2023-24 Celtics
- Thread starter 128
- Start date
Not sure how many here listen to Zach Lowe's pod, but he had the following mini-rant about the Celtics on last Friday's edition. Added the bold for a few pieces I found interesting:
"They're starting to, leak oil is a little strong, they're starting to wobble a little bit and have put together some performances that have made me be (concerned). They started the season 26-6 , they're now 11-6 since. When they were 26-6 , they averaged 121 points per 100 possessions , #2 in the league, 110 points allowed, that was #2 in the league. In the 11-6 stretch, they're down to 119/100 on offense, which doesn't sound like a big drop off, it's just two points, but offense trends up over the season, and we've seen this movie with Boston before, that's 9th in that span. 112 points allowed, that's 5th. Their net rating has dropped from 10.4 plus to plus 6.8, which is the difference between great and just good, they're like 5th in net rating in the 11-6 stretch. Still good, still like, everything's fine, I just haven't loved, just haven't loved it. Their starting, the version of their starting five with Horford in Porzingis's place is minus five in 215 minutes, it just hasn't played well and Porzingis feels just so essential to their team.
And I'm just gonna read you…what percentile thier share of shot attempts at the rim in their last 15-ish games since January 10th would rank on a zero to 100 scale: 5th, 25th, 7th, 9th, 13th, 8th,19th ,38th, 16th, 26th, 10th, 6th, 9th. I'm sorry to keep harping on this, it's not a crisis, but it's a thing. They shoot too many threes, they do not get to the rim enough, I know some people within their team would say: when we get to the rim and we miss, it's a jailbreak going to the other end, with threes we can set our defense. The math says this. The math says you're amazing, and you are amazing. I am not sitting here advocating that the Boston Celtics turn into like the 1994 Knicks. I'm just saying those numbers I just read are problematic, and they suggest that a 5-10% shift of your philosophy, a small but noticeable shift, is necessary. Because they had seven free throws last night. They don't get to the line a lot, they don't offensive rebound a lot, they are just too dependent on threes. Maybe this is just the malaise of we're awesome, we're good, we're gonna take the easier way out in some of these games, but those numbers, it's just ice cold blue up and down those numbers, and it's just something I'm monitoring. They're down to like 28th for the season, in percentage of shot attempts at the rim, it's too low, it needs to get higher."
A few thoughts:
-Starters minus KP + Al being negative in 200ish minutes is surprising, but maybe not shocking. I'd be intersted to look at the 3 point shooting splits, but a big part of the KP only big lineups dominating is his insane efficiency posting up, something Al doesn't bring at nearly the same level.
-I hadn't fully considered the 2nd bolded line about spacing/not crowding the rim helping transition defense, but it appears to be working. The Celtics rank first in transition Points Per Possession leaguewide. Cleary their offense into defense philosophy is helping so some degree. But I do agree that they need to get to the rim more, both for higher % shots and the free throws that rim pressure generates. Long playoff series, with games every other day played at insane levels of intensity, absolutely impact teams ability to shoot.
-KP is a great bellwether against bad shooting, but finding other ways to get to the rim more will help this team weather shooting variance storms, something everyone who has watched their last 2 playoff runs has been asking for. JB + Tatum are ok finishers (for elite players), and below-average handlers + decision makers when faced with multiple bodies in the paint. Just telling them to drive isn't the answer- we saw this get ugly for stretches during both the GSW and 23 MIA series. You can't just tell them to get to the hoop- leveraging spacing provided by a 5 out lineup is a great way too.
I know he's a topic of consternation here, but I thought Exasperated Guy did pretty well to point out some situations where BOS players could have cut towards the rim for very high percentage looks against the LAL and just didn't (0:42, 1:40, 2:24 marks)
View: https://www.youtube.com/watch?v=Qzz7Wc62WdM&t=228s
"They're starting to, leak oil is a little strong, they're starting to wobble a little bit and have put together some performances that have made me be (concerned). They started the season 26-6 , they're now 11-6 since. When they were 26-6 , they averaged 121 points per 100 possessions , #2 in the league, 110 points allowed, that was #2 in the league. In the 11-6 stretch, they're down to 119/100 on offense, which doesn't sound like a big drop off, it's just two points, but offense trends up over the season, and we've seen this movie with Boston before, that's 9th in that span. 112 points allowed, that's 5th. Their net rating has dropped from 10.4 plus to plus 6.8, which is the difference between great and just good, they're like 5th in net rating in the 11-6 stretch. Still good, still like, everything's fine, I just haven't loved, just haven't loved it. Their starting, the version of their starting five with Horford in Porzingis's place is minus five in 215 minutes, it just hasn't played well and Porzingis feels just so essential to their team.
