Celtics in 18-19

Eddie Jurak

canderson-lite
Lifetime Member
SoSH Member
Dec 12, 2002
44,475
Melrose, MA
Brian Robb had a nice article breaking down the responsibility for the loss. Here were his 5 reasons:

5. Lack of hustle (featuring Marcus Morris).
4.Tatum thinking he had a mismatch with Beverley
3. Poor shot selection/decision making from the usual suspects (Rozier, Brown)
2. Rotations (Stevens had no answer for the Clippers 3 small quick guard lineups [involving 3 of Williams, Beverley, Alexander, Shamet] and sat Brown for long stretches while watching Tatum and Hayward get torched against this group)
1. Transition defense

Also, I think it's time for Morris to sit.
 

nighthob

Member
SoSH Member
Jul 15, 2005
12,678
He's owed $30 million next year if he wants it. Why not exercise the option, play one more year with Boston (or possibly be traded) and then sign a 3 year deal as a FA?
Because every less productive year that his next team is signing him for is going to reduce his annual pay. A four year deal this summer might carry a total premium more than a three year deal following him picking up the thirty million option. The glimpses of age showing in his game this year are definitely a warning sign.
 

Auger34

used to be tbb
SoSH Member
Apr 23, 2010
9,273
Brian Robb had a nice article breaking down the responsibility for the loss. Here were his 5 reasons:

5. Lack of hustle (featuring Marcus Morris).
4.Tatum thinking he had a mismatch with Beverley
3. Poor shot selection/decision making from the usual suspects (Rozier, Brown)
2. Rotations (Stevens had no answer for the Clippers 3 small quick guard lineups [involving 3 of Williams, Beverley, Alexander, Shamet] and sat Brown for long stretches while watching Tatum and Hayward get torched against this group)
1. Transition defense

Also, I think it's time for Morris to sit.
I’m glad Brian wrote this article because I think point #5 highlights the biggest problem with Morris trying to talk to the media about any of this teams problems.

Who the fuck is he and why does he think that he has the clout to say any of these things? As others have pointed out....Morris himself is probably the biggest offender of the lack of team spirit that he wants to blame on others.

I’m not going to relitigate the whole yelling at Jaylen Brown episode, but it’s interesting that a guy who also has a pention to not hustle thought he was the best person to yell at someone else for not hustling.

I really hope that Brad takes him up on the “I don’t care if I start I just want to win!” Portion of his diatribe last night. I think it was self righteous bullshit on Morris’s part, I know that he was the player that many writers/fans targeted as someone that would get traded this summer because he DOES care so much about playing time but alas, however he’s on record as saying it and probably deserves to get less playing time anyway.

I really hope this thing turns around but the decisions of some players to always Air things out through the media has made this season not very enjoyable for me as a fan. I can’t omagine what the other dozen or so players in the locker room think
 

Jimbodandy

Member
SoSH Member
Jan 31, 2006
11,403
around the way
Morris is who we thought he was. He fooled some of us with his BIP luck early in the year. Jaylen needed to get out of his own head, so the switch was a logical move (also Smart for GH, who was clearly struggling).

What Morris does not get is that considering yourself a leader and being a leader are two different things. Being a leader includes leading by example. You can't bitch at guys for doing things that you do.

The kids need to tune him out successfully, and they will be fine. I worry about his influence on Rozier a bit, but the other two will have to get over him. Working with guys who are a bit off is part of being a professional.
 
Last edited:

HomeRunBaker

bet squelcher
SoSH Member
Jan 15, 2004
30,096
Morris is who we thought he was. He fooled some of us with his BIP luck early in the year. Jaylen needed to get out of his own head, so the switch was a logical move (also Smart for GH, who was clearly struggling).

What Morris does not get is that considering yourself a leader and being a leader are two different things. Being a leader includes leading by example. You can't bitch at guys for doing things that you do.

The kids need to tune him out successfully, and they will be fine. I worry about his influence on Rozier a bit, but the other two will have to get over him. Working with guys who are a bit off is part of being a professional.
So MaMo, the guy/leader who has been calling out Tatum and Jaylen all year for lackadaisical play, should be tuned out by young guys who haven't shown consistent mental toughness? Each player has a similar iso-style game but the issue MaMo and Kyrie have had with the youth on this team is their lack of toughness and preparedness to compete every night, aka being soft. He isn't wrong.
 

Jimbodandy

Member
SoSH Member
Jan 31, 2006
11,403
around the way
So MaMo, the guy/leader who has been calling out Tatum and Jaylen all year for lackadaisical play, should be tuned out by young guys who haven't shown consistent mental toughness? Each player has a similar iso-style game but the issue MaMo and Kyrie have had with the youth on this team is their lack of toughness and preparedness to compete every night, aka being soft. He isn't wrong.
He's right about toughness. But his game is often selfish, and he takes plays off too. It's not selfish because it's iso. It's selfish because he takes bad shots without even attempting to move the ball, i.e. play the offense. That was fine last year when he was the only second team option, and it was fine this year when his TS was off the charts.

I don't think that they should tune him out. I said that wrong. I just mean that they can't let it sap their confidence and take them out of their games.

