Should Joe keep playing Kornet?

[SCeltic34]

First off, this is not a thread meant to bash Kornet, and please do not turn it into one.  Kornet has given us some solid minutes in some of the games this season.  As a whole, I'm proud of our bench players.  Each player so far has had at least one game in which they've had a positive impact and helped us win the game.

However, I would assert that Kornet should not get regular playing time moving forward, even if we are short on bigs.  Our defense suffers too much with him in the game, and his offensive contributions do not outweigh his defensive liabilities.  I especially dislike double big lineups in which Kornet is one of the bigs on the floor.  It compromises our defense far too much and makes us particularly vulnerable to 3-point shooting.

The stats support the eye test.





Now, I'm not sure which set of stats is more accurate.  The top screen capture is from CleaningTheGlass.  The second (bottom) screen capture is from basketball reference.  (Click on the screenshot and it should enlarge.)

Both set of statistics show just how much our defense suffers with Kornet on the floor.  In the top screen capture, I put a green box over the on/off stats for this season.  Unless I am misreading the stats, it shows the following:

- As a whole, we are -4.8 points per 100 possessions with Kornet on the court
- We apparently score more points +1.8 per 100 possessions with Kornet on the floor, but shoot -3.9% worse on eFG%
- We give up +6.6 points per 100 possessions with Kornet on the floor compared to when he is off the floor
- Our defensive rebounding suffers with him on the floor, +7.4% for the opponent to grab an offensive rebound with Kornet on the floor as opposed to off.  He does help our offensive rebounding significantly.

The second set of stats shows the same story.  With Kornet on the court, opposing teams have:

- A higher eFG% (.545% on vs. .516% off)
- A higher offensive rebounding % by a significant margin (+8.7%)
- Very significant higher offensive rating of 117.4 with Kornet on, as opposed to 109 with Kornet off the floor
- In this set of stats, our eFG% suffers significantly with Kornet on the court, but the offensive rating is essentially the same

All of our bigs play drop coverage in PnR.  But from what I've observed, Kornet is especially prone to getting caught in no man's land where he cannot do anything effective, including grabbing a defense rebound.

The C's, as all teams in the NBA now, pay attention to advanced stats.  Unless there's something I'm missing here, I don't quite understand why Joe plays him as often as he does.  There have been a few games where we won despite Joe inexplicably putting Kornet back in the game in the 3rd/4th quarter so it might be overlooked by some fans. 

I'm still open to using Kornet is certain situations.  However, given the eye test and the stats above, I really hope that Joe cuts back on his playing time.  I'd much prefer playing a more mobile player such as Stevens or Brissett, even with their offensive shortcomings.  They give us much more toughness, defensive versatility and speed.  They should be able to contribute if they are given more consistent playing time.  If time reveals that they are not a net positive, then another adjustment can be made at a later time.

[jambr380]

Thanks for providing actual statistics - I agree. I get that he is trusted because he knows the system, but besides a highlight block or alley-oop every now and then, he seemingly plays way below his 7'2" listed height.

I realize he is a good vibes guy and it isn't his fault that he vaulted into the top 9 of our rotation, but he should move back in the 12-13 range where he has always been before. With injuries or Al sitting, Luke moves up to 8th, 7th, or even 6th -  that's just crazy. I know Queta lacks experience, but with Rob no longer on the team, his athleticism and grit around the rim is much appreciated.

[liam]

It looks like Luke has a PER 19.2 this year. 3rd on the team behind Tatum and Porzingis. Interesting.

[keevsnick]

Hè's the third string center, he's fine for what he is.

[jpotter33]

Agreed. Queta has shown enough - both now and in terms of potential - that he should have the third big spot. And then both Stevens and Brissett are there for the extra big spot when we need to go smaller.

Hopefully this time with him being out will allow for one of the others to shine enough to take his place.

[ozgod]

First off, this is not a thread meant to bash Kornet, and please do not turn it into one.  Kornet has given us some solid minutes in some of the games this season.  As a whole, I'm proud of our bench players.  Each player so far has had at least one game in which they've had a positive impact and helped us win the game.

However, I would assert that Kornet should not get regular playing time moving forward, even if we are short on bigs.  Our defense suffers too much with him in the game, and his offensive contributions do not outweigh his defensive liabilities.  I especially dislike double big lineups in which Kornet is one of the bigs on the floor.  It compromises our defense far too much and makes us particularly vulnerable to 3-point shooting.

