The BBD, Murphy, Krstic and Perk Debate

[Roy H.]

I don't see how there is really any debate at all.  It's clear we sacrificed the next two years of contention for the hope of building a contendor again in 2012.

Simple

  There's a debate because many people (if not most) would disagree with your statement.

Yeah, I don't like the trade because I think it was too risky, but I don't think we've sacrificed anything.  Gambled it, maybe.

[Megatron]

I don't see how there is really any debate at all.  It's clear we sacrificed the next two years of contention for the hope of building a contendor again in 2012.

Simple

  There's a debate because many people (if not most) would disagree with your statement.

Yeah, I don't like the trade because I think it was too risky, but I don't think we've sacrificed anything.  Gambled it, maybe.

If Shaq and JO can return healthy and STAY healthy throughout the playoffs, I think this trade made us BETTER.

If those two cant stay healthy, it made us worse. I have faith in Danny though, he knows better then I do about play health.

[ballin]

There is some validity to the "useage" argument; i.e., when a player needs to take more shots he will become less efficient. But this is only true if a player takes more shots of a particular type that he already wasn't efficient at to begin with. (check out David Berri's writing over at Wages of Wins for more info)

To illustrate, let's say a player like Marcin Gortat averages .555 shooting on 4 attempts per game. The argument goes that if Gortat were to increase his attempts to 8 per game, his shooting percentage would decrease. However, Berri's statistical research has shown that this is only true IF Gortat starts taking different types of shots in order to increase his attempts.

In other words, somebody that finishes around the rim 60% of time will always finish around the rim 60% of the time no matter how many shots at the rim a player takes. But players who increase their shot attempts usually do so by taking more shots further away from the rim or other shots they are not proficient at. ("heat checks").

So what does all this mean? It means that a low useage rate is a GOOD thing if you have other players who are better at taking the shots the lower useage player would take inefficiently. As it applies to our team, it means Glen Davis shouldn't take jumpers when you can have Paul Pierce iso and hit that shot much more efficiently.

The only player in the discussion that has an offensive move that is "worthwhile" enough to take away shots from our stars is Murphy, with his 3-point .391%. So if Murphy comes in and sticks to 3's and putbacks, then I'll agree that Murphy was an offensive upgrade. If he comes in and him, Krstic, and Davis start shooting deep 2's, they will hurt a lot more than they will help.

[jmei]

Let's try using a few more advanced stats-- offensive rating and offensive win shares. These statistics are based on field goals made, field goals missed, assists, offensive rebounds, and turnovers and are meant to serve as a catch-all offensive statistic. Details about how they work in more detail are at the bottom of my post.

Kendrick Perkins
Career offensive rating: 103
Offensive rating, last three years (2010-11 first): 97, 105, 105
Offensive Win Shares, last three years: -0.1, 1.4, 1.1

Glen Davis
Career offensive rating: 103
Offensive rating, last three years: 101, 103, 106
Offensive Win Shares, last three years: 0.4, 0.4, 1.0

Troy Murphy
Career offensive rating: 111
Offensive rating, last three years: 83, 115, 120
Offensive Win Shares, last three years: -0.4, 4.1, 5.2

Nenad Krstic
Career offensive rating: 105
Offensive rating, last three years: 112, 111, 104
Offensive Win Shares, last three years: 1.4, 2.0, 0.6


Based on this admittedly basic analysis, it seems like the order is as follows: Murphy > Krstic > Perkins > Davis. Which, for me at least, backs up the eye test.

------------
Glossary
http://www.basketball-reference.com/about/glossary.html
http://www.basketball-reference.com/about/ws.html
http://en.wikipedia.org/wiki/Offensive_rating

[jmei]

Sidenote: offensive rating is a rate stat (per-minute production) and offensive win shares is a counting stat which accounts for playing time. Also note small sample sizes for Perkins' and Murphy's production this season, which probably means you ignore them entirely or at least take them with a grain of salt.

[BballTim]

There is some validity to the "useage" argument; i.e., when a player needs to take more shots he will become less efficient. But this is only true if a player takes more shots of a particular type that he already wasn't efficient at to begin with. (check out David Berri's writing over at Wages of Wins for more info)

To illustrate, let's say a player like Marcin Gortat averages .555 shooting on 4 attempts per game. The argument goes that if Gortat were to increase his attempts to 8 per game, his shooting percentage would decrease. However, Berri's statistical research has shown that this is only true IF Gortat starts taking different types of shots in order to increase his attempts.