His toughness and edge is a good thing. But the Hanibal Lecter bullshit has to stop. I wouldn't take Kryie complaining about defense either, for that matter. Play some first, then complain.
 
Last edited:

DJnVa

Dorito Dawg
SoSH Member
Dec 16, 2010
53,840
A 5 game winning streak will take care of this crap. As would finishing with the #3 seed.
With 3 of the 5 against Philly, Milwaukee, and Toronto....

Win 2 of those 3 and we're happy. Lose 2 of 3 or worse and it's Armageddon.

ALL THEY HAD TO DO WAS CLOSE OUT GAMES WITH 18 AND 28 POINT LEADS!! WTF!
 

BigSoxFan

Member
SoSH Member
May 31, 2007
47,090
With 3 of the 5 against Philly, Milwaukee, and Toronto....

Win 2 of those 3 and we're happy. Lose 2 of 3 or worse and it's Armageddon.

ALL THEY HAD TO DO WAS CLOSE OUT GAMES WITH 18 AND 28 POINT LEADS!! WTF!
Let this be a lesson to these Celtics. No more big leads!
 

Auger34

used to be tbb
SoSH Member
Apr 23, 2010
9,273
So MaMo, the guy/leader who has been calling out Tatum and Jaylen all year for lackadaisical play, should be tuned out by young guys who haven't shown consistent mental toughness? Each player has a similar iso-style game but the issue MaMo and Kyrie have had with the youth on this team is their lack of toughness and preparedness to compete every night, aka being soft. He isn't wrong.
I don’t know if they should tune him out but he should definitely follow his own advice first before spreading it onto others.

As Brian Robb wrote about Morris took off 4-5 plays that directly led to buckets before his speech about having fun.

As others pointed out, Morris is one of, if not the biggest offender of never looking like he has fun but still thought he should tell the media about not having fun.

Jimbo already commented on his offense and why his shots are often bad (nothing to do with iso heavy, it’s that he pretty much refuses to pass the ball)

So, no I don’t think they should tune him out and I think in a lot of cases he was right with what he was calling the young guys out about.
It’s that he needs to look in the mirror and figure out if he’s practicing what he’s preaching and if he’s not (which clearly he isn’t all of the time) he needs to focus on himself before calling out anyone else. It’s pretty emblematic of the Celtics team as a whole season this year honestly
 

wade boggs chicken dinner

Member
SoSH Member
Mar 26, 2005
30,491
Here are some more numbers to chew on, courtesy of Himmelsbach: https://www.bostonglobe.com/sports/celtics/2019/02/10/why-has-celtics-defense-been-poor-second-and-third-quarters/8a5tfshtv75ZuM5ERqIHMJ/story.html.
  • Rough proxy for Cs offense finds it consistent through the quarters. 4Q Ortg of 113.1 is the highest quarter; lower is 2Q, with ORtg of 109.7.
  • DRtg not as consistent and by quarter is: 1Q = 99.2; 2Q = 109.3; 3Q = 109.5; and 4Q = 104.5. This is somewhat consistent because the players who play the bulk of the 2Q are not the players who play the bulk of the 3Q.
  • In 1Q, MaMo and Al have great DRtg numbers: 96.7 and 97.2, respectively
  • However, MaMo has DRtg of 111.3 in 2Q and 111.6 in 3Q.
  • Al has DRtg of 110.8 in 2Q and 112.1 in 3Q. Note that Al has a team-worst 105.7 DRtg.
  • "When the Celtics’ defense has been gashed this season, it is usually related to some combination of Horford, Hayward, and Morris being on the court together. Of the team’s 16 worst defensive ratings among two-man groups with at least 250 minutes played together, 15 have Hayward, Horford or Morris in them."
  • GH and MaMo have been Cs worst defensive pairing with DRtg of 111.3 over 446 minutes together.
  • GH and Al are second-worst, with a 108.3 DRtg over 528 minutes.
  • MaMo and Al are fourth-worst, with a 107.6 rating over 762 minutes together.
  • "When isolated to the troubling second quarters, the defensive numbers are even more grisly: Hayward/Horford (125.1), Hayward/Morris (116.7), and Horford/Morris (114.9) have been the team’s three worst pairings with at least 50 second-quarter minutes together. In third quarters, Horford and Morris have a team-worst 118.7 defensive rating, followed by Morris/Terry Rozier and Horford/Smart."
From my memory of games, I think that Brad has been playing GH and MaMo together to close out a lot of 2Q and 4Q. That probably should stop.

I will also note that while everyone I'm sure is rooting for GH, and I know Brad is playing the long game, but if I'm on the Cs, having him be first guy off the bench would be tough to digest day-in and day-out.
 

cardiacs

Admires Neville Chamberlain
SoSH Member
Jul 15, 2005
2,993
Milford, CT
Jaylen needs to start again, IMO. It isn't the end of the world for this team and I am still optimistic they will get it together before the playoffs. We just lost two in a row we should have won but before that we won 10 of 11 and the only loss was against GS. Have a good showing in the playoffs and rework the roster for next years with the lessons learned about chemistry.
 

joe dokes

Member
SoSH Member
Jul 18, 2005
30,238
So, no I don’t think they should tune him out and I think in a lot of cases he was right with what he was calling the young guys out about.
It’s that he needs to look in the mirror and figure out if he’s practicing what he’s preaching and if he’s not (which clearly he isn’t all of the time) he needs to focus on himself before calling out anyone else. It’s pretty emblematic of the Celtics team as a whole season this year honestly
Tommy Heinsohn said something to that effect at the end of the LAC game. Gorman asked him something like "what should a leader do here"? Tommy replied, "Stand up in the middle of the room and ask all his teammates what HE (the guy in the middle of the room) needs to do differently to help the team have better results.
 