The stats support the eye test.

~snip~

Now, I'm not sure which set of stats is more accurate.  The top screen capture is from CleaningTheGlass.  The second (bottom) screen capture is from basketball reference.  (Click on the screenshot and it should enlarge.)

Both set of statistics show just how much our defense suffers with Kornet on the floor.  In the top screen capture, I put a green box over the on/off stats for this season.  Unless I am misreading the stats, it shows the following:

- As a whole, we are -4.8 points per 100 possessions with Kornet on the court
- We apparently score more points +1.8 per 100 possessions with Kornet on the floor, but shoot -3.9% worse on eFG%
- We give up +6.6 points per 100 possessions with Kornet on the floor compared to when he is off the floor
- Our defensive rebounding suffers with him on the floor, +7.4% for the opponent to grab an offensive rebound with Kornet on the floor as opposed to off.  He does help our offensive rebounding significantly.

The second set of stats shows the same story.  With Kornet on the court, opposing teams have:

- A higher eFG% (.545% on vs. .516% off)
- A higher offensive rebounding % by a significant margin (+8.7%)
- Very significant higher offensive rating of 117.4 with Kornet on, as opposed to 109 with Kornet off the floor
- In this set of stats, our eFG% suffers significantly with Kornet on the court, but the offensive rating is essentially the same

All of our bigs play drop coverage in PnR.  But from what I've observed, Kornet is especially prone to getting caught in no man's land where he cannot do anything effective, including grabbing a defense rebound.

The C's, as all teams in the NBA now, pay attention to advanced stats.  Unless there's something I'm missing here, I don't quite understand why Joe plays him as often as he does.  There have been a few games where we won despite Joe inexplicably putting Kornet back in the game in the 3rd/4th quarter so it might be overlooked by some fans. 

I'm still open to using Kornet is certain situations.  However, given the eye test and the stats above, I really hope that Joe cuts back on his playing time.  I'd much prefer playing a more mobile player such as Stevens or Brissett, even with their offensive shortcomings.  They give us much more toughness, defensive versatility and speed.  They should be able to contribute if they are given more consistent playing time.  If time reveals that they are not a net positive, then another adjustment can be made at a later time.

I couldn't see the screenshot you posted (probably something to do with my settings) so I just reposted the On-Off for the Celtics overall below. You're right in how you read it, Kornet has the second worst net differential on the team at -5.4%, which means he's in the bottom 28% of all players in the NBA in that metric. The team had an eFG% of -4.3% when he was on the court which places him in the bottom 10% of players in terms of his impact on their eFG. The only thing he really has a significant positive impact on while on the court is in offensive rebounding, where the team grabs 11% more offensive rebounds when he is on the court, where he is in the top 100th percentile of players. If the numbers are slightly different it's probably a timing issue.

click to enlarge


If I had to answer the question on why he keeps getting the nod, if we assume that Joe is looking at advanced metrics (I'm sure he has access to much more from Second Spectrum, including actual ball tracking data), it's because as mediocre as Luke's advanced metrics are, the alternatives on our team are currently worse. For example Brissett has a -10.4 net rating when he is on the court, and Stevens (including his performance yesterday) has a -24.2. The obvious caveat is that Luke has played 192 min, Oshae 76 and Lamar 29, so their sample size is much lower, and many of those minutes were probably in blowouts. Queta in 47 mins has a 3.2 net rating differential when he is on the court. All three alternatives fall below the minimum playing time to be considered in the actual rankings so we don't know where they fall but it's fair to say Queta might be a bit above Luke and the other two below. So if we're debating the No9 spot, I think that's one of the spots that Joe could possibly experiment a little with his rotations (even though he said after the last game he wasn't really going to anymore).

I didn't know if you or anyone else noticed the elephant in the room though - of all our rotation players, the player with the worst on-off net rating differential is Jaylen Brown. The team's net rating is 7.7 points worse when he is on the court than off, which puts him in the bottom 21% of players in the league. The team is 5.6 points worse in points per possession which also puts him in the bottom 21% of players in terms of his impact on that metric. The team's offensive rebound % drops by 10.6% when Jaylen is on the floor, which puts his impact on that metric in the bottom 1% of players in the league. The only two Efficiency and Four Factors metrics that Jaylen has an impact above the median on is on free throw rate, because the team's free throw rate increases by 4.4 per 100 FGA when he is on the court, which puts him in the 85th percentile, and in causing opponent turnovers, which increases by 1.5% when he is on the floor, putting him in the 78th percentile for all players. So he has a lot more detrimental impact on our team than Luke does, and he plays a hell of a lot more minutes too.

