On Assets and Liabilities: Celtics Playoff Bench vs their Peers

Koufax · Aug 27, 2020

That is a commendable effort and really informative. Thank you for putting it together. It must have sucked to finish that up just as the prospect of seeing the playoffs blow up entirely emerged.

Mugsy's Jock · Aug 27, 2020

Thanks, @InstaFace -- I'm not nearly as facile with basketball analytics as I want to be, but that was super straightforward, well-written and informative. I feel a little more smart. Also a little more concerned.

Sille Skrub · Aug 27, 2020

Wow, nice post and thread! Thanks!

RedOctober3829 · Aug 27, 2020

This thread is why this draft is important for the Celtics despite everybody saying the picks they have don't mean much. They have a chance to pick players at 14, 26, and 30 that can better compliment their starters coming off the bench.

radsoxfan · Aug 27, 2020

RedOctober3829 said:
This thread is why this draft is important for the Celtics despite everybody saying the picks they have don't mean much. They have a chance to pick players at 14, 26, and 30 that can better compliment their starters coming off the bench.

I definitely agree, though they had almost the exact same chance last year and now are stuck with a terrible bench because they are forced to play some of these guys.

14:Langford
22:Williams
33:Edwards

The jury is still out on them, I don't mean to imply they are all busted already, and I agree with your overall point. The draft is important, but unfortunately it's hard to find good players outside the lottery to step in right away to be part of the rotation.

And I'll echo everyone else, thanks for the opening post, strong work!

Saints Rest · Aug 27, 2020

InstaFace said:
So in the Hayward thread, radsoxfan and I had a brief exchange on whether our end-of-rotation players were, collectively, a liability. This is a tough question to answer, because any deep-bencher is going to look bad by comparison to NBA starters. Let's compare.

My impression is that the answer is no, but I'm certainly happy to be corrected by those more knowledgeable here. TLDR: Turns out, answer is actually "a little yes".

So, Hayward was our #4 non-big, arguably #5. With BW as the clear #6 non-big, Hayward's absence presses our 7-9 on that list into service: Semi, Grant and Langford.

Here are the #7-9 non-bigs (so excluding anyone with >50% minutes at Center) on the other top contenders, with VORP numbers attached, SSS caveat notwithstanding. This should be a decent comparison of "deep bench" (just outside playoff rotation or fringe playoff rotation), and adequacy relative to who they'll be taking the floor against. N.B. this has nothing to say about particular matchups, fitness for a particular purpose, health, etc.

Format:
- Position Name, Mins/G (VORP, PER).

All numbers are regular season. Remember PER has 15 as a league average.

Milwaukee: 6 players with >10 starts, exclude R. Lopez & B. Lopez, then George Hill (backup PG) as #6.

7: SG/SF Pat Connaughton, 18.6 (0.8, 11.5)
8: SF Kyle Korver, 16.6 (0.5, 11.3)
9: PF Ersan Ilyasova, 15.7 (0.6, 14.9)

A very veteran squad for Deep Bench.

Toronto: 8 players with >10 starts, exclude Ibaka, Marc Gasol & Chris Boucher, then SG Norman Powell #5, SF Patrick McCaw #6.

7: SF/PF Rondae Hollis-Jefferson, 18.7 (0.6, 15.1)
8: SG/SF Terence Davis, 16.8 (0.8, 13.8)
9: PG/SG Matt Thomas, 10.7 (0.3, 13.3) N.B. Rookie

Two 3/4 year vets with ~6 FGA/gm, and a rookie who got 440 minutes (compare Grant Williams: 1043, Romeo Langford: 370).

Miami: 6 players with >10 starts, exclude Olynyk and Meyers Leonard (injured as of February); also exclude Justice Winslow (injured all season), and Bam Adebayo (57% PF, 43% C regular season, but 90% C in playoffs). That puts their non-big lineup at Butler, Dragic, Herro, Robinson, Nunn (starter all season, but some personal incidents and maybe an injury or illness lately), and Jae Crowder at #6. Crowder has lately started in Nunn's place, but might get bumped back down if Nunn is truly back for the Milwaukee series.

7: SG Andre Iguodala, 19.9 (0.2, 10.5)
8: SF/SG Derrick Jones, 23.3 (1.0, 14.9)
9: PF Chris Silva, 7.9 (0.0, 17.3)

Despite having way more regular-season minutes, Derrick Jones has been relegated to spot duty in the playoffs while Crowder (4 starts, 25.5 mpg, might revert to bench) and Iguodala (26.8 mpg, mostly there for D) have seen roles much greater than their regular-season contributions. Part of that is Iguodala's injury that only let him begin his season in February; Crowder was traded to Miami at the deadline but has been at near-starter minutes all season. The biggest difference is that in the Pacers series, Miami has largely thrown aside Derrick Jones, and hardly played Olynyk at all.

