Trade Irving for the sake of the team

[tarheelsxxiii]

I posted this in another thread and think it's meaningful.

For the 2017-2018 season, Boston's most used lineup was excellent:
Brown-Tatum-Horford-Baynes-KYRIE (354 minutes played). 

The Celtics had only two better lineups all season, actually; at least with sample sizes reasonable enough to interpret:

Brown-Tatum-Horford-Baynes-ROZIER (115 mins)

Brown-Tatum-Horford-Baynes-SMART (43 mins)

That means the Celtics' best lineups consisted of their starting 5 after swapping out Kyrie for Rozier or Smart.  Pretty powerful data.



Were the Celtics very good with Kyrie? Yes. He's a net negative defender, but very good scorer that offsets his defense.  But that the Celtics are better with two younger, borderline-starting PGs speaks volumes about his value as a teammate.

Add the fact that Kyrie (injury-prone, very reliant on quickness) will be due somewhere ~$40 million in his mid 30s, and there is plenty of cause for concern. 

https://hardwoodhoudini.com/2018/11/12/boston-celtics-holiday-gift-guide/

It doesn't really contribute much since it's 2017-18 stats. For instance, this year, when 5 man combos have played at least 20 minutes together, Kyrie is in 12 of the top 16 best net rated 5 man combos this year.

Here's the important stat: Kyrie is literally the only player on the team who when he sits the Celtics are outscored by the opponent per 100 possessions.

Kyrie also has the highest raw +/- and overall net rating.  The team is simply a considerable amount better with Kyrie playing than not playing, small sample sizes be [dang]ed.

https://stats.nba.com/team/1610612738/onoffcourt-summary/?sort=DIFF_NET_RATING&dir=1

That data is meaningful.  I agree that he is an overall plus, primarily because he's an elite scorer.  That said, the team was better over an entire season when you swapped him out for either of his two back-ups, and that is a much larger sample size.

You also have to consider context.  The 2017-2018 data was seen in the context of a deep playoff run made in Kyrie's absence; your 2018-2019 data is shown in the context of a slow start to the season in which nearly every player (minus Mook) has been worse since Kyrie returned.  That's also meaningful.

[celticsclay]

I often appreciate the points you make, Moranis, but I have no idea why you keep removing the 16-2 (or 15-2 with Irving) to begin last year. Just because it was 'unsustainable' doesn't mean it didn't happen. That start is even more impressive seeing as how we had just lost our max FA signing for at least a year.

It's hard to disagree that Rozier plays better when he starts, but guys like Tatum and Brown more seem to be affected by # of touches/shots than anything Kyrie is doing. Brown and Hayward finally commingling was just something that needed time.

If anything, it seems that Horford takes on more of the leadership role when Kyrie is out and all of the other guys fall into place. I do have confidence that as the season continues, players will feel more and more comfortable with their 'roles'.

Because it was a fluke.  Everyone knew that team wasn't anywhere near that good, which was borne out the rest of the season.  Sometimes teams go on incredible runs both good and bad that aren't indicative of the level of the team.  The other 64 games the team played gave a much better representation of the quality of that team.

There's nothing fluky about a 16-2 start.

And even if there is, it's basically picking and choosing what's a fluke.

We had stretches where we lost 5 out of 6 games, and 4 out of 5.  Were we a sub-.200 squad?  No?  Then aren't these flukes in comparison to our overall record?  Do we throw them out?

It's just fuzzy math, manipulated to "prove" a point.

This is how I feel a 100%. We have more data available than ever before, and it makes it easier to cherry pick random stuff to make an argument. However attempting to omit an entire 18 game stretch by merely calling it a fluke is about as egregious as it gets in that department

[The Oracle]

I posted this in another thread and think it's meaningful.

For the 2017-2018 season, Boston's most used lineup was excellent:
Brown-Tatum-Horford-Baynes-KYRIE (354 minutes played). 

The Celtics had only two better lineups all season, actually; at least with sample sizes reasonable enough to interpret:

Brown-Tatum-Horford-Baynes-ROZIER (115 mins)

Brown-Tatum-Horford-Baynes-SMART (43 mins)

That means the Celtics' best lineups consisted of their starting 5 after swapping out Kyrie for Rozier or Smart.  Pretty powerful data.