TripleOT

Member
SoSH Member
Jul 4, 2007
7,758
I put the blame mostly on the FA veterans. Morris is a journeyman in a contract year. He's was shooting lights out, but since he pushed Brown in a huddle, he's playing lousy, 22-70 from three, 31.4%, and 66 for 160 overall, 41.2%.

Kyrie turned himself into a distraction when he called out his young teammates in public, and then again when backsliding on his commitment to staying in Boston.

Rozier is a RFA and deserves blame for playing like trash with the second unit. Tatum for trying to play too much iso ball, and Brown in the beginning of the season for not handling his demotion to the bench well.

I was a big Morris guy, but as also mentioned by other posters, he's the last Celtic who should be spouting off about selfishness. Theis is in a contract year, and has made chump change the past two years with Boston. He doesn't get a lot of minutes, even though he's having a monster year statistically, while only playing 15.2 mpg (58/47% shooting, 61.7eFG. If anyone in that locker room should be griping, it's Theis.

I'd like to see more of Theis and Baynes if he can ever get out of the MASH unit, especially if Morris isn't willing to muck it up in the paint. Horford needs to play at least half his minutes with another big body, or he's going to break down again.
 

BaseballJones

ivanvamp
SoSH Member
Oct 1, 2015
24,375
First world problems for sure. Boston is 35-21 (.625), good for 5th in the conference. Seems bad, given expectations. But they have the 3rd best point differential in the league. They're #2 in the NBA in opposing FG% and 2nd in opposing adjusted FG%. They're also 5th in the NBA in adjusted FG%. They suck at rebounding, but they've struggled with that in the past and still been really good.

The Celtics are a really good NBA team. They're not as good as most of us hoped for or expected. But they're really good. They have things to work out, but this is a good team. A really good team.
 

DeJesus Built My Hotrod

Well-Known Member
Lifetime Member
SoSH Member
Dec 24, 2002
48,205
Here are some more numbers to chew on, courtesy of Himmelsbach: https://www.bostonglobe.com/sports/celtics/2019/02/10/why-has-celtics-defense-been-poor-second-and-third-quarters/8a5tfshtv75ZuM5ERqIHMJ/story.html.
  • Rough proxy for Cs offense finds it consistent through the quarters. 4Q Ortg of 113.1 is the highest quarter; lower is 2Q, with ORtg of 109.7.
  • DRtg not as consistent and by quarter is: 1Q = 99.2; 2Q = 109.3; 3Q = 109.5; and 4Q = 104.5. This is somewhat consistent because the players who play the bulk of the 2Q are not the players who play the bulk of the 3Q.
  • In 1Q, MaMo and Al have great DRtg numbers: 96.7 and 97.2, respectively
  • However, MaMo has DRtg of 111.3 in 2Q and 111.6 in 3Q.
  • Al has DRtg of 110.8 in 2Q and 112.1 in 3Q. Note that Al has a team-worst 105.7 DRtg.
  • "When the Celtics’ defense has been gashed this season, it is usually related to some combination of Horford, Hayward, and Morris being on the court together. Of the team’s 16 worst defensive ratings among two-man groups with at least 250 minutes played together, 15 have Hayward, Horford or Morris in them."
  • GH and MaMo have been Cs worst defensive pairing with DRtg of 111.3 over 446 minutes together.
  • GH and Al are second-worst, with a 108.3 DRtg over 528 minutes.
  • MaMo and Al are fourth-worst, with a 107.6 rating over 762 minutes together.
  • "When isolated to the troubling second quarters, the defensive numbers are even more grisly: Hayward/Horford (125.1), Hayward/Morris (116.7), and Horford/Morris (114.9) have been the team’s three worst pairings with at least 50 second-quarter minutes together. In third quarters, Horford and Morris have a team-worst 118.7 defensive rating, followed by Morris/Terry Rozier and Horford/Smart."
From my memory of games, I think that Brad has been playing GH and MaMo together to close out a lot of 2Q and 4Q. That probably should stop.

I will also note that while everyone I'm sure is rooting for GH, and I know Brad is playing the long game, but if I'm on the Cs, having him be first guy off the bench would be tough to digest day-in and day-out.
Thank you for sharing - very interesting.

While it would be ideal if the Cs players refrained from trying to "lead" via the press and took advice similar to what some in this thread are proposing with regards to self/team evaluation, the most realistic fix is likely from adjusted rotations.

Stevens likely sees data that shows that Horford, Hayward and Morris are relatively poor together defensively so we should expect him to change things, all else equal. Of course, he is managing people and not raw inputs so all else isn't truly equal. But after the last game, he acknowledged that he had to something different.