Lakers: 6 players with >10 starts, exclude Dwight Howard, JaVale McGee, count AD as a PF. You have SG Avery Bradley as #6.

7: PG Alex Caruso, 18.4 (0.5, 10.8)
8: PG Rajon Rondo, 20.5 (0.2, 12.4) N.B. opted out of bubble
8: PG Quinn Cook, 11.5 (-0.2, 10.0)
9: SF/SG Troy Daniels, 11.1 (0.0, 9.0)

Caruso is in a role that's roughly as big as Hayward, and from there it's a bunch of dreck they don't trust. The Lakers have 2 players at 35mpg, 8 who are between 18 and 25.5 mpg, and then not a lot of true deep-bench minutes.

Clippers: 8 players with >10 starts, exclude Harrell and Zubac, but we gotta add in Lou Will because he leads the team in minutes despite only 8 starts. Regardless of starts, I'd put their non-bigs ordered as Kawhi, George, Morris, Beverly, Lou Will, Shamet. Patrick Patterson began the year as a starter (18 starts) but gradually fell out, and now after the Marcus Morris trade gets only very occasional minutes. With Beverley's injury, Reggie Jackson has temporarily entered the playoff rotation as backup PG behind Shamet.

7: PG Reggie Jackson, 21.3 (0.2, 12.6)
8: PF JaMychal Green, 20.7 (0.5, 11.1)
9: SG/SF Rodney McGruder, 15.6 (-0.2, 6.7)

With Beverley returning, we'll see less of Jackson. Green has gotten proper playoff rotation minutes (12-18 per game), but McGruder only garbage time. This might be the most direct comparison we have for the loss of Hayward, though, because they put in Jackson and so we got to see what their rotation would shift to do.

Denver: 8 players with >10 starts, exclude Jokic and Plumlee, and let's name the top 6 others as Murray, Grant, Porter, Monte Morris, Millsap, and Torrey Craig in place of the opted-out Will Barton and Gary Harris. They also traded two deep benchers to Minnesota at the deadline (Malik Beasley & Juan Hernangomez). Right now they're rolling with:

5: PF Michael Porter Jr., 16.4 (0.8, 19.8)
6: SF Torrey Craig, 18.5 (-0.1, 10.8)
7: PG PJ Dozier, 14.2 (-0.2, 11.6)
8: SG Troy Daniels, 12.7 (-0.1, 5.3 lol)
9: SF Keita Bates-Diop, 14.0 (-0.1, 12.1)

Of those, only Dozier really can be considered "in the playoff rotation" right now, the others (new acquisitions from Minnesota) have 19 and 16 playoff minutes in 5 games, including 2-3 DNPs. In a sense, Denver has already had to promote two deep-benchers to the playoff rotation (Porter and Craig) after they lost their MPG leader (Barton) and preferred 6th-man (Harris). That's why I put Porter and Craig's numbers above, because they're sort of the Boston comps right now.

---

Boston: 6 players with >10 starts, exclude Theis, Kanter & Time Lord, then PG Wanamaker as #6, counting Hayward as a shadow #5.

7: SF/PF Semi Ojeleye, 15.1 (0.1, 7.7)
8: PF Grant Williams, 14.7 (-0.2, 7.9)
9: SG Romeo Langford, 11.6 (-0.3, 5.4 oof)

Can also compare Javonte Green, who actually got more total NBA minutes this year than Langford, though he's now out. Either way, our #7-9 non-bigs are rated pretty low-value by both VORP and especially by PER, at least compared to our peer teams. And while many teams have given (non-blowout) playoff minutes to their #7s on these lists, few have had to give meaningful minutes to #s 8 or 9. Playoff minutes for our 3 boys here: 29, 39, and 37 respectively.

---

Clearly better than Boston (at readiness to absorb injuries to playoff rotation), on this list:
- Milwaukee
- Toronto
- Clippers
- Miami

Comparable to Boston:
- Denver
- Lakers

...and Denver may be about to get unceremoniously bounced tomorrow night, too. So while I had thought our depth was pretty good, I'm forced to conclude that once you get beyond Wanamaker and have to look for a Hayward replacement, we really aren't that good, at least aside from our 4-headed Center rotation.

Get well soon, Gordon Hayward, and maybe eat some spicy food and some castor oil, Robyn Hayward.

This is a terrific post.

