpg between them and 3.5 wins on the year.
I gave Raja 9 mpg (down from 31) with Jeremy Evans and Favors improving, just 14 mpg combined for Okur and Josh Howard.

The Suns have a bunch of limited players, other than Nash who is still great. I figured him at 30 mpg. Gortat and Lopez are good, but they share one position. Everyone else has peaked or is borderline unusable.

Warriors' rookies given 33 total mpg should bring 2.4 wins. Lee should bounce back, Curry is rising, Kwame is something. However, I've assumed good health from all major players. They could add some free agents.

[J.E.]

JJ Hickson had the worst On/Off on that godawful Cavaliers team. Maybe he looks OK in some BoxScore stats* but he makes so many things wrong on the court. He may be one "poster child" for +/- statistics

*his TS%, which is arguably the most important statistics you can get from a BoxScore, was awful. Yet he had a 25.5% USG% for a whopping -0.2 Offensive WinShares

The Jazz were 7-9 with Devin Harris

Good catch. Four of those wins came with Kirilenko in the lineup though