Greg Stiemsma's Potential = Perk or Better?

[clover]

I can't figure out how Stiemsma's defensive rating is so out of line with his overall on court/off court numbers.

Both basketball value and 82games.com have him allowing 102 and 103 pp/100 on the court and the team at around 98 with him off the court.

I know Dean Olliver and other places rate possesions differently but that's a pretty big difference.

I bet ya his defensive numbers when he's playing with KG (like Perk) are much better.

Yeah but both DRTG and on/off court numbers should have the same amount of minutes next to KG, neither are adjusted for teammate quality from what I understand.

Yes, I was somewhat answering a different question.  But the difference between the two stats you quote is the 'per possession' aspect of one, no?

[Fafnir]

I can't figure out how Stiemsma's defensive rating is so out of line with his overall on court/off court numbers.

Both basketball value and 82games.com have him allowing 102 and 103 pp/100 on the court and the team at around 98 with him off the court.

I know Dean Olliver and other places rate possesions differently but that's a pretty big difference.

I bet ya his defensive numbers when he's playing with KG (like Perk) are much better.

Yeah but both DRTG and on/off court numbers should have the same amount of minutes next to KG, neither are adjusted for teammate quality from what I understand.

Yes, I was somewhat answering a different question.  But the difference between the two stats you quote is the 'per possession' aspect of one, no?

Has to be.

[clover]

And again, im saying that the metrics are flawed yes, im not saying they are useless and should be thrown out completely, but there are certain examples where they just don't make sense and therefore need to be considered with caution, ellsbury being one of them and Stiemsma being better than Perk is another.  

The examples that don't make sense are the biggest reasons to use statistics.

We don't need advanced measures to know that LeBron is awesome, or that Dwight Howard is a very good defender. Its the cases that conflict with what we "know" are the most interesting and most enlightening.

Sure, but if you are looking at stats and it tells you that Dwight is the second worst defensive center in the league will you look at it and say, "Man how big of a fool am I, I thought he was good!" or "something must be wrong with the equation producing this stat or some of the input data must be flawed."

You know what?  I'm just going to go ahead and say that the hypothesis that Stiemsma is better than Perkins on defense right now does not fail the eyeball test.  I don't think that it is blindingly obvious who is better just from casually watching several games, so I am open to the numbers persuading me in either direction.  So, if there are metrics that say Stiemsma is the better defender, I don't think that is evidence that those metrics are flawed.

Take a look here: http://www.basketball-reference.com/teams/BOS/2012.html

Stiemer has the best defensive rating on the team and he is tied for third on the team in win shares per 48.

[Rondo2287]

And again, im saying that the metrics are flawed yes, im not saying they are useless and should be thrown out completely, but there are certain examples where they just don't make sense and therefore need to be considered with caution, ellsbury being one of them and Stiemsma being better than Perk is another.  

The examples that don't make sense are the biggest reasons to use statistics.

We don't need advanced measures to know that LeBron is awesome, or that Dwight Howard is a very good defender. Its the cases that conflict with what we "know" are the most interesting and most enlightening.

Sure, but if you are looking at stats and it tells you that Dwight is the second worst defensive center in the league will you look at it and say, "Man how big of a fool am I, I thought he was good!" or "something must be wrong with the equation producing this stat or some of the input data must be flawed."

You know what?  I'm just going to go ahead and say that the hypothesis that Stiemsma is better than Perkins on defense right now does not fail the eyeball test.  I don't think that it is blindingly obvious who is better just from casually watching several games, so I am open to the numbers persuading me in either direction.  So, if there are metrics that say Stiemsma is the better defender, I don't think that is evidence that those metrics are flawed.

Take a look here: http://www.basketball-reference.com/teams/BOS/2012.html

Stiemer has the best defensive rating on the team and he is tied for third on the team in win shares per 48.

Using the same metric Pierce and Sasha are equally good defensively.  Interesting

[Vermont Green]

I think if you gave most GMs a stratigh up choice for this season between Perk or Stiems, money not a factor, most would take Perk.  Factor in money and upside, and I would think it would be much more even with Stiems having the potential to grow to be a better overall player.

Perk is a fine player but the Celtics played some of their best basketball (early season last) with him on the bench injured and either Shaq or Erden playing.  They are now playing some pretty good ball again with basically Bass in for Perk.  I think we can be pretty good with Steims playing 16-20 also.  It has a lot more to do with PP and KG.  Right now they are healthy and it changes the whole team.

[Rondo2287]

Just doing a little Perk research, I didnt remember that he averaged a double double in 08-09 in the playoffs.  Pretty Beastly.