In other words, somebody that finishes around the rim 60% of time will always finish around the rim 60% of the time no matter how many shots at the rim a player takes. But players who increase their shot attempts usually do so by taking more shots further away from the rim or other shots they are not proficient at. ("heat checks").

  This isn't the case. Take Rondo, for instance. He hits over 60% of his layups. But he only attempts them when he thinks he can get to the rim without getting blocked. If he decided to take 20 layups in the next game instead of 5 then he'd have to take a lot of them with the opposing center waiting for him. He wouldn't hit 60% of them.

  If Perk could take as many close shots as he wanted without any drop in his FG% or TS% then our offense would consist almost entirely of Perk scoring from in close and PP and RA taking threes, and we'd be unstoppable.

[BballTim]

Sidenote: offensive rating is a rate stat (per-minute production) and offensive win shares is a counting stat which accounts for playing time. Also note small sample sizes for Perkins' and Murphy's production this season, which probably means you ignore them entirely or at least take them with a grain of salt.

  I don't think that offensive win shares accounts for playing time (although I could be wrong). I think you have Win Shares (which is offensive win shares + defensive win shares) and you also have Win Shares/48.

[ballin]

There is some validity to the "useage" argument; i.e., when a player needs to take more shots he will become less efficient. But this is only true if a player takes more shots of a particular type that he already wasn't efficient at to begin with. (check out David Berri's writing over at Wages of Wins for more info)

To illustrate, let's say a player like Marcin Gortat averages .555 shooting on 4 attempts per game. The argument goes that if Gortat were to increase his attempts to 8 per game, his shooting percentage would decrease. However, Berri's statistical research has shown that this is only true IF Gortat starts taking different types of shots in order to increase his attempts.

In other words, somebody that finishes around the rim 60% of time will always finish around the rim 60% of the time no matter how many shots at the rim a player takes. But players who increase their shot attempts usually do so by taking more shots further away from the rim or other shots they are not proficient at. ("heat checks").

  This isn't the case. Take Rondo, for instance. He hits over 60% of his layups. But he only attempts them when he thinks he can get to the rim without getting blocked. If he decided to take 20 layups in the next game instead of 5 then he'd have to take a lot of them with the opposing center waiting for him. He wouldn't hit 60% of them.

  If Perk could take as many close shots as he wanted without any drop in his FG% or TS% then our offense would consist almost entirely of Perk scoring from in close and PP and RA taking threes, and we'd be unstoppable.

Without realizing it, you changed the circumstances and invalidated your argument.

"tak[ing] a lot of them with the opposing center waiting for him" is not the same as him taking more of the same shot.

Obviously, teams adjust, but there are ways to get the same shot without doing the same thing over and over again. Different plays can end up with the player getting to the same spot and taking the same shot.

[BballTim]

There is some validity to the "useage" argument; i.e., when a player needs to take more shots he will become less efficient. But this is only true if a player takes more shots of a particular type that he already wasn't efficient at to begin with. (check out David Berri's writing over at Wages of Wins for more info)

To illustrate, let's say a player like Marcin Gortat averages .555 shooting on 4 attempts per game. The argument goes that if Gortat were to increase his attempts to 8 per game, his shooting percentage would decrease. However, Berri's statistical research has shown that this is only true IF Gortat starts taking different types of shots in order to increase his attempts.

In other words, somebody that finishes around the rim 60% of time will always finish around the rim 60% of the time no matter how many shots at the rim a player takes. But players who increase their shot attempts usually do so by taking more shots further away from the rim or other shots they are not proficient at. ("heat checks").

  This isn't the case. Take Rondo, for instance. He hits over 60% of his layups. But he only attempts them when he thinks he can get to the rim without getting blocked. If he decided to take 20 layups in the next game instead of 5 then he'd have to take a lot of them with the opposing center waiting for him. He wouldn't hit 60% of them.

  If Perk could take as many close shots as he wanted without any drop in his FG% or TS% then our offense would consist almost entirely of Perk scoring from in close and PP and RA taking threes, and we'd be unstoppable.

Without realizing it, you changed the circumstances and invalidated your argument.

"tak[ing] a lot of them with the opposing center waiting for him" is not the same as him taking more of the same shot.

  Of course I realized it. I was pointing out the fallacy of your argument, that somebody that finishes around the rim 60% of time will always finish around the rim 60% of the time no matter how many shots at the rim a player takes. Taking more of the same shot will change the circumstances of your shot. If Perk takes 6 shots a game they could average out to 4 uncontested shots and 2 putbacks (for example). If he were to increase his number of shots then he'd be taking shots from the same spot but he'd be going up against defenders.