Were the Celtics very good with Kyrie? Yes. He's a net negative defender, but very good scorer that offsets his defense.  But that the Celtics are better with two younger, borderline-starting PGs speaks volumes about his value as a teammate.

Add the fact that Kyrie (injury-prone, very reliant on quickness) will be due somewhere ~$40 million in his mid 30s, and there is plenty of cause for concern. 

https://hardwoodhoudini.com/2018/11/12/boston-celtics-holiday-gift-guide/

It doesn't really contribute much since it's 2017-18 stats. For instance, this year, when 5 man combos have played at least 20 minutes together, Kyrie is in 12 of the top 16 best net rated 5 man combos this year.

Here's the important stat: Kyrie is literally the only player on the team who when he sits the Celtics are outscored by the opponent per 100 possessions.

Kyrie also has the highest raw +/- and overall net rating.  The team is simply a considerable amount better with Kyrie playing than not playing, small sample sizes be [dang]ed.

https://stats.nba.com/team/1610612738/onoffcourt-summary/?sort=DIFF_NET_RATING&dir=1

That data is meaningful.  I agree that he is an overall plus, primarily because he's an elite scorer.  That said, the team was better over an entire season when you swapped him out for either of his two back-ups, and that is a much larger sample size.

You also have to consider context.  The 2017-2018 data was seen in the context of a deep playoff run made in Kyrie's absence; your 2018-2019 data is shown in the context of a slow start to the season in which nearly every player (minus Mook) has been worse since Kyrie returned.  That's also meaningful.

115 and 43 minute sample sizes have nearly zero statistical value, they are merely a novelty.  You need much larger samples before you even begin to approach any level of relevance. 

[tarheelsxxiii]

I posted this in another thread and think it's meaningful.

For the 2017-2018 season, Boston's most used lineup was excellent:
Brown-Tatum-Horford-Baynes-KYRIE (354 minutes played). 

The Celtics had only two better lineups all season, actually; at least with sample sizes reasonable enough to interpret:

Brown-Tatum-Horford-Baynes-ROZIER (115 mins)

Brown-Tatum-Horford-Baynes-SMART (43 mins)

That means the Celtics' best lineups consisted of their starting 5 after swapping out Kyrie for Rozier or Smart.  Pretty powerful data.



Were the Celtics very good with Kyrie? Yes. He's a net negative defender, but very good scorer that offsets his defense.  But that the Celtics are better with two younger, borderline-starting PGs speaks volumes about his value as a teammate.

Add the fact that Kyrie (injury-prone, very reliant on quickness) will be due somewhere ~$40 million in his mid 30s, and there is plenty of cause for concern. 

https://hardwoodhoudini.com/2018/11/12/boston-celtics-holiday-gift-guide/

It doesn't really contribute much since it's 2017-18 stats. For instance, this year, when 5 man combos have played at least 20 minutes together, Kyrie is in 12 of the top 16 best net rated 5 man combos this year.

Here's the important stat: Kyrie is literally the only player on the team who when he sits the Celtics are outscored by the opponent per 100 possessions.

Kyrie also has the highest raw +/- and overall net rating.  The team is simply a considerable amount better with Kyrie playing than not playing, small sample sizes be [dang]ed.

https://stats.nba.com/team/1610612738/onoffcourt-summary/?sort=DIFF_NET_RATING&dir=1

That data is meaningful.  I agree that he is an overall plus, primarily because he's an elite scorer.  That said, the team was better over an entire season when you swapped him out for either of his two back-ups, and that is a much larger sample size.

You also have to consider context.  The 2017-2018 data was seen in the context of a deep playoff run made in Kyrie's absence; your 2018-2019 data is shown in the context of a slow start to the season in which nearly every player (minus Mook) has been worse since Kyrie returned.  That's also meaningful.

115 and 43 minute sample sizes have nearly zero statistical value, they are merely a novelty.  You need much larger samples before you even begin to approach any level of relevance.

They are relatively small sample sizes, yet larger than any other being cited in the thread.  So if we're going to quantify this argument in any way, that's where you should start... and anything else should be similarly criticized. 

That is certainly not "zero statistical value," but I'm curious to know how many minutes of data would be meaningful to you.  Further, if you maintain this position, will you render any data from the first half of this 2018-2019 season equally meaningless?

[KGs Knee]

I posted this in another thread and think it's meaningful.