In short, if the Celtics hope to play deep into the playoffs, it won't be from better locker room chemistry which is difficult if not impossible to engineer especially in season. Instead it will be as a result of rotational adjustments and fortunately those are easier to implement and measure.
 

Big John

New Member
Dec 9, 2016
2,086
I wish folks would not keep bringing up point differential as if it meant something. Win game 1 by 50 points and lost the next three by a combined 6 points and your point differential is a gaudy +11. Who cares if you can beat up on bad teams if you can't win the close games against good teams (not to mention blowing big leads against mediocre teams)? What matters is wins and losses. Milwaukee's point differential doesn't matter. What matters is that they have a 7 game lead on the Celtics in the loss column.
 

Sam Ray Not

Member
SoSH Member
Jul 19, 2005
8,848
NYC
No Kyrie tomorrow
No idea how serious that "strain" was, but you might as well shut him down through the AS break, no? (i.e. v. Philly and Detroit). Have to think he'd be crazy to play in the AS game, too, especially with the Bucks looming right after the break. Seems like a perfect moment for a nice 12-day break.
 

benhogan

Granite Truther
SoSH Member
Nov 2, 2007
20,112
Santa Monica
Here are some more numbers to chew on, courtesy of Himmelsbach: https://www.bostonglobe.com/sports/celtics/2019/02/10/why-has-celtics-defense-been-poor-second-and-third-quarters/8a5tfshtv75ZuM5ERqIHMJ/story.html.
  • Rough proxy for Cs offense finds it consistent through the quarters. 4Q Ortg of 113.1 is the highest quarter; lower is 2Q, with ORtg of 109.7.
  • DRtg not as consistent and by quarter is: 1Q = 99.2; 2Q = 109.3; 3Q = 109.5; and 4Q = 104.5. This is somewhat consistent because the players who play the bulk of the 2Q are not the players who play the bulk of the 3Q.
  • In 1Q, MaMo and Al have great DRtg numbers: 96.7 and 97.2, respectively
  • However, MaMo has DRtg of 111.3 in 2Q and 111.6 in 3Q.
  • Al has DRtg of 110.8 in 2Q and 112.1 in 3Q. Note that Al has a team-worst 105.7 DRtg.
  • "When the Celtics’ defense has been gashed this season, it is usually related to some combination of Horford, Hayward, and Morris being on the court together. Of the team’s 16 worst defensive ratings among two-man groups with at least 250 minutes played together, 15 have Hayward, Horford or Morris in them."
  • GH and MaMo have been Cs worst defensive pairing with DRtg of 111.3 over 446 minutes together.
  • GH and Al are second-worst, with a 108.3 DRtg over 528 minutes.
  • MaMo and Al are fourth-worst, with a 107.6 rating over 762 minutes together.
  • "When isolated to the troubling second quarters, the defensive numbers are even more grisly: Hayward/Horford (125.1), Hayward/Morris (116.7), and Horford/Morris (114.9) have been the team’s three worst pairings with at least 50 second-quarter minutes together. In third quarters, Horford and Morris have a team-worst 118.7 defensive rating, followed by Morris/Terry Rozier and Horford/Smart."
From my memory of games, I think that Brad has been playing GH and MaMo together to close out a lot of 2Q and 4Q. That probably should stop.

I will also note that while everyone I'm sure is rooting for GH, and I know Brad is playing the long game, but if I'm on the Cs, having him be first guy off the bench would be tough to digest day-in and day-out.
Once again thanks Wade, just more evidence of what the problem is here.

Masquerading Horford/MaMo OR Horford/Hayward as a 5/4 creates defensive mismatches throughout the lineup and will continue to not work. We get away with it when MaMo shoots at unrealistic levels. It will get crumpled/crushed by Philly in any kind of playoff series.

Also, Jaylen Brown starting for MaMo isn't the answer, we tried that already. Sticking Jaylen in the corner to be the 4th scoring option with Kyrie, Al and Tatum is beyond idiotic. We need size, not another ball stopping, me-first offensive wing.

Here is my mea culpa with this team: I had Brad Stevens as a top 3 hoops Head Coach (NBA & NCAA) entering the season. While he is tremendous with young, scrappy players/teams, he has been below average with a group of established players and up/coming stars. Its not hard to imagine several established NBA coaches could have done a better job with this group. This may be a learning/transition year for Brad. By no means am I calling for his scalp, even, if we get bounced in round 1.
 
Last edited:

Strike4

Member
SoSH Member
Jul 19, 2005
3,895
Portland, Maine
First world problems for sure. Boston is 35-21 (.625), good for 5th in the conference. Seems bad, given expectations. But they have the 3rd best point differential in the league. They're #2 in the NBA in opposing FG% and 2nd in opposing adjusted FG%. They're also 5th in the NBA in adjusted FG%. They suck at rebounding, but they've struggled with that in the past and still been really good.