A couple things that I wonder about, that maybe you have data on from your original research:

How many MPG do each team's excluded bigs play? Theoretically, it would be close to 48, but some teams may feel more comfortable playing two bigs together.
How many MPG do the top 6 non-bigs play?
Then relatedly, how many MPG are going to #7, #8, #9, both individually and collectively?

My point being, how big a factor is a deep bench during the playoffs, when rotation conceivably shorten. It's a bit hard to tell how much they shorten, but I think we would all agree that they do.

My sense is that the answer to my final bullet, is something around 10-15 MPG, collectively.

My other sense is that those players 7-9 per your definition, are almost be definition "role" players, who would be coming in not as a replacement for one of a team's top 6, but rather to fill a specific, match-up and/or context-specific role. For example, Semi is not a replacement for anyone on the top 6 but against Giannis, he has a specific value, and one that can't be replicated by Romeo, for instance.

Finally, I believe that a smart coach (which I think Stevens is) will maximize the talents of those deep bench guys more so when you have a player who is more of a Master on One vs a Jack of All.

Smokey Joe · Aug 27, 2020

RedOctober3829 said:
This thread is why this draft is important for the Celtics despite everybody saying the picks they have don't mean much. They have a chance to pick players at 14, 26, and 30 that can better compliment their starters coming off the bench.

I would agree with this except our window is not 2-3 years from now, the window to compete is now. We will not be drafting for contributors this draft, we are drafting for trade fodder and the picks themselves are trade fodder. With the clock ticking on Hayward and Walker, we need that bench next year.

wade boggs chicken dinner · Aug 27, 2020

Thanks for doing this. A couple of things.

(1) Instead of MPG, I think you should minutes played. For example, you have Romeo at 11.6 mpg. However, he only played 320 minutes this year. Compare to Terence Davis, who you have at 16.8 mpg - but he played 1209 minutes, or almost 4x what Romeo played. Which brings me to (2)

(2) Aren't WS and VORP counting stats? Sorry I am not ~~a facile~~ as skilled with these stats as I would like to be either. But if they are, that would explain Romeo's poor counting stat numbers. Also, other teams rely way more on their benches than BOS does. For example, BOS's bench would never score 100 points, even if you gave them two games to do it. They just don't get the minutes.

Finally, as far as the overall evaluation of the various benches, I guess I don't see a lot of daylight between what BOS has and what other teams have - except, of course, that DA was never going to sign anyone like AI, Jae, or Reggie Jackson because BOS needs to develop players on rookie deals, not try to add a one-year player for a title run. Just as one example, maybe Connaughton is a better overall basketball player than Semi but I think BOS needs what Semi provides - ability to switch; ability to guard specific players like Giannis; and he's shooting 38% from 3P this year - way more than what Connaughton might provide. YMMV.

As for Romeo, I'm a big fan of his and think he's going to have a long career in the NBA. I posted this in the playoff thread but I'll post again here: a deep dive into his defense, particularly his on-ball defense, which may be as good as anyone on the Cs.

JakeRae · Aug 27, 2020

I think the non-big formulation makes little sense because different teams run different rotations. At the end of the day, the Celtics are a bit hamstrung on depth because they only use one big and 3 of their 8 players who contribute positively are bigs. I’d also add that cutting Wanamaker out of the list of guys compared masks just how terrible our bench is because he is a big part of why it is bad. Finally, I would avoid PER as a metric as a general rule and would suggest looking to impact metrics like those compiled by Basketball Index.

I’m on my phone so I’m going to do a pretty simple comparison of depth by looking at the point where a team shifts from having value positive players to negative. To do this, I’m using Basketball Index’s Simple Average Impact, which I believe is an average of PIPM, RPM, RAPTOR, and BPM. These stats are all pre bubble and I will only look at the top 4 teams in the East.

Celtics: the top 8 players are all above average. Jaylen Brown has the lowest rating at 0.88. Robert Williams is in that grouping and may deserve SSS caveats. After that, things drop off severely. Green, Grant, and Semi are all between -1.5 and -2, Wanamaker is a hair below -2, and Romeo is around -2.75.

Raptors: They also have 8 guys who cut off around where the Celtics do (Rondae is at 0.89), with that grouping including the likely suspects and Boucher. The difference is that the Raptors then have two more guys who are positive impact players in Anunoby and Serge (also Matt Thomas who had sufficiently limited minutes I’m not counting him). McCaw is the only player the Raptors give real minutes to who is a big negative.

Bucks: They only run 8 deep with positive rates players and their 7 and 8 guys are closer to 0.5 than 1. Korver is barely negative and Connaughton is above -1, so again the depth still runs deeper with quality players. Robin Lopez is awful and plays and Sterling Brown is also around -1.