The ironic thing is I'm actually not a Jaylen hater - my opinion is that he adds value to the team, even if I feel his contract will eventually cause him to be traded to keep us below the cap. I didn't know if I should bring it up because this is a Kornet thread. But stats are stats and it's very obvious when you look at the team's on-off metrics

[BitterJim]

Considering that Luke being on the court generally means Al and KP are off the court, I'm not sure why anyone would expect his On/Off stats to be anything but negative. KP and Al are so much better than him (and any other 3rd string center we could bring in) that looking at those stats is always going to make him look bad.

The real question isn't if Kornet is a net positive compared to KP and Al, it's if he's more of a net positive (or less of a net negative) than the other 3rd string Center/big options we have. And right now, the answer is a clear yes (except maybe Queta). I'd love for us to bring in a better 3rd string big, but until then Kornet should continue playing when Al and KP need rest, even though he isn't as good as them.

Also, for Offensive and Defensive rating basketball-reference uses a different formula than everyone else (IIRC it's to allow calculating those for players that played before possessions were tracked) so they are less accurate.

[Neurotic Guy]

Considering that Luke being on the court generally means Al and KP are off the court, I'm not sure why anyone would expect his On/Off stats to be anything but negative. KP and Al are so much better than him (and any other 3rd string center we could bring in) that looking at those stats is always going to make him look bad.

The real question isn't if Kornet is a net positive compared to KP and Al, it's if he's more of a net positive (or less of a net negative) than the other 3rd string Center/big options we have. And right now, the answer is a clear yes (except maybe Queta). I'd love for us to bring in a better 3rd string big, but until then Kornet should continue playing when Al and KP need rest, even though he isn't as good as them.

Also, for Offensive and Defensive rating basketball-reference uses a different formula than everyone else (IIRC it's to allow calculating those for players that played before possessions were tracked) so they are less accurate.

TP. You wrote exactly what I was thinking. Got to compare with other 3rd string centers. Not just statistically, but take a look at who they are.  Agree Brad should try to upgrade, but I definitely don’t see Luke the way many here do.

Could look at JBs stats through the lens that he’s on the court virtually all of the 12 minutes or so that our MVP candidate is off the court. Stats may not support what I just said, but JB also has been figuring out his role and my guess is that his net rating is trending positively.

[Surferdad]

Thanks for providing actual statistics - I agree. I get that he is trusted because he knows the system, but besides a highlight block or alley-oop every now and then, he seemingly plays way below his 7'2" listed height.

I realize he is a good vibes guy and it isn't his fault that he vaulted into the top 9 of our rotation, but he should move back in the 12-13 range where he has always been before. With injuries or Al sitting, Luke moves up to 8th, 7th, or even 6th -  that's just crazy. I know Queta lacks experience, but with Rob no longer on the team, his athleticism and grit around the rim is much appreciated.

TP, I agree with all of this. He certainly seems to play way below his 7'2" height. We've seen slashers go right at him, give a little up-fake and lay it in. He is not a rim protector.

[timpiker]

He's fine for what he is.

[liam]

Considering that Luke being on the court generally means Al and KP are off the court, I'm not sure why anyone would expect his On/Off stats to be anything but negative. KP and Al are so much better than him (and any other 3rd string center we could bring in) that looking at those stats is always going to make him look bad.

The real question isn't if Kornet is a net positive compared to KP and Al, it's if he's more of a net positive (or less of a net negative) than the other 3rd string Center/big options we have. And right now, the answer is a clear yes (except maybe Queta). I'd love for us to bring in a better 3rd string big, but until then Kornet should continue playing when Al and KP need rest, even though he isn't as good as them.

Also, for Offensive and Defensive rating basketball-reference uses a different formula than everyone else (IIRC it's to allow calculating those for players that played before possessions were tracked) so they are less accurate.

Bingo. It's a huge step down from Porzingis to Kornet.

[Phantom255x]

He's a 4th center at best, which is the role he was in last season. G-Will and even Griffin sort of functioned as our 3rd PF/C last season. Obviously we no longer have both + traded Timelord last minute, so Kornet is thrust into this role.

