How'd "Wages of Wins" Midseason Predictions Go?

[shrink]

I like WoW because of two reasons. First, its run by economists, and they value numerical formulae over our emotional attachment to our own ideas. Second, its because they're usually right.

Still, its always good to go back and check the validity of any information source. Fortunately, the WoW guys made predictions right after Al's injury.

http://dberri.wordpress.com/2009/02/12/ ... minnesota/

Had Jefferson stayed healthy, the team’s efficiency differential - offensive efficiency minus defensive efficiency - suggested this team might win 31 games (it would be close to 40 if Gomes was playing like he did last year). With Jefferson gone, though, this team will struggle to win 10 more times this year. And that means this team will finish closer to 25 wins. In other words, the loss of Jefferson is going to cost this team about five wins in the final standings.

The Wolves won 7 more times over their last 30 games, to finish 24-58

And the really bad news in this story is that there isn’t the usual silver lining to the team’s accumulation of additional losses. When we think of the draft, the loss of Jefferson will probably not change Minnesota’s draft position much. Currently Minnesota has the 6th worst record in the league. If the T-Wolves only wins 5-7 more games, they may still have the 6th worst record. It’s possible they might get passed by Oklahoma City or Memphis, but it’s possible that won’t happen also.

Bullseye. MIN finished 6th (24-58), but MEM (24-58) tied them on the last day for a 5th-6th finish. OKC finished 23-59.

So let’s summarize the news for Minnesota.

1. The loss of Jefferson means Kevin Love will produce more wins. He will likely lead all rookies in Wins Produced, although that was probably true before this happened.

True. Love improved from 6.4 WP to 9.97 WP, which was #1 among rookies

http://www.wagesofwins.com/Rookies2009.html

2. The team will lose more often. Instead of winning more than 30 games, Minnesota will probably finish closer to the 25 win mark.

True. Over the last 30 games, they went 7-23, and finished 24-58

3. These additional losses will probably not change the team’s draft position (barring a lucky lottery bounce).

True. MEM won its last game to finish with the same record, and the other team they mentioned, OKC, finished 23-59

If we put the whole picture together, the loss of Jefferson just means more losses and not much gain. And really that’s too bad. Kevin McHale - the much maligned architect of this roster - made a bold move last summer. In exchange for O.J. Mayo, McHale got both Love and Miller. Both players are much more productive than Mayo, so this move was quite good. Unfortunately, the loss of Jefferson means we are not going to see much overall improvement in Minnesota. So the one brilliant move McHale has made might ultimately go unnoticed (at least, outside of this forum).

True?

[Worm Guts]

I like WoW because of two reasons. First, its run by economists, and they value numerical formulae over our emotional attachment to our own ideas.

I don't think economists are less emotional than anyone else who does statistical analysis. Anyone who does statistical analysis is attempting to remove subjectivity.

[shrink]

Some people knock the "Wins produced" stat, but it has been historically shown to be extremely accurate in calculating a team's eventual number of wins. It did it again here.

[Jonathan Watters]

Yeah, WOW just doesn't make sense at all. It doesn't take into account the fact that the shot clock forces a team to take a shot within 24 seconds. It basically penalizes players for shooting and rewards the rebounder, even though the shooter who missed often did nothing wrong and the player that grabbed the defensive rebound did nothing he shouldn't be expected to do. This causes WOW to ridiculously overvalue even your average rebounding big man. Every single "call" Berri touts is the same thing, replacing a guard with a big man. And a team with 5 big men is never going to win any games. So while the point about big men being important is a relevant one, it is simply a point - not a valid statistical model for evaluating individual players.

[shrink]

How do you disagree with the results, that it has very accurately predicted team wins for several years?

And why was I watching the internet every night for those last 2.5 months, when WoW predicted the final standings almost exactly?

[Jonathan Watters]

How do you disagree with the results, that it has very accurately predicted team wins for several years?

And why was I watching the internet every night for those last 2.5 months, when WoW predicted the final standings almost exactly?

I guess the team argument is more valid than the individual argument. Because if that forced shot at the end of the shot clock is poor, that is a sign of a negative for the team's offense. And the rebound by the team that forced the poor shot is a sign of a positive for that team's defense.

But at an individual level, it isn't even remotely close to accurate. The shooter that took the poor shot may or may not be responsible for the team not getting a good shot in 24 seconds. The rebounder may or may not be responsible for his team forcing the bad shot.

I always forget about the team evaluation stuff though...just don't pay attention to the individual stuff. It is borderline meaningless. Berri likes Jefferson because he gets a lot of rebounds. That has very little to do with why he is the Wolves' most important player.