The Celtics are a really good NBA team. They're not as good as most of us hoped for or expected. But they're really good. They have things to work out, but this is a good team. A really good team.
This is true and is good for perspective. BUT - we can pretty easily point to the issues with almost every other team in the league. We know what's wrong with the Knicks, or the Pacers, or the Mavs, etc., and those teams can take steps to address the issues. It's proving to be really hard to figure out what's wrong with the Celtics. In years past you could say they didn't have the talent to compete with the better teams in the EC, but talent is no longer lacking. It's probably a combination of factors, but it's kind of alarming that glaring problems have persisted for most of the season and they haven't been fixed.
 

Average Game James

Member
SoSH Member
Apr 28, 2016
4,346
I wish folks would not keep bringing up point differential as if it meant something. Win game 1 by 50 points and lost the next three by a combined 6 points and your point differential is a gaudy +11. Who cares if you can beat up on bad teams if you can't win the close games against good teams (not to mention blowing big leads against mediocre teams)? What matters is wins and losses. Milwaukee's point differential doesn't matter. What matters is that they have a 7 game lead on the Celtics in the loss column.
Except for all its flaws, point differential is still a better predictor of future performance than current W/L records. Now, we aren’t looking at this in a vacuum, of course, and MIL being up 7 games in the loss column sets them up for an easier path in the playoffs with HCA... but, for all their inconsistencies, the Celtics still do have a winning record against the top 4 in the East.
 

Sam Ray Not

Member
SoSH Member
Jul 19, 2005
8,848
NYC
I wish folks would not keep bringing up point differential as if it meant something. Win game 1 by 50 points and lost the next three by a combined 6 points and your point differential is a gaudy +11. Who cares if you can beat up on bad teams if you can't win the close games against good teams (not to mention blowing big leads against mediocre teams)? What matters is wins and losses. Milwaukee's point differential doesn't matter. What matters is that they have a 7 game lead on the Celtics in the loss column.
W-L is what matters in terms of the standings, but net rating tends to be more indicative than W-L of a team's true talent level, so more predictive of what we can expect going forward. This is pretty basic stuff.

It's not a perfect predictor, of course, but it's far less imperfect than an emotional, recency-bias-influenced fan ranting on a fan board that net rating is bunk because his team just lost a couple of coin-flip games.
 

camneely

New Member
May 9, 2018
15
On the one hand, dealing with the turmoil and struggle that this season has presented may toughen the team up some for the stretch run. They're obviously just as talented as the other top EC teams (though in fairness we haven't yet seen them against the new look 76ers). When they're on they can hang with the best in the league. Maybe they'll be able to build on this experience, and rely on it when facing down adversity come playoff time? Or it will give them a bit of an edge over teams whose seasons were less rocky? Fingers crossed, I guess.

On the other hand, it might get worse before it gets better. Tomorrow night could be a bloodbath given Kyrie's absence.
 

bowiac

Caveat: I know nothing about what I speak
Lifetime Member
SoSH Member
Dec 18, 2003
12,945
New York, NY
To add to Sam Ray's point, while some teams do outperform or undeperform their W-L rating consistently (e.g. the Gasol/Conley/Z-Bo Grizzlies for years or the Kevin Love Wolves on the other end), this is somewhat rare, and is usually to a relatively limited degree.

I would expect the Celtics to slightly underperform their net rating since they get a lot of value out of having a good bench, but that's a pretty small factor.
 

HomeRunBaker

bet squelcher
SoSH Member
Jan 15, 2004
30,096
W-L is what matters in terms of the standings, but net rating tends to be more indicative than W-L of a team's true talent level, so more predictive of what we can expect going forward. This is pretty basic stuff.

It's not a perfect predictor, of course, but it's far less imperfect than an emotional, recency-bias-influenced fan ranting on a fan board that net rating is bunk because his team just lost a couple of coin-flip games.
Are you implying blowing a 28-point lead st home to the depleted Clippers a “coin-flip game?”

There is so much noise in those net ratings with so many stars missing/sitting out games. Of course we know reg season point diff means zero to a team like the Warriors but when it’s skewed against teams missing guys like LeBron for a month, Giannis for a week or Utah’s early season death schedule you can’t simply point to one number without making adjustments to it.
 

Sam Ray Not

Member
SoSH Member
Jul 19, 2005
8,848
NYC
Are you implying blowing a 28-point lead st home to the depleted Clippers a “coin-flip game?”
Pretty much, yeah. In fact I think it's a pretty classic example of a case where "fan PTSD" from a painful one-game sample clouds a reasonable accounting of the significance of that game in terms of the team's chances going forward. You'd have found similar recency bias on Golden State of Mind earlier this season after the Harden game.

There is so much noise in those net ratings with so many stars missing/sitting out games. Of course we know reg season point diff means zero to a team like the Warriors but when it’s skewed against teams missing guys like LeBron for a month, Giannis for a week or Utah’s early season death schedule you can’t simply point to one number without making adjustments to it.
Agreed with all this, but I'd also file it under "net rating is not a perfect predictor, but it's less imperfect than most others." It's a better baseline for predictions than W-L, for example. Of course, as you point out, it can get better still with more context and finer-grained analysis (both statistical and scouting-based).
 

RedOctober3829

Member
SoSH Member
Jul 19, 2005
55,298
deep inside Guido territory
Pretty much, yeah. In fact I think it's a pretty classic example of a case where "fan PTSD" from a painful one-game sample clouds a reasonable accounting of the significance of that game in terms of the team's chances going forward. You'd have found similar recency bias on Golden State of Mind earlier this season after the Harden game.