And I'm just gonna read you…what percentile thier share of shot attempts at the rim in their last 15-ish games since January 10th would rank on a zero to 100 scale: 5th, 25th, 7th, 9th, 13th, 8th,19th ,38th, 16th, 26th, 10th, 6th, 9th. I'm sorry to keep harping on this, it's not a crisis, but it's a thing. They shoot too many threes, they do not get to the rim enough, I know some people within their team would say: when we get to the rim and we miss, it's a jailbreak going to the other end, with threes we can set our defense. The math says this. The math says you're amazing, and you are amazing. I am not sitting here advocating that the Boston Celtics turn into like the 1994 Knicks. I'm just saying those numbers I just read are problematic, and they suggest that a 5-10% shift of your philosophy, a small but noticeable shift, is necessary. Because they had seven free throws last night. They don't get to the line a lot, they don't offensive rebound a lot, they are just too dependent on threes. Maybe this is just the malaise of we're awesome, we're good, we're gonna take the easier way out in some of these games, but those numbers, it's just ice cold blue up and down those numbers, and it's just something I'm monitoring. They're down to like 28th for the season, in percentage of shot attempts at the rim, it's too low, it needs to get higher."
A few thoughts:
-Starters minus KP + Al being negative in 200ish minutes is surprising, but maybe not shocking. I'd be intersted to look at the 3 point shooting splits, but a big part of the KP only big lineups dominating is his insane efficiency posting up, something Al doesn't bring at nearly the same level.
-I hadn't fully considered the 2nd bolded line about spacing/not crowding the rim helping transition defense, but it appears to be working. The Celtics rank first in transition Points Per Possession leaguewide. Cleary their offense into defense philosophy is helping so some degree. But I do agree that they need to get to the rim more, both for higher % shots and the free throws that rim pressure generates. Long playoff series, with games every other day played at insane levels of intensity, absolutely impact teams ability to shoot.
-KP is a great bellwether against bad shooting, but finding other ways to get to the rim more will help this team weather shooting variance storms, something everyone who has watched their last 2 playoff runs has been asking for. JB + Tatum are ok finishers (for elite players), and below-average handlers + decision makers when faced with multiple bodies in the paint. Just telling them to drive isn't the answer- we saw this get ugly for stretches during both the GSW and 23 MIA series. You can't just tell them to get to the hoop- leveraging spacing provided by a 5 out lineup is a great way too.
I know he's a topic of consternation here, but I thought Exasperated Guy did pretty well to point out some situations where BOS players could have cut towards the rim for very high percentage looks against the LAL and just didn't (0:42, 1:40, 2:24 marks)
View: https://www.youtube.com/watch?v=Qzz7Wc62WdM&t=228s
Last edited:
This is picking nits. but +6.8 would be #3 in the NBA right now. The Thunder are at +8.0, the Clips are 3rd at +6.4.
Lowe's rant aside, the idea that 3s allow you to set the defense better than rim attempts hadn't occurred to me, but is hard to unsee. I imagine they also feel that corner crashes on 3-point attempts are less risky for that reason.
I'm sure it can be overdone, but that's the type of mildly outside-the-box thinking I like to see from a coaching staff.
And, as noted, it's shown up in the numbers. Transition attempts are incredibly valuable, and limiting that value is a big part of defense, regular season and playoffs.
I'm sure it can be overdone, but that's the type of mildly outside-the-box thinking I like to see from a coaching staff.
And, as noted, it's shown up in the numbers. Transition attempts are incredibly valuable, and limiting that value is a big part of defense, regular season and playoffs.