11.9 ppg, 11.6 rpg, 2.6 blks

[clover]

And again, im saying that the metrics are flawed yes, im not saying they are useless and should be thrown out completely, but there are certain examples where they just don't make sense and therefore need to be considered with caution, ellsbury being one of them and Stiemsma being better than Perk is another.  

The examples that don't make sense are the biggest reasons to use statistics.

We don't need advanced measures to know that LeBron is awesome, or that Dwight Howard is a very good defender. Its the cases that conflict with what we "know" are the most interesting and most enlightening.

Sure, but if you are looking at stats and it tells you that Dwight is the second worst defensive center in the league will you look at it and say, "Man how big of a fool am I, I thought he was good!" or "something must be wrong with the equation producing this stat or some of the input data must be flawed."

You know what?  I'm just going to go ahead and say that the hypothesis that Stiemsma is better than Perkins on defense right now does not fail the eyeball test.  I don't think that it is blindingly obvious who is better just from casually watching several games, so I am open to the numbers persuading me in either direction.  So, if there are metrics that say Stiemsma is the better defender, I don't think that is evidence that those metrics are flawed.

Take a look here: http://www.basketball-reference.com/teams/BOS/2012.html

Stiemer has the best defensive rating on the team and he is tied for third on the team in win shares per 48.

Using the same metric Pierce and Sasha are equally good defensively.  Interesting

Yes, equal to each other--not to Stiemer! 

[LooseCannon]

I think if you gave most GMs a stratigh up choice for this season between Perk or Stiems, money not a factor, most would take Perk.

GMs tend to have a bias towards veterans over rookies (especially rookies who aren't first round picks), so that's not a great argument in favor of your point.  Those same GMs would take Keyon Dooling over E'Twuan Moore, a choice that would have been strongly disputed on this board a few weeks ago.

[Rondo2287]

I think if you gave most GMs a stratigh up choice for this season between Perk or Stiems, money not a factor, most would take Perk.

GMs tend to have a bias towards veterans over rookies (especially rookies who aren't first round picks), so that's not a great argument in favor of your point.  Those same GMs would take Keyon Dooling over E'Twuan Moore, a choice that would have been strongly disputed on this board a few weeks ago.

Also especially not towards rookies who are only one year younger than the veteran in question.  If you take injuries into account I would buy that though.

[BballTim]

And again, im saying that the metrics are flawed yes, im not saying they are useless and should be thrown out completely, but there are certain examples where they just don't make sense and therefore need to be considered with caution, ellsbury being one of them and Stiemsma being better than Perk is another.  

The examples that don't make sense are the biggest reasons to use statistics.

We don't need advanced measures to know that LeBron is awesome, or that Dwight Howard is a very good defender. Its the cases that conflict with what we "know" are the most interesting and most enlightening.

Sure, but if you are looking at stats and it tells you that Dwight is the second worst defensive center in the league will you look at it and say, "Man how big of a fool am I, I thought he was good!" or "something must be wrong with the equation producing this stat or some of the input data must be flawed."

You know what?  I'm just going to go ahead and say that the hypothesis that Stiemsma is better than Perkins on defense right now does not fail the eyeball test.  I don't think that it is blindingly obvious who is better just from casually watching several games, so I am open to the numbers persuading me in either direction.  So, if there are metrics that say Stiemsma is the better defender, I don't think that is evidence that those metrics are flawed.

Take a look here: http://www.basketball-reference.com/teams/BOS/2012.html

Stiemer has the best defensive rating on the team and he is tied for third on the team in win shares per 48.

  You also have to consider that, beyond the fact that many of Steamer's minutes come against backups, he's 11th on the team in minutes per game. When he's in a bad matchup he won't play a lot, when he's playing well he'll get more minutes.

[clover]

And again, im saying that the metrics are flawed yes, im not saying they are useless and should be thrown out completely, but there are certain examples where they just don't make sense and therefore need to be considered with caution, ellsbury being one of them and Stiemsma being better than Perk is another.  

The examples that don't make sense are the biggest reasons to use statistics.

We don't need advanced measures to know that LeBron is awesome, or that Dwight Howard is a very good defender. Its the cases that conflict with what we "know" are the most interesting and most enlightening.

Sure, but if you are looking at stats and it tells you that Dwight is the second worst defensive center in the league will you look at it and say, "Man how big of a fool am I, I thought he was good!" or "something must be wrong with the equation producing this stat or some of the input data must be flawed."