Agreed with all this, but I'd also file it under "net rating is not a perfect predictor, but it's less imperfect than most others." It's a better baseline for predictions than W-L, for example. Of course, as you point out, it can get better still with more context and finer-grained analysis (both statistical and scouting-based).
There is no way whatsoever that a team that leads by 28 points should lose a game ever.
 

Average Game James

Member
SoSH Member
Apr 28, 2016
4,346
There is no way whatsoever that a team that leads by 28 points should lose a game ever.
We’ve had blown 25 point leads in the playoffs in backs to back years. It happens, even to good teams. I want to do some research on this, but my suspicion is that large blown leads have become more common as 3 pt shooting has become a bigger share of overall shot attempts...

Edit: I’m also wondering if pace plays a role
 
Last edited:

Sam Ray Not

Member
SoSH Member
Jul 19, 2005
8,848
NYC
There is no way whatsoever that a team that leads by 28 points should lose a game ever.
Agreed from an emotional standpoint, but sh*t happens. In January the Warriors squandered a 20-point third quarter lead on their home floor over the depleted (CP3-less) Rockets, a bitter rival whom they should have been motivated to stomp, and lost in OT in the most crushing way imaginable.

With a month and a half of distance (plus a nice stretch of winning), it now looks like just another regular season loss on the schedule, but at the time it felt a lot more significant more than that. I suspect you'll feel the same about the Clips debacle if/when the Cs reel off a bunch in a row.

Edit: also, great point by AGJ about how the increased variance from higher-volume three-point shooting makes that kind of comeback/squander more commonplace.
 

HomeRunBaker

bet squelcher
SoSH Member
Jan 15, 2004
30,096
Yes while leads over 20-points in the NBA are not nearly as safe as the casual fan believes (just listen to those asking in game threads why a starter is in the game up 20 with 6 min to go) my point was how can you consider that a "coin flip game" when you are a heavy favorite prior to the opening tip and an enormous one throughout the game. My definition of a "coin flip game" was when the Warriors came to Boston, the point spread was like 1.5, it was tight the entire game, and whichever team executed down the stretch wins.
 

jon abbey

Shanghai Warrior
Moderator
SoSH Member
Jul 15, 2005
70,713
That was the single biggest comeback in Clippers history, not exactly commonplace.
 

Big John

New Member
Dec 9, 2016
2,086
W-L is what matters in terms of the standings, but net rating tends to be more indicative than W-L of a team's true talent level, so more predictive of what we can expect going forward. This is pretty basic stuff.

It's not a perfect predictor, of course, but it's far less imperfect than an emotional, recency-bias-influenced fan ranting on a fan board that net rating is bunk because his team just lost a couple of coin-flip games.
Is it basic stuff? It's better than W-L record itself using the same sample sizes? I predict based on the Celtics' .current .625 winning percentage over 56 games, that they will win between 51-52 regular season games this year. What does net rating predict?
 

Deathofthebambino

Drive Carefully
SoSH Member
Apr 12, 2005
41,947
it was the second time they set that record that week - they won after being down 25 to Detroit on 2/2
It wasn't even the same team. The team that beat the Celtics had 5 new players on the roster from the team came back from 25 earlier in the week. As Tommy said, these guys "haven't even had a chance to exchange phone numbers," and yet, they were able to come together and come back against the C's, while outscoring the Celtics 81-44 from the 4:00 minute mark of the 2nd quarter.

I can't imagine that's ever happened before to an NBA team around the trading deadline.
 

Average Game James

Member
SoSH Member
Apr 28, 2016
4,346
Is it basic stuff? It's better than W-L record itself using the same sample sizes? I predict based on the Celtics' .current .625 winning percentage over 56 games, that they will win between 51-52 regular season games this year. What does net rating predict?
53.5 based on point differential. I don’t know the net rating formula off the top of my head. Which doesn’t sound much different vs. your 51-52, but there’s only 26 games left. If they finish at 51 wins, that means they play at a 53 win full year pace for the rest of the year vs. a 59 win pace projected by point differential.
 

Deathofthebambino

Drive Carefully
SoSH Member
Apr 12, 2005
41,947
Agreed from an emotional standpoint, but sh*t happens. In January the Warriors squandered a 20-point third quarter lead on their home floor over the depleted (CP3-less) Rockets, a bitter rival whom they should have been motivated to stomp, and lost in OT in the most crushing way imaginable.

With a month and a half of distance (plus a nice stretch of winning), it now looks like just another regular season loss on the schedule, but at the time it felt a lot more significant more than that. I suspect you'll feel the same about the Clips debacle if/when the Cs reel off a bunch in a row.

Edit: also, great point by AGJ about how the increased variance from higher-volume three-point shooting makes that kind of comeback/squander more commonplace.
I think the bolded part is the issue. At any time this season, were you ever concerned that the Warriors wouldn't "figure" it out and eventually end up right where they need to be at the end of the season? With these Celtics, there has never been a point in this season where the fans should have become confident that was the case. They started the year 10-10, then won a bunch of games, but when you dig into those winning streaks, you realize that they haven't beaten a single team on the road with a .500 record since the 5th game of the season (and that was OKC, who started the season 0-4 at that point). So we know they can't win on the road against good teams, and now they are blowing 18-28 point leads against bad teams at home.

