Sullinger is the worst defender on the team

[TheFlex]

Truth be told, they both are bad on defense, but KO is slightly worse.  Those stats on Sully by Boris ( TP +1) show that he is bad defender, a really bad defender but KO is even worse.  This should not be a surprise given their combine scores in athletic ability.   Both did poorly in physical tests.
Now, the positive thing is both are nifty offensive players and Sully is a good but not elite rebounder.   Both of them are good passers.

When we play both of them at the same time, we are hosed.   They both stink to high heck on D.   We need a rim protector bad to cover for them and sorry, Zeller is a backup.

I agree with your points.

Lol. If you agree, is there any point to loyally challenging anyone who asserts that KO was not the better defender in an individual game (in this case, vs. DET)?

[Celtics4ever]

KO was not the better defender in an individual game (in this case, vs. DET)?

I was speaking of overall tendacies but I would like to point out that neither played good D that game as well.

Their bigs scored 29 and 27 did any one play well on D?   I think not.

[The Oracle]

  The statistics used in this thread by many different posters are at best next to useless!
 
1.  You can't cherry pick statistics from small sample sizes and use them to prove anything!        Differences between EFG%, Orat., Drat. and the like are all way to volatile in 16 game sample sizes!

2.  Using PER for anything is a horrible idea!

3.  Using Def. PER should be a criminal offense!  It is not even remotely correct in many instances and it certainly is not in the case of Sully versus K.O.!

 
 

[Granath]

The only appropriate reaction to the OP.

[TheFlex]

  The statistics used in this thread by many different posters are at best next to useless!
 
1.  You can't cherry pick statistics from small sample sizes and use them to prove anything!        Differences between EFG%, Orat., Drat. and the like are all way to volatile in 16 game sample sizes!

2.  Using PER for anything is a horrible idea!

3.  Using Def. PER should be a criminal offense!  It is not even remotely correct in many instances and it certainly is not in the case of Sully versus K.O.!

Sully's defensive stats from 2013-14 are also better than Olynyk's. You can look them up for yourself if you'd like.

Your second point is completely wrong. Using PER as an overall evaluation of a player is flawed because it doesn't allot defensive performance appropriate weight. However, when assessing a player's impact on defense, their opponent's PER is quite effective as it is essentially an offensive advanced stat. A player's defensive assignment's PER doesn't speak as to how well that assignment plays defense on the other end of the floor but it is one stat that exhibits the assignment's offensive performance against that player

I didn't use defensive PER, I don't know if anyone else did.

[The Oracle]

  If PER was an accurate measure of a players actual performance and If said players opponent was accurately recorded then yes it would be a be a good statistic!  The truth is that whatever site you are using is most certainly not recording who a player is guarding!  They are making assumptions and marking who they assume was the center and power forward and doing the same for the other team!  They then assume one center is guarding the other teams center and so forth!  In the case of K.O. and Sully this results in Sully being marked as the PF and K.O. the center when in truth K.O. is more often then not this year actually defending the PF position when on the floor with Sully!  The same is true for Rondo and Avery, Rondo is being marked as defending the other teams PG when in fact he is not more often then he is when on the floor with Avery!  This results in a very muddy mess and is useless!

[mmmmm]

  If PER was an accurate measure of a players actual performance and If said players opponent was accurately recorded then yes it would be a be a good statistic!  The truth is that whatever site you are using is most certainly not recording who a player is guarding!  They are making assumptions and marking who they assume was the center and power forward and doing the same for the other team!  They then assume one center is guarding the other teams center and so forth!  In the case of K.O. and Sully this results in Sully being marked as the PF and K.O. the center when in truth K.O. is more often then not this year actually defending the PF position when on the floor with Sully!  The same is true for Rondo and Avery, Rondo is being marked as defending the other teams PG when in fact he is not more often then he is when on the floor with Avery!  This results in a very muddy mess and is useless!

Look, I personally am a huge critic of PER - it is a lousy stat for comparing different players for a host of reasons that I have ranted about frequently.

But the way it is used by 82games.com to record the opposition performance and the positional 'Net PER' is actually pretty reasonable because it isolates the role context.