Specific to those lineups:
With KP:
Offensive: 2pFG%: 60.75%, 3pFG%: 39.42%
Defensive: 2pFG%: 49.89%, 3pFG%: 35.11%
With Al:
Offensive: 2pFG%: 55.56%, 3pFG%: 36.79%
Defensive: 2pFG%: 57.32%, 3pFG%: 35.68%
Not specific to those lineups:
Nothing massive pops out in the individual on-offs. KP seems to have more of an impact on SMR shots than Horford. Horford seems to be the better rebounder. The difference in rim shooting (make or frequency) isn't as large as I was expecting given those 2pFG% allowed numbers:
KP on: Opp Rim FG%: 61.2% KP off: 65.2%
Al on: Opp Rim FG%: 62.2% Al off: 64.3%
Here's the non-lineup specific on/off:
Yeah that Zach Lowe bit is like a living example of the xkcd "weighted random number generator" joke*. When they were going great, they were great, and now that they're not going so great, they're not as great. OK, do ya think this might be why we look at things using season averages, so that the extra-good parts and the not-quite-as-good parts get averaged down to a better sense of what we might expect in the future? Like, the predictive power of "having more data / no longer a small sample size" comes from minimizing the error in estimating the average. Choosing to turn around and subdivide the season so that you can still have small sample sizes and conveniently include or exclude portions of them, or play up the individual-game team-rankings (as if they're more meaningful to a human than just giving the average)... well, it doesn't reflect strong statistical understanding, I'll say that.
We attempt and make the most 3s in the league (per-game and per-100), although particularly so for straight-ahead 3s; for 20-24' shots (which include corner 3s and at-the-break 3s), we are league-average, 15th. Given the higher eFG value of a corner 3, this distribution would be surprising, in the absence of a competing explanation (such as, they're happier to take those because it lets them set the defense).
Is there a correlation between 3PA / game and Transition D PPP? Not really:
R^2 < 1%, basically noise, although the trendline is at least negative. Taking 10 more 3PA per game will tend to reduce your Transition D PPP by 0.01, or about 1 point per game. I suppose when you put it like that it sounds like that's not nothin', but from the correlation, clearly the impact of doing so is swamped by many other factors.
Does it at least limit the number of transition D possessions faced? Again, no, not really: very weak correlation, R^2 of 3.3%, and a positive slope, oddly. More 3PA = more time spent playing transition D, suggesting the opposite effect to what's postulated:
Take 10 more 3PA per game (32 -> 42), face ~2 more defensive transition possessions.
I'm not going to go into the shot-profile stuff to look at whether it's a function of how we're getting our 3Ps, like the 20-24' vs 25'+ numbers. Someone else could. But this seems like something that you're told and we come up with an intuitive explanation, but the reality is very different. The reality is probably just "having Jrue and White getting back limits opportunities and forces opponents to pull it out and reset for a half-court possession, and Tatum is always up for a chasedown attempt". I'm sure game-state affects how furiously teams try to turn rebounds into transition, as well, and god only knows how that would affect the conclusions here.
* I swear I didn't realize this was SteveF's avatar too, until after posting.
I don't mind the notion in theory, but does it hold up to scrutiny? Part of the effect we should see is that they are reducing the overall number of transition opportunities (by getting a set defense), and, well, they're 20th-fewest in possessions of transition defense (both overall and per-game).
We attempt and make the most 3s in the league (per-game and per-100), although particularly so for straight-ahead 3s; for 20-24' shots (which include corner 3s and at-the-break 3s), we are league-average, 15th. Given the higher eFG value of a corner 3, this distribution would be surprising, in the absence of a competing explanation (such as, they're happier to take those because it lets them set the defense).
Is there a correlation between 3PA / game and Transition D PPP? Not really:
R^2 < 1%, basically noise, although the trendline is at least negative. Taking 10 more 3PA per game will tend to reduce your Transition D PPP by 0.01, or about 1 point per game. I suppose when you put it like that it sounds like that's not nothin', but from the correlation, clearly the impact of doing so is swamped by many other factors.
Does it at least limit the number of transition D possessions faced? Again, no, not really: very weak correlation, R^2 of 3.3%, and a positive slope, oddly. More 3PA = more time spent playing transition D, suggesting the opposite effect to what's postulated:
Take 10 more 3PA per game (32 -> 42), face ~2 more defensive transition possessions.