We've got a veteran throwing everyone under the bus and talking about how it's not fun to be a Celtic and hasn't been for a while. And on top of that, we're about to go on one of our most important road trips of the year, with Philly, Milwaukee and Toronto in the next like10 days, and may have do it without Kyrie.

I love the Celtics, and hadn't been as excited for a season to start as I was this year in recent memory, and it's been like a long, slow root canal with short periods in which the novocaine kicked in, but mostly, just painful.
 

Big John

New Member
Dec 9, 2016
2,086
53.5 based on point differential. I don’t know the net rating formula off the top of my head. Which doesn’t sound much different vs. your 51-52, but there’s only 26 games left. If they finish at 51 wins, that means they play at a 53 win full year pace for the rest of the year vs. a 59 win pace projected by point differential.
Well, the predictive validity of W-L improves over time, yes? If a team's record is 50-31 after 81 games, I can safely predict that the team will win between 50 and 51 games. Maybe net rating is a better predictor earlier in the season, although you still have to account for injuries, personnel changes, coaching changes and all the rest.
 

DeJesus Built My Hotrod

Well-Known Member
Lifetime Member
SoSH Member
Dec 24, 2002
48,205
There is no way whatsoever that a team that leads by 28 points should lose a game ever.
I will republish an academic study I posted in the game thread but it happens quite frequently in the NBA. Heck, the Rockets blew a 26 point lead at home on the same night the Cs lost to the Clippers.
 

DeJesus Built My Hotrod

Well-Known Member
Lifetime Member
SoSH Member
Dec 24, 2002
48,205
I short, the size of a lead doesn't matter as much as when a team is leading. Big leads in the first halves of NBA games tend to be whittled down over time as substitutions occur, adjustments are made and teams revert to the mean on offense and defense.
 

Sam Ray Not

Member
SoSH Member
Jul 19, 2005
8,848
NYC
Well, the predictive validity of W-L improves over time, yes?
Yep, and same goes for net rating. But since net rating — especially an adjusted net rating like BB-Ref's SRS that factors in strength of schedule, etc. — contains a smaller component of luck, it will always tend to have more predictive value (though as the sample gets larger, the two numbers should start to converge).

If a team's record is 50-31 after 81 games, I can safely predict that the team will win between 50 and 51 games.
Not sure if you're being serious here, but the only thing you're still "predicting" after 81 games is the team's final game. And yes, all other things being equal, a team's adjusted net rating should give you a better prediction about that one game (and its subsequent playoff performance, if applicable) than its W-L record.

you still have to account for injuries, personnel changes, coaching changes and all the rest.
Of course. Same goes for W-L.
 

Big John

New Member
Dec 9, 2016
2,086
Yep, and same goes for net rating. But since net rating — especially an adjusted net rating like BB-Ref's SRS that factors in strength of schedule, etc. — contains a smaller component of luck, it will always tend to have more predictive value (though as the sample gets larger, the two numbers should start to converge).


Not sure if you're being serious here, but the only thing you're still "predicting" after 81 games is the team's final game. And yes, all other things being equal, a team's adjusted net rating should give you a better prediction about that one game (and its subsequent playoff performance, if applicable) than its W-L record.


Of course. Same goes for W-L.
I'm just trying to get my head around the predictive validity of the two methods at various points in the season. So let's say there are 5 games to go, and you have a .500 team that has won many games by huge scores, but lost most of the close ones. Why would net rating be a better predictor of the team's record for the season at that point? Wouldn't that team tend to exhibit the same behaviors (i.e. win by alot, lose by a little) going forward? Now go back to that same team with 10 games to go, or 20 or 30 or 60 and ask the same question. I'm not implying that I know the answer, but if I were a gambler I'd love to know it.
 

mauf

Anderson Cooper × Mr. Rogers
Moderator
SoSH Member
Well, the predictive validity of W-L improves over time, yes? If a team's record is 50-31 after 81 games, I can safely predict that the team will win between 50 and 51 games. Maybe net rating is a better predictor earlier in the season, although you still have to account for injuries, personnel changes, coaching changes and all the rest.
During a Warriors game last week, they quoted a statistic that the Warriors had the 2nd-best record in games decided by 3 points or less since the beginning of the 2014-15 season.

Who was 1st? The Sacramento Kings.

It’s nice to think that there’s a skill to winning close games, but the empirical evidence is clear: it’s mostly luck, at least at the NBA level.

It wouldn’t surprise me if winning close games is a repeatable skill at the high-school level — just as it wouldn’t surprise me to learn that high-school baseball players play better defense when the pitcher works quickly, or that a high-school team can be coached so that it recovers more than 50% of total fumbles. It’s not necessarily that these things are purely luck-driven; it’s that the things one can do are straightforward enough that every professional team does them.
 