The position assignments are certainly subject to error, but in the aggregate are probably pretty close.    Especially on the defensive end.

Absolute precision isn't really necessary when the differences illustrated are this big.   Further, the Net PER results used here are completely consistent with what has been measured on the scoreboard, as data posted in this thread illustrate. 

It is certainly true that 15 games is ultimately indeed a small sample and I personally wouldn't read too much into the numbers from just that.

But the numbers are a record of what has happened so far and even if you don't want to accept them as proof that Sully is a better defender than Olynyk, you still have to reconcile that they are completely at odds with the OP assertion of the contrary.

And the numbers from last year's much larger sample also don't support the OP.

[The Oracle]

  I don't have time right now to properly respond to this as the game is going to start but I will come on later and respond to this!  The data that places like 82 games is spitting out is garbage or at least some of it is and I will explain why tonight if not tomorrow! 

[Tr1boy]

  If PER was an accurate measure of a players actual performance and If said players opponent was accurately recorded then yes it would be a be a good statistic!  The truth is that whatever site you are using is most certainly not recording who a player is guarding!  They are making assumptions and marking who they assume was the center and power forward and doing the same for the other team!  They then assume one center is guarding the other teams center and so forth!  In the case of K.O. and Sully this results in Sully being marked as the PF and K.O. the center when in truth K.O. is more often then not this year actually defending the PF position when on the floor with Sully!  The same is true for Rondo and Avery, Rondo is being marked as defending the other teams PG when in fact he is not more often then he is when on the floor with Avery!  This results in a very muddy mess and is useless!

+1

[Celtics18]

  If PER was an accurate measure of a players actual performance and If said players opponent was accurately recorded then yes it would be a be a good statistic!  The truth is that whatever site you are using is most certainly not recording who a player is guarding!  They are making assumptions and marking who they assume was the center and power forward and doing the same for the other team!  They then assume one center is guarding the other teams center and so forth!  In the case of K.O. and Sully this results in Sully being marked as the PF and K.O. the center when in truth K.O. is more often then not this year actually defending the PF position when on the floor with Sully!  The same is true for Rondo and Avery, Rondo is being marked as defending the other teams PG when in fact he is not more often then he is when on the floor with Avery!  This results in a very muddy mess and is useless!

Look, I personally am a huge critic of PER - it is a lousy stat for comparing different players for a host of reasons that I have ranted about frequently.

But the way it is used by 82games.com to record the opposition performance and the positional 'Net PER' is actually pretty reasonable because it isolates the role context.

The position assignments are certainly subject to error, but in the aggregate are probably pretty close.    Especially on the defensive end.

Absolute precision isn't really necessary when the differences illustrated are this big.   Further, the Net PER results used here are completely consistent with what has been measured on the scoreboard, as data posted in this thread illustrate. 

It is certainly true that 15 games is ultimately indeed a small sample and I personally wouldn't read too much into the numbers from just that.

But the numbers are a record of what has happened so far and even if you don't want to accept them as proof that Sully is a better defender than Olynyk, you still have to reconcile that they are completely at odds with the OP assertion of the contrary.

And the numbers from last year's much larger sample also don't support the OP.

I've got to say that I agree with Oracle on this one.  Maybe, "in aggregate," meaning taking the whole league, the position assignments might end up being pretty close.  However, in specific instances, like Rondo and Bradley, and Sullinger and Olynyk, I don't think they are close at all.  If 82games simply uses the starting positions of each of those players as listed in the box score to determine who they are guarding, then they are getting it wrong--very wrong. 

Those of us who watch the Celtics play know that the majority of the time (when Olynyk was still the starter), Kelly was guarding fours and Sully was guarding fives.  We know the same is true of Rondo and Bradley at the 1 and 2.  In a number of instances, Rondo has even spent more time guarding opposing threes.

I'm sure we are not the only team that regularly employs this type of "cross matching."  For this reason, I completely agree that it is hard to find those sorts of head to head defensive stats reliable.

I prefer the more straightforward defensive rating system.  Of course, that comes with a variety of well-documented flaws, as well.