[SCeltic34]

First off, this is not a thread meant to bash Kornet, and please do not turn it into one.  Kornet has given us some solid minutes in some of the games this season.  As a whole, I'm proud of our bench players.  Each player so far has had at least one game in which they've had a positive impact and helped us win the game.

However, I would assert that Kornet should not get regular playing time moving forward, even if we are short on bigs.  Our defense suffers too much with him in the game, and his offensive contributions do not outweigh his defensive liabilities.  I especially dislike double big lineups in which Kornet is one of the bigs on the floor.  It compromises our defense far too much and makes us particularly vulnerable to 3-point shooting.

The stats support the eye test.

~snip~

Now, I'm not sure which set of stats is more accurate.  The top screen capture is from CleaningTheGlass.  The second (bottom) screen capture is from basketball reference.  (Click on the screenshot and it should enlarge.)

Both set of statistics show just how much our defense suffers with Kornet on the floor.  In the top screen capture, I put a green box over the on/off stats for this season.  Unless I am misreading the stats, it shows the following:

- As a whole, we are -4.8 points per 100 possessions with Kornet on the court
- We apparently score more points +1.8 per 100 possessions with Kornet on the floor, but shoot -3.9% worse on eFG%
- We give up +6.6 points per 100 possessions with Kornet on the floor compared to when he is off the floor
- Our defensive rebounding suffers with him on the floor, +7.4% for the opponent to grab an offensive rebound with Kornet on the floor as opposed to off.  He does help our offensive rebounding significantly.

The second set of stats shows the same story.  With Kornet on the court, opposing teams have:

- A higher eFG% (.545% on vs. .516% off)
- A higher offensive rebounding % by a significant margin (+8.7%)
- Very significant higher offensive rating of 117.4 with Kornet on, as opposed to 109 with Kornet off the floor
- In this set of stats, our eFG% suffers significantly with Kornet on the court, but the offensive rating is essentially the same

All of our bigs play drop coverage in PnR.  But from what I've observed, Kornet is especially prone to getting caught in no man's land where he cannot do anything effective, including grabbing a defense rebound.

The C's, as all teams in the NBA now, pay attention to advanced stats.  Unless there's something I'm missing here, I don't quite understand why Joe plays him as often as he does.  There have been a few games where we won despite Joe inexplicably putting Kornet back in the game in the 3rd/4th quarter so it might be overlooked by some fans. 

I'm still open to using Kornet is certain situations.  However, given the eye test and the stats above, I really hope that Joe cuts back on his playing time.  I'd much prefer playing a more mobile player such as Stevens or Brissett, even with their offensive shortcomings.  They give us much more toughness, defensive versatility and speed.  They should be able to contribute if they are given more consistent playing time.  If time reveals that they are not a net positive, then another adjustment can be made at a later time.

I couldn't see the screenshot you posted (probably something to do with my settings) so I just reposted the On-Off for the Celtics overall below. You're right in how you read it, Kornet has the second worst net differential on the team at -5.4%, which means he's in the bottom 28% of all players in the NBA in that metric. The team had an eFG% of -4.3% when he was on the court which places him in the bottom 10% of players in terms of his impact on their eFG. The only thing he really has a significant positive impact on while on the court is in offensive rebounding, where the team grabs 11% more offensive rebounds when he is on the court, where he is in the top 100th percentile of players. If the numbers are slightly different it's probably a timing issue.

click to enlarge


If I had to answer the question on why he keeps getting the nod, if we assume that Joe is looking at advanced metrics (I'm sure he has access to much more from Second Spectrum, including actual ball tracking data), it's because as mediocre as Luke's advanced metrics are, the alternatives on our team are currently worse. For example Brissett has a -10.4 net rating when he is on the court, and Stevens (including his performance yesterday) has a -24.2. The obvious caveat is that Luke has played 192 min, Oshae 76 and Lamar 29, so their sample size is much lower, and many of those minutes were probably in blowouts. Queta in 47 mins has a 3.2 net rating differential when he is on the court. All three alternatives fall below the minimum playing time to be considered in the actual rankings so we don't know where they fall but it's fair to say Queta might be a bit above Luke and the other two below. So if we're debating the No9 spot, I think that's one of the spots that Joe could possibly experiment a little with his rotations (even though he said after the last game he wasn't really going to anymore).