For the 2017-2018 season, Boston's most used lineup was excellent:
Brown-Tatum-Horford-Baynes-KYRIE (354 minutes played). 

The Celtics had only two better lineups all season, actually; at least with sample sizes reasonable enough to interpret:

Brown-Tatum-Horford-Baynes-ROZIER (115 mins)

Brown-Tatum-Horford-Baynes-SMART (43 mins)

That means the Celtics' best lineups consisted of their starting 5 after swapping out Kyrie for Rozier or Smart.  Pretty powerful data.



Were the Celtics very good with Kyrie? Yes. He's a net negative defender, but very good scorer that offsets his defense.  But that the Celtics are better with two younger, borderline-starting PGs speaks volumes about his value as a teammate.

Add the fact that Kyrie (injury-prone, very reliant on quickness) will be due somewhere ~$40 million in his mid 30s, and there is plenty of cause for concern. 

https://hardwoodhoudini.com/2018/11/12/boston-celtics-holiday-gift-guide/

It doesn't really contribute much since it's 2017-18 stats. For instance, this year, when 5 man combos have played at least 20 minutes together, Kyrie is in 12 of the top 16 best net rated 5 man combos this year.

Here's the important stat: Kyrie is literally the only player on the team who when he sits the Celtics are outscored by the opponent per 100 possessions.

Kyrie also has the highest raw +/- and overall net rating.  The team is simply a considerable amount better with Kyrie playing than not playing, small sample sizes be [dang]ed.

https://stats.nba.com/team/1610612738/onoffcourt-summary/?sort=DIFF_NET_RATING&dir=1

That data is meaningful.  I agree that he is an overall plus, primarily because he's an elite scorer.  That said, the team was better over an entire season when you swapped him out for either of his two back-ups, and that is a much larger sample size.

You also have to consider context.  The 2017-2018 data was seen in the context of a deep playoff run made in Kyrie's absence; your 2018-2019 data is shown in the context of a slow start to the season in which nearly every player (minus Mook) has been worse since Kyrie returned.  That's also meaningful.

This is factually incorrect.

Go back a page or two and you'll see I posted the 2017-18 data, which shows the Celtics had a better point differential and winning percentage with Kyrie in the lineup as opposed to out of the lineup.  They outscored their opponents by 3 points more per game in games Kyrie played in.

The data you posted isn't team data, it's lineup specific data, and includes games that Kyrie both played in and did not play in.  A few select lineups having better numbers without Kyrie, and with either Smart or Rozier in his stead, does not equate to the team as a whole playing better.

[The Oracle]

I posted this in another thread and think it's meaningful.

For the 2017-2018 season, Boston's most used lineup was excellent:
Brown-Tatum-Horford-Baynes-KYRIE (354 minutes played). 

The Celtics had only two better lineups all season, actually; at least with sample sizes reasonable enough to interpret:

Brown-Tatum-Horford-Baynes-ROZIER (115 mins)

Brown-Tatum-Horford-Baynes-SMART (43 mins)

That means the Celtics' best lineups consisted of their starting 5 after swapping out Kyrie for Rozier or Smart.  Pretty powerful data.



Were the Celtics very good with Kyrie? Yes. He's a net negative defender, but very good scorer that offsets his defense.  But that the Celtics are better with two younger, borderline-starting PGs speaks volumes about his value as a teammate.

Add the fact that Kyrie (injury-prone, very reliant on quickness) will be due somewhere ~$40 million in his mid 30s, and there is plenty of cause for concern. 

https://hardwoodhoudini.com/2018/11/12/boston-celtics-holiday-gift-guide/

It doesn't really contribute much since it's 2017-18 stats. For instance, this year, when 5 man combos have played at least 20 minutes together, Kyrie is in 12 of the top 16 best net rated 5 man combos this year.

Here's the important stat: Kyrie is literally the only player on the team who when he sits the Celtics are outscored by the opponent per 100 possessions.

Kyrie also has the highest raw +/- and overall net rating.  The team is simply a considerable amount better with Kyrie playing than not playing, small sample sizes be [dang]ed.

https://stats.nba.com/team/1610612738/onoffcourt-summary/?sort=DIFF_NET_RATING&dir=1

That data is meaningful.  I agree that he is an overall plus, primarily because he's an elite scorer.  That said, the team was better over an entire season when you swapped him out for either of his two back-ups, and that is a much larger sample size.