Sam Ray Not

Member
SoSH Member
Jul 19, 2005
8,848
NYC
Why would net rating be a better predictor of the team's record for the season at that point?
Obviously it's not a better predictor of its record "for the season" since you have 77 games of W-L baked into the cake. But it should give you a better prediction than W-L for those five remaining games. (For example, assuming five opponents with 0.0 net ratings, you'd expect a .500 team with a great Pythag record to go 3-2 rather than 2-3, whereas just looking at W-L would tell you it's a coin flip).

But sure, W-L is a fantastic "predictor" of W-L for the games that have already been played. :p
 

The Needler

New Member
Dec 7, 2016
1,803
I will republish an academic study I posted in the game thread but it happens quite frequently in the NBA. Heck, the Rockets blew a 26 point lead at home on the same night the Cs lost to the Clippers.
I short, the size of a lead doesn't matter as much as when a team is leading. Big leads in the first halves of NBA games tend to be whittled down over time as substitutions occur, adjustments are made and teams revert to the mean on offense and defense.
That study says a lead of 18 at halftime is 90 percent safe. The Celtics led by 21 at the half, and also as late as 9:08 in the 3rd. The article also acknowledges leads are even safer when held by the stronger team, presumably even moreso at home.

As to “happens quite frequently,” the biggest comeback in NBA history was 36 points, and there have only been 7 comebacks in history of more than 28 points.
 

NomarsFool

Member
SoSH Member
Dec 21, 2001
8,157
Just as leads can be more easily diminished in today's 3 point heavy era of basketball, I would imagine they can also be more easily accumulated. Two teams each throw up 5 three pointers. One team hits all of them and the other hits none of them and you've got yourself a 15 point lead. It's an extreme example, but stuff similar to that happens.
 

DeJesus Built My Hotrod

Well-Known Member
Lifetime Member
SoSH Member
Dec 24, 2002
48,205
That study says a lead of 18 at halftime is 90 percent safe. The Celtics led by 21 at the half, and also as late as 9:08 in the 3rd. The article also acknowledges leads are even safer when held by the stronger team, presumably even moreso at home.

As to “happens quite frequently,” the biggest comeback in NBA history was 36 points, and there have only been 7 comebacks in history of more than 28 points.
According to this link, your data is incorrect. There have been nine comebacks of 28 points or more. There have been 32 comebacks of 24 points or more in the last ~ 44 years in the NBA and of those 32 games, 29 have happened in the last 23 years. And the majority of those have happened in the aughts when rule changes made offense easier.

As for the study, the point isn't that a significant lead at halftime isn't safe. Its a response to the idea that teams should never lose big leads. The fact is that they can, and do.

We have three instances of teams at home losing leads of 25 points or more in the past week (Detroit vs LAC, Houston vs OKC and the Celtics game).

Finally, for those who keep citing the fact that the Clippers were running completely new guys out on the floor, that is a fact. But if you are going to note that, you have to also acknowledge that the Celtics lost their leading scorer and the one player who can reliably create his own shot for the second half of the game. Also, its not like a guy like Shamet had to learn an all new playbook on offense. The guy did exactly what he did for the 76ers which is go to the corner and wait for an open look.

The C's loss was bad on Saturday and people are entitled to be concerned. But that sort of loss is not as unprecedented as some might think.
 

lexrageorge

Member
SoSH Member
Jul 31, 2007
18,096
I'm just trying to get my head around the predictive validity of the two methods at various points in the season. So let's say there are 5 games to go, and you have a .500 team that has won many games by huge scores, but lost most of the close ones. Why would net rating be a better predictor of the team's record for the season at that point? Wouldn't that team tend to exhibit the same behaviors (i.e. win by alot, lose by a little) going forward? Now go back to that same team with 10 games to go, or 20 or 30 or 60 and ask the same question. I'm not implying that I know the answer, but if I were a gambler I'd love to know it.
I'll try the following hypothetical:

Team A is 0.500. All of their wins have been by 15 points or more, while all of their losses have been by 2 points.

Team B is 0.500. All of their losses have been by 15 points or more, while all of their wins have been by 2 points.

Both teams have played 60 games, all against the same opponents. So Team A's differential would be +390, while Team B's would be -390.

Team A and Team B are scheduled to play each other in a neutral site tomorrow night. Which team do you think would be most likely to win that game?

Similarly, looking ahead to the remaining 20 games, which team do you think would perform better, assuming similar schedules, injury luck, etc.

The reality is that there is a large degree of randomness in very close games. While the aggregate of blowouts tends to illustrate the underlying talent on the roster (or lack thereof). This phenomenon is true in nearly all sports: the record in 1-run games in baseball is considered to be largely a fluke and therefore not predictive going forward. Same with 3 point wins in football.

Of course, the hypothetical above is unlikely to ever happen. Just pointing out that studies have shown that point differential, while imperfect, is a better predictor than record alone in terms of predicting a team's future. If you consider 2 point games to be a 50/50 proposition, or even a 60/40 proposition, then it seems very unlikely that Team A would continue to go 0-whatever in their remaining 2 point games.

The other "of course" is that Team B could still win the above game, and the model would not be invalidated by that outcome.

EDIT: Back to the real world: The Celtics point differential standings will be cold comfort if the Celtics find themselves having to play 4 games in Philly or Indiana in the opening round.