I didn't know if you or anyone else noticed the elephant in the room though - of all our rotation players, the player with the worst on-off net rating differential is Jaylen Brown. The team's net rating is 7.7 points worse when he is on the court than off, which puts him in the bottom 21% of players in the league. The team is 5.6 points worse in points per possession which also puts him in the bottom 21% of players in terms of his impact on that metric. The team's offensive rebound % drops by 10.6% when Jaylen is on the floor, which puts his impact on that metric in the bottom 1% of players in the league. The only two Efficiency and Four Factors metrics that Jaylen has an impact above the median on is on free throw rate, because the team's free throw rate increases by 4.4 per 100 FGA when he is on the court, which puts him in the 85th percentile, and in causing opponent turnovers, which increases by 1.5% when he is on the floor, putting him in the 78th percentile for all players. So he has a lot more detrimental impact on our team than Luke does, and he plays a hell of a lot more minutes too.

The ironic thing is I'm actually not a Jaylen hater - my opinion is that he adds value to the team, even if I feel his contract will eventually cause him to be traded to keep us below the cap. I didn't know if I should bring it up because this is a Kornet thread. But stats are stats and it's very obvious when you look at the team's on-off metrics

Good contribution, TPs.  We need a larger sample size from some of the other role players.  Queta has done a good job in his minutes.

[michigan adam]

First off, this is not a thread meant to bash Kornet, and please do not turn it into one.  Kornet has given us some solid minutes in some of the games this season.  As a whole, I'm proud of our bench players.  Each player so far has had at least one game in which they've had a positive impact and helped us win the game.

However, I would assert that Kornet should not get regular playing time moving forward, even if we are short on bigs.  Our defense suffers too much with him in the game, and his offensive contributions do not outweigh his defensive liabilities.  I especially dislike double big lineups in which Kornet is one of the bigs on the floor.  It compromises our defense far too much and makes us particularly vulnerable to 3-point shooting.

The stats support the eye test.





Now, I'm not sure which set of stats is more accurate.  The top screen capture is from CleaningTheGlass.  The second (bottom) screen capture is from basketball reference.  (Click on the screenshot and it should enlarge.)

Both set of statistics show just how much our defense suffers with Kornet on the floor.  In the top screen capture, I put a green box over the on/off stats for this season.  Unless I am misreading the stats, it shows the following:

- As a whole, we are -4.8 points per 100 possessions with Kornet on the court
- We apparently score more points +1.8 per 100 possessions with Kornet on the floor, but shoot -3.9% worse on eFG%
- We give up +6.6 points per 100 possessions with Kornet on the floor compared to when he is off the floor
- Our defensive rebounding suffers with him on the floor, +7.4% for the opponent to grab an offensive rebound with Kornet on the floor as opposed to off.  He does help our offensive rebounding significantly.

The second set of stats shows the same story.  With Kornet on the court, opposing teams have:

- A higher eFG% (.545% on vs. .516% off)
- A higher offensive rebounding % by a significant margin (+8.7%)
- Very significant higher offensive rating of 117.4 with Kornet on, as opposed to 109 with Kornet off the floor
- In this set of stats, our eFG% suffers significantly with Kornet on the court, but the offensive rating is essentially the same

All of our bigs play drop coverage in PnR.  But from what I've observed, Kornet is especially prone to getting caught in no man's land where he cannot do anything effective, including grabbing a defense rebound.

The C's, as all teams in the NBA now, pay attention to advanced stats.  Unless there's something I'm missing here, I don't quite understand why Joe plays him as often as he does.  There have been a few games where we won despite Joe inexplicably putting Kornet back in the game in the 3rd/4th quarter so it might be overlooked by some fans. 

I'm still open to using Kornet is certain situations.  However, given the eye test and the stats above, I really hope that Joe cuts back on his playing time.  I'd much prefer playing a more mobile player such as Stevens or Brissett, even with their offensive shortcomings.  They give us much more toughness, defensive versatility and speed.  They should be able to contribute if they are given more consistent playing time.  If time reveals that they are not a net positive, then another adjustment can be made at a later time.

Of course the on off is better with him off.  If he's off that means al or kp is on.  That us what those stats are saying. Are they horrible numbers for a third bug....no.  would I prefer a better backup big...yes. 

[Big333223]

I'm pretty high on Luke, for what he is. I think he was rounding into form as a lob threat before he got hurt. That said, I'd rather give more minutes to Queta and see what he has. It feels like his ceiling is higher and I want to know how close he can get to it with some real minutes.