I'm not going to go into the shot-profile stuff to look at whether it's a function of how we're getting our 3Ps, like the 20-24' vs 25'+ numbers. Someone else could. But this seems like something that you're told and we come up with an intuitive explanation, but the reality is very different. The reality is probably just "having Jrue and White getting back limits opportunities and forces opponents to pull it out and reset for a half-court possession, and Tatum is always up for a chasedown attempt". I'm sure game-state affects how furiously teams try to turn rebounds into transition, as well, and god only knows how that would affect the conclusions here.
* I swear I didn't realize this was SteveF's avatar too, until after posting.
I have zero proof, but have always felt that Corner3 attempts help with transition Defense.
Having a defender try to block a Corner3 has them heading into the front row (farthest point away from their basket),
while a defender challenging an above-the-break 3 has them heading toward their basket.
View: https://www.youtube.com/watch?v=uAumBjFwtGU
There's got to be some tracking data that has this - I'd love to know the percentage of layup attempts Tatum and Brown fall down on versus other players that drive a lot. It seems like they fall down all. The. Time. Which results in a great shot going the other way.
But maybe every other guy in the NBA does, too, and I just clock it the way I do with those two.
Missed rim attempts do likely turn into transition more often than missed 3 pointers, but teams miss rim attempts less often, and misses (of all kinds of shots) turn into transition attempts more often than makes. Plus, rim attempts result in more offensive rebounds.
I don't doubt that there's an advantage when comparing missed shots, but taken as a whole, I'm not surprised the correlation is weak/nonexistent given make/miss and offensive rebounding.
Edit: Another issue here is "transition" is sort of a blunt category. A better but imperfect way to look at it is when a shot occurs in a fresh 24 shot clock (nba.com's shot clock stats lack sufficient granularity). And even there I think you'd have to be careful since I doubt there's a purely linear relationship between EFG% and time left on the 24 shot clock, though that information is probably out there somewhere.
Last edited:
I'm not sure if they define transition for these stats the same way that the NBA stats define "fast break", but if so it has to be off a missed shot or live ball turnover to begin with anyway. They define fast break as a shot or shooting foul within 8 seconds of a live ball change of possession in the other half of the court.
They do. It’s like a game of ring around the rosey out there sometimes. They all fall down.
My impression is that falling to the court in case you get a call has become something players not only do, but train themselves to do -- or should be, at least. There's certainly a technique to it. Done well, it's handled with a minimum possibility of genuine injury, and with the ability (one hopes) to spring quickly back into action when that needs to happen.
The theatrical aspect of it can get ridiculous. My read is, shrewd players do it dramatically enough to ping the refs when warranted, but otherwise get right back up and into the game again -- a split second decision.
Jaylen Brown, IMO does it pretty skillfully, smoothly.
I grew up watching a very different, much rougher version of the game. But I've been adjusting. The falling-just-in-case thing is not necessarily displeasing to watch. Just another element of one's game, layered in
So I guess that means misses turn into transition attempts way, way more often than makes!
Infinitely more so! Thanks for the info.
In todays game this is a major problem in a teams transistion defense. If you don’t convert you are going to be giving up numbers the other way which is nearly always going to at worst result in a great look. The 3-pt shot with corners crashing to where their sight lines are clear with the ability to quickly retreat on the miss is just so damn efficient.
Obligatory:
(I don't have a problem with it. Only way I'd be tuning in to the dunk contest).
Yeah.. the contest feels like it’s more for young guys or lesser knowns rather than players like Brown. The days of Jordan or Dominique being in this contest are long gone.
With his injury history don’t really want him to jump over people or cars or anything where he can come down awkwardly.
So I guess that means misses turn into transition attempts way, way more often than makes!
Chose the right spots (those split second decisions...), and you employ the technique without necessarily wrecking your transition D. It's a proportion thing.
edit (expanding a bit after some more thought)
Let me add I definitely take the points quoted above. It's just, the right degree of working the refs, depending on the night, sometimes might make sense strategically, if you can pick your spots.