You also have to consider context.  The 2017-2018 data was seen in the context of a deep playoff run made in Kyrie's absence; your 2018-2019 data is shown in the context of a slow start to the season in which nearly every player (minus Mook) has been worse since Kyrie returned.  That's also meaningful.

115 and 43 minute sample sizes have nearly zero statistical value, they are merely a novelty.  You need much larger samples before you even begin to approach any level of relevance.

They are relatively small sample sizes, yet larger than any other being cited in the thread.  So if we're going to quantify this argument in any way, that's where you should start... and anything else should be similarly criticized. 

That is certainly not "zero statistical value," but I'm curious to know how many minutes of data would be meaningful to you.  Further, if you maintain this position, will you render any data from the first half of this 2018-2019 season equally meaningless?

They are not relatively small, they are extremely small irrelevant samples.  Here is the numbers from the most used lineup during the 2015-2016 season, followed by their numbers broken down by month to show how wildly variable these small samples can be.

Johnson/Bradley/Crowder/Sullinger/Thomas  played 723 minutes and had a net rating of +2.3

through November 2015   117 minutes  -11.3 net rating
            December            160             +15.0
            January                158              -1.7
            February              157              -2.7
            March                   77              +14.4
            April                     54               +4.6

You don't even begin to reach any true relevance until you have much larger, more stable samples. 

[nickagneta]

I posted this in another thread and think it's meaningful.

For the 2017-2018 season, Boston's most used lineup was excellent:
Brown-Tatum-Horford-Baynes-KYRIE (354 minutes played). 

The Celtics had only two better lineups all season, actually; at least with sample sizes reasonable enough to interpret:

Brown-Tatum-Horford-Baynes-ROZIER (115 mins)

Brown-Tatum-Horford-Baynes-SMART (43 mins)

That means the Celtics' best lineups consisted of their starting 5 after swapping out Kyrie for Rozier or Smart.  Pretty powerful data.



Were the Celtics very good with Kyrie? Yes. He's a net negative defender, but very good scorer that offsets his defense.  But that the Celtics are better with two younger, borderline-starting PGs speaks volumes about his value as a teammate.

Add the fact that Kyrie (injury-prone, very reliant on quickness) will be due somewhere ~$40 million in his mid 30s, and there is plenty of cause for concern. 

https://hardwoodhoudini.com/2018/11/12/boston-celtics-holiday-gift-guide/

It doesn't really contribute much since it's 2017-18 stats. For instance, this year, when 5 man combos have played at least 20 minutes together, Kyrie is in 12 of the top 16 best net rated 5 man combos this year.

That data is non-contributory to the conversation because it is from the 2017-2018 season?  I am not clear what you are citing, but it is necessarily half the sample size; moreover, assuming your correct, any interpretation of this year's data still needs to take into account last year's.

You're doing exactly what you're criticizing moranis of, yet rather than a stretch of games, you'd like to omit an entire season?

Moranis, please, please, for the sake of mathematics, please stop with the " unsustainable 15-2 start needs to be disregarded". These discussions don't work that way and I am sure you know that. If it was unsustainable it wouldn't have happened. It did happen so you have to count it.
 

You can't discount the Lakers record 33 game winning streak and argue that it wasn't sustainable so we should look at that Lakers juggernaut as a 36 win team. You can't discount the Philly 16 game win streak to end last season and say they were really a barely above .500 team in reality.

You just can't do that. If the games were played then they count and have to be taken into consideration.

So please, for the love of God, stop with the "you have to ignore this very large swarth of games" stuff simply to try to prove a point. It doesn't work that way.

Last year's stats have nothing to do with this year's stats and so are not contributory when discussing this year. And when trying to discuss this year, you absolutely do not need to look at last year's stats.

And my telling Moranis you just can not ignore 18 games from lahst year in evaluating last year's numbers isn't even close to the same as me saying you can't bring in last year's numbers when reviewing this year's numbers. If I am reviewing last year, I take into consideration all last year's numbers. If I am reviewing this year, I review all this year's numbers. I don't conflate the numbers of both years to discuss the combined numbers as one year.

This year is this year.

Last year was last year.

Different players. Different teams. Different variables.