You know what?  I'm just going to go ahead and say that the hypothesis that Stiemsma is better than Perkins on defense right now does not fail the eyeball test.  I don't think that it is blindingly obvious who is better just from casually watching several games, so I am open to the numbers persuading me in either direction.  So, if there are metrics that say Stiemsma is the better defender, I don't think that is evidence that those metrics are flawed.

Take a look here: http://www.basketball-reference.com/teams/BOS/2012.html

Stiemer has the best defensive rating on the team and he is tied for third on the team in win shares per 48.

  You also have to consider that, beyond the fact that many of Steamer's minutes come against backups, he's 11th on the team in minutes per game. When he's in a bad matchup he won't play a lot, when he's playing well he'll get more minutes.

Like other players, Stiemer tends to play better with the other starters though as a backup that's not where he gets most of his minutes.  Perk, for example, almost always played with the benefit of KG beside him.

[BballTim]

And again, im saying that the metrics are flawed yes, im not saying they are useless and should be thrown out completely, but there are certain examples where they just don't make sense and therefore need to be considered with caution, ellsbury being one of them and Stiemsma being better than Perk is another.  

The examples that don't make sense are the biggest reasons to use statistics.

We don't need advanced measures to know that LeBron is awesome, or that Dwight Howard is a very good defender. Its the cases that conflict with what we "know" are the most interesting and most enlightening.

Sure, but if you are looking at stats and it tells you that Dwight is the second worst defensive center in the league will you look at it and say, "Man how big of a fool am I, I thought he was good!" or "something must be wrong with the equation producing this stat or some of the input data must be flawed."

You know what?  I'm just going to go ahead and say that the hypothesis that Stiemsma is better than Perkins on defense right now does not fail the eyeball test.  I don't think that it is blindingly obvious who is better just from casually watching several games, so I am open to the numbers persuading me in either direction.  So, if there are metrics that say Stiemsma is the better defender, I don't think that is evidence that those metrics are flawed.

Take a look here: http://www.basketball-reference.com/teams/BOS/2012.html

Stiemer has the best defensive rating on the team and he is tied for third on the team in win shares per 48.

  You also have to consider that, beyond the fact that many of Steamer's minutes come against backups, he's 11th on the team in minutes per game. When he's in a bad matchup he won't play a lot, when he's playing well he'll get more minutes.

Like other players, Stiemer tends to play better with the other starters though as a backup that's not where he gets most of his minutes.  Perk, for example, almost always played with the benefit of KG beside him.

  Perkins was a very good defender before we traded for KG.

[MBunge]

Perkins is better.


Perkins is a starting caliber C in the NBA.

This season, Perk is averaging 4.8 pts, 6.3 boards, 1.2 blocks and just under 50% from the field in over 26 minutes a game.  That's a "starting caliber C" in the sense that Erick Dampier has been a starting caliber center over the last 6 years.

Perk's a much greater physical force and seems to be a better rebounder.  Stiems is a vastly superior shot blocker and seems much more fluid on the offensive end.  I'd say Perk's the better starter, but Stiems' talents actually make him more effective off the bench than Perk would be.

Mike

[clover]

Perkins is better.


Perkins is a starting caliber C in the NBA.

This season, Perk is averaging 4.8 pts, 6.3 boards, 1.2 blocks and just under 50% from the field in over 26 minutes a game.  That's a "starting caliber C" in the sense that Erick Dampier has been a starting caliber center over the last 6 years.

Perk's a much greater physical force and seems to be a better rebounder.  Stiems is a vastly superior shot blocker and seems much more fluid on the offensive end.  I'd say Perk's the better starter, but Stiems' talents actually make him more effective off the bench than Perk would be.

Mike

At this point Stiemer is also a .667% career FT shooter to Perk's .601%.

[Neurotic Guy]

Things Stiem already does better:
1. Jumper
2. Quick up with the ball at the hoop
3. Sets picks without fouling (Perk sets a more imposing pick, but used to foul frequently)
4. Blocks shots

The most interesting difference I've observed (an observation that I'll continue to watch for), is that Perk used to keep people out of the paint with his presence and intensity; Stiem is less fearsome and guys get to the rim -- but notice, his pesence alters a ton of shots inside.  Note how many bunnies are missed when a guy catches Stiemer's arms rising out of the corner of his eye.

I am very excited about the opportunity Greg has for playing high intensity minutes over the next stretch.  The games against Miami, San Antonio, Chicago, Indiana, etc will all have playoff intensity. Stiem's experience grows each game.  I think it is a great plus for him that the schedule played out like this -- really will help him be more playoff ready.