Last edited:
I feel like this stat is missing context because as we all know, the Cs have been really good in the post. I don't have time to do a deep dive on this, but it feels like this stretch has both the huge blowouts of the Cs (MIL and LAC) as well as other games that were blowouts on the Cs side (first HOU game, MIA, SAS, off the top of my head). And that stretch also (I believe) includes the MIN and DEN games.
Plus guys have been in and out of the lineup.
Plus, I'd be interested in knowing just what counts as "shots at the rim" - if they including only shots in the restricted area and not shots in the paint, well I think shots in the paint would probably be a good indicator since that's where KP usually shoots.
Maybe for the Cs, instead of 3s and layups being the optimal offense, it's 3s and post-ups.
There is tracking data on this. If you're interested, I ran across this 2020 article - https://fansided.com/2020/06/08/nylon-calculus-crash-the-glass-or-get-back-on-defense/ - that compared how LAL and OKC defended transition using tracking data. I haven't read the article yet but it looks interesting.
I believe that if Jayson wants to move to the next level NBA player he needs to start rocking the baby ala the Doctor. I think he‘d be great at it and has about an opportunity or two per game to do it.
edit: I know Jaylen is being discussed but JT is the one who should be “rocking the baby”!
View: https://youtu.be/SrlBgLdGKZM
He does seem more like a power dunker than a high flyer. This would've made more sense 5-6 years ago, if he participates JB would be the biggest star in the dunk contest since Dwight Howard in 2009. There has been plenty of big names since (Blake Griffin, Paul George, Giannis, Donovan Mitchell), but none who had made an all-NBA team or three all-star games (also excluding Dwight Howard in 2020, way past his prime).
Sometimes I think about how Tacko Fall was basically bullied into being a prop during a dunk contest (it wasn't planned, and Tacko has said he was uncomfortable) and how shitty that is. Then I imagine Jaylen Brown recreating it with various EC playoff rivals (Giannis, Bam, an Embiid standee), but instead of jumping over them he just dunks on them and stares, and I smile.
It's possible they're pulling from a different subset, I just know based on the nba rulebook what they consider "fast break" to be and where those stats come from, I'm not sure where the transition stats are coming from. It's pretty likely they're coming from people who don't necessarily know what the NBA wants different shots and play types to be considered to be because the NBA itself never really explains it well enough anywhere, so I wouldn't trust them to be exact, though they're probably pretty close.So I guess that means misses turn into transition attempts way, way more often than makes!
Surprised, the Celtics 2-for-1's have been better than I thought.
View: https://twitter.com/SamLaFranceNBA/status/1754913455985832081
I went back and watched all of the #Celtics' two-for-one possessions. Here's what I found:
https://www.celticsblog.com/2024/2/6/24057052/boston-celtics-2-for-1-breakdown-jayson-tatum-brian-scalabrine-joe-mazzulla-jaylen-brown-porzingis
View: https://twitter.com/SamLaFranceNBA/status/1754913455985832081
I went back and watched all of the #Celtics' two-for-one possessions. Here's what I found:
https://www.celticsblog.com/2024/2/6/24057052/boston-celtics-2-for-1-breakdown-jayson-tatum-brian-scalabrine-joe-mazzulla-jaylen-brown-porzingis
Nice work on his part to actually go and collect all of this data. Not totally surprised by this:
"The strange thing here is that more of Boston’s final attempts have been more open than the ones that they’d taken about 30 seconds prior"
Feel like when they (usually Tatum) come down on the 1st of the 2 possessions there's rarely any movement and it's just a no-pass jumper, something the author spoke to in the article. At the end of the quarter they seem to run actual actions more. As a fan of the 2 for 1 I appreciate someone putting numbers behind what we've been watching!
"The strange thing here is that more of Boston’s final attempts have been more open than the ones that they’d taken about 30 seconds prior"
Feel like when they (usually Tatum) come down on the 1st of the 2 possessions there's rarely any movement and it's just a no-pass jumper, something the author spoke to in the article. At the end of the quarter they seem to run actual actions more. As a fan of the 2 for 1 I appreciate someone putting numbers behind what we've been watching!
The 1st possession of a 2-for-1 requires the shot to be taken within a certain timeframe to effectively take advantage of the 2-for-1. A Tatum Iso controls when the shot is taken whereas running motion does not.