I mean if I want to discuss Hayward's impact on the Celtics last year, I can only bring up last year. It makes no sense to bring up this year's stats or his last year in Utah's stats to discuss his impact on the 2017-18 Celtics. He got hurt and missed last year so his impact is he hurt the C's last year because he didn't contribute.

[ozgod]

No sure about anyone else but I’m sure glad we haven’t traded Kyrie yet 

[gouki88]

No sure about anyone else but I’m sure glad we haven’t traded Kyrie yet 

Lol. Glad this thread got revived

[Moranis]

No sure about anyone else but I’m sure glad we haven’t traded Kyrie yet 

Lol. Glad this thread got revived

Fabulous game again and yet Boston was actually better again without him on the floor, just as they were in the Toronto game i.e. +4 in 36 minutes is worse than +2 in 12 minutes, just as +4 in 38 minutes is worse than +5 in 10 minutes.  This is where it is always tricky in these things and were many of the same arguments being made when IT4 was torching people, especially in the 4th quarter, yet the team often performed better in the only place it really matters when IT4 was on the bench.

[gouki88]

No sure about anyone else but I’m sure glad we haven’t traded Kyrie yet 

Lol. Glad this thread got revived

Fabulous game again and yet Boston was actually better again without him on the floor, just as they were in the Toronto game i.e. +4 in 36 minutes is worse than +2 in 12 minutes, just as +4 in 38 minutes is worse than +5 in 10 minutes.  This is where it is always tricky in these things and were many of the same arguments being made when IT4 was torching people, especially in the 4th quarter, yet the team often performed better in the only place it really matters when IT4 was on the bench.

Here we go

[Q_FBE]

No sure about anyone else but I’m sure glad we haven’t traded Kyrie yet 

Lol. Glad this thread got revived

No comment.

[Phantom255x]

No sure about anyone else but I’m sure glad we haven’t traded Kyrie yet 

Lol. Glad this thread got revived

Fabulous game again and yet Boston was actually better again without him on the floor, just as they were in the Toronto game i.e. +4 in 36 minutes is worse than +2 in 12 minutes, just as +4 in 38 minutes is worse than +5 in 10 minutes.  This is where it is always tricky in these things and were many of the same arguments being made when IT4 was torching people, especially in the 4th quarter, yet the team often performed better in the only place it really matters when IT4 was on the bench.

Yeah if you actually think we're winning tonight and against Toronto with Kyrie on the bench, then frankly I don't know what to tell you. Without Kyrie I honestly think we get blown out in both games the way they unfolded (yes, even tonight w/o Kyrie).

[gouki88]

No sure about anyone else but I’m sure glad we haven’t traded Kyrie yet 

Lol. Glad this thread got revived

Fabulous game again and yet Boston was actually better again without him on the floor, just as they were in the Toronto game i.e. +4 in 36 minutes is worse than +2 in 12 minutes, just as +4 in 38 minutes is worse than +5 in 10 minutes.  This is where it is always tricky in these things and were many of the same arguments being made when IT4 was torching people, especially in the 4th quarter, yet the team often performed better in the only place it really matters when IT4 was on the bench.

Yeah if you actually think we're winning tonight and against Toronto with Kyrie on the bench, then frankly I don't know what to tell you. Without Kyrie I honestly think we get blown out (yes, even tonight w/o Kyrie).

It's really funny. Using stats from when our bench, which is loaded with starting calibre players, pick on Toronto's lineup without Anunoby and JV (not to mention Kawhi, Lowry and Ibaka) or abuse a lowly Memphis bench to somehow imply we're better without Kyrie is just hilarious. Some crazy context blinders being worn by Moranis

[Vox_Populi]

No sure about anyone else but I’m sure glad we haven’t traded Kyrie yet 

Lol. Glad this thread got revived

Fabulous game again and yet Boston was actually better again without him on the floor, just as they were in the Toronto game i.e. +4 in 36 minutes is worse than +2 in 12 minutes, just as +4 in 38 minutes is worse than +5 in 10 minutes.  This is where it is always tricky in these things and were many of the same arguments being made when IT4 was torching people, especially in the 4th quarter, yet the team often performed better in the only place it really matters when IT4 was on the bench.

You know there are metrics that regularize what you're amateurishly trying to, much more robustly, right? Over the season, Irving has been a massive positive for Boston.