I know you know this but not all 3s are created equal. Other than a dunk, you can't get much better than Jrue or Hauser for a Corner3
Tatum is excellent with the clock, ISO PU3s are very worth it on the front end of a 2 for 1
It's not complicated math. Two bad shots are better than one good shot. The bad shots would have to be sub 27% efg% for that to not be true, and really even less than that since bad shots probably entail less turnover risk. There really aren't shots like that inside the logo.
I'm surprised anyone still needs to be convinced.
I'm surprised anyone still needs to be convinced.
Yeah aesthetically it can be very displeasing, but the math doesn’t lie. And yet I still complain about it when it doesn’t work anyway.
Mazzulla is a different kind of cat
View: https://twitter.com/NoaDalzellNBA/status/1754953984605925589
View: https://twitter.com/NoaDalzellNBA/status/1754953984605925589
yeah, and the math is pretty overwhelming - this isn't one of those cases where the math might change depending on the circumstances. glad to see a skeptic go and actually test it out and see the math don't lie.
Weird finish to the game: they ran very good crunchtime offense from about 9 minutes to 1:30 left in the 4th, but the defense and rebounding were uncharacteristically bad for late-game.
Jaylen allowed a couple layup attempts/fouls, no rebounding effort on Bogie missed 3, and they didn't set transition defense a couple times in standard situations.
That kept Atlanta in the game, and so then when the offense bogged down for 2 possessions, they were right there when Bey missed the late 3.
Also felt like they could have been more intentional about attacking Trae in the last couple minutes.
Jaylen allowed a couple layup attempts/fouls, no rebounding effort on Bogie missed 3, and they didn't set transition defense a couple times in standard situations.
That kept Atlanta in the game, and so then when the offense bogged down for 2 possessions, they were right there when Bey missed the late 3.
Also felt like they could have been more intentional about attacking Trae in the last couple minutes.
Although, to play devil's advocate to my own "they should have attracted Trae more" point, they've done a better job this year in crunch time of not predictably getting bogged down hunting mismatches. The looks were consistently high-quality until the end.
I thought they used a little too much clock. It’s great to drain it, but they could have easily started their movement with 10-12 seconds on the shot clock instead of 5-6 and not gotten pinned into being unable to pass it. If they are getting a quality look with 5 seconds on the shot clock instead of forced look with 1 second they are likely better off.
I thought so too, but if you rewatch the 4th quarter, it's only under 2 minutes that they really did this. Prior to that, they moved it well and got good shots.
They stopped trying on D with ~3-4 minutes left, and stopped trying on O with ~2 minutes left.
Yup, that’s right. Good distinction. Either way, still a comfortable win, but a good reminder how quickly 10+ leads can disappear if you stop trying.
I always remind myself that an 8-0 run in a minute is just:
- opponent hits an open 3
- you miss a good look
- they hit a contested 3
- you get frustrated and turn it over, they get a dunk the other way
It happens so fast, and barely requires a lapse (just very minor bad luck/focus).
This is also why I laugh at the idea that crunchtime is just 5 point games with 2 minutes left or whatever.
It's funny, at the Lakers game, I had some folks sitting around me screaming all game long about how the C's shooters don't "follow their shots" as if they were watching a middle school game. I think one of them said multiple times "we were taught to follow our shots when we were little kids." I was so disgusted by the effort on the court that I didn't bother to turn to them and say "and if Tatum crashes from the top of the key to follow his shot, you'll be bitching your head off that nobody got back on defense."
They are now 22-10 against teams over 500. That winning percentage would have them as ... the #1 seed in the east. Ha!
20-10 gets them to 60 wins. It takes a contender to win 60. The Cs won 62 or more from 1980-86. The one year they won 56 was probably the weakest team. Though they knocked on the door after 86 they were never rerallu close to a title again. So thios group getting to 60 would be the best way, all more detailed analysis aside to show they are the best or ready to take a run.
I have more substantative comments for later, but for now it's funny that their biggest issue the past 2 games has been Prevent Offense with 2 mins left, up big.
I would guess they are relatively confident in their ability to fix that, and I'd mostly chalk it up to pre-ASB boredom.
I would guess they are relatively confident in their ability to fix that, and I'd mostly chalk it up to pre-ASB boredom.