Heat: The Heat only have five positive players, but they have another 7 guys who are better than the 9th best Celtic. Herro is the guy they play who has a rating below our bench guys.

I don’t think these metrics are a perfect analytical lens, but the clear point is that the Celtics have an unusual cliff like drop off from their core players to their bench.The Bucks and Raptors are both significantly deeper, going 10 or 11 deep with quality or almost quality NBA players. The Heat are much shallower in terms of good players but they have a large collection of slightly below average players too.

chilidawg · Aug 27, 2020

InstaFace said:
So in Romeo's case, -0.3 VORP in short minutes AND a PER of only 5.4, in combination, suggests that he sucked out loud on the court this year and if he's seeing non-blowout playoff minutes it's because either he's had a total Eureka, or more likely, we're desperate and he's the best body available to throw at the problem.

Those stats would indicate that he sucked offensively, which would be true. They tell you very little about his defensive contributions, which seem to me to be significant.

ManicCompression · Aug 27, 2020

Awesome post - I think one thing that works in the Celtics favor is the depth of our starting lineup with playmakers (particularly when Hayward's healthy) vs. other teams.

The Celtics can always play at least two of the following:
Kemba
Smart
Jaylen
JT
Hayward (please come back)

Though Smart is not a traditional scorer, he's an excellent passer, and can keep the offense cromulent when some of the other guys are off the floor. I'm not including Wanamaker even though I don't think he's as bad as he's made out to be.

Compared with the other top-tier contenders:
- LA - when Lebron's off the floor, it's been well documented that they've been terrible and can't create offense
- Milwaukee - they really need to have one of Giannis or Middleton on the floor in the playoffs IMO - their bench numbers are good, but George Hill is more of an off-ball shooter and Bledsoe has his struggles in half-court offense. Their wings are shooters and not creators. I think we'll see how much they miss Brogdon when they're playing Non-Orlando Magic quality teams
- LAC - They always need to have Kawhi, PG or Lou Will on the floor, and preferably two of them (though Kawhi could play alone and be effective) - Morris can get hot, but he's not really someone you want to run an offense through. Neither is Beverley.
- Toronto - this is the team that can compete with the Celtics in terms of starting depth with FVV, Siakam, Lowry and even Norm Powell in a pinch (though the high-end quality may not be the same as other teams). That said, they have a tough time scoring in the half court even when they're all healthy.

No one has the same kind of scoring depth that we have, particularly when it comes to off-the-dribble jumpers (which annihilated Philly's drop coverage and could do the same to Mil). They may have higher top-end talent (Tatum obviously is not in the Lebron/Davis/Kawhi/Giannis stratosphere yet) but none can deploy multiple offensive hinges the same way we can.

So while the Celtics have worse depth, I think that issue is mitigated because those players don't really have to do much because they're always playing with multiple playmaking ballhanders. Semi, Romeo, Grant, etc. still get to play in their very defined, limited responsibility roles - defend well and shoot open corner threes (Semi above the break gives me hives). I think when we look at bench PPG after games, it's a little unfair because our starters will usually all hit double digits in points - there aren't a lot of shots left for 7-9 (and for good reason). These guys don't contribute much by design, whereas the load for a team like Milwaukee is much more stacked towards relying on bench contributors. If the lineup is ever Semi, Romeo, Grant, Kanter, and Wanamaker, we're either up by 30 or something has gone horribly wrong.

JCizzle · Aug 27, 2020

radsoxfan said:
I definitely agree, though they had almost the exact same chance last year and now are stuck with a terrible bench because they are forced to play some of these guys.

14:Langford
22:Williams
33:Edwards

The jury is still out on them, I don't mean to imply they are all busted already, and I agree with your overall point. The draft is important, but unfortunately it's hard to find good players outside the lottery to step in right away to be part of the rotation.

And I'll echo everyone else, thanks for the opening post, strong work!

Those picks tell me that we need to focus on finding a shooter in the draft. Romeo and Grant seem to project as good defenders with questionable shots. Edwards might not even be on the team next year and didn’t show much ability to shoot against NBA quality players. Seeing all the shooters that the Heat can run out makes me jealous in an offensive driven league.

—

Also wanted to echo thanks for the topic, it was an informative read.

Auger34 · Aug 27, 2020

RedOctober3829 said:
This thread is why this draft is important for the Celtics despite everybody saying the picks they have don't mean much. They have a chance to pick players at 14, 26, and 30 that can better compliment their starters coming off the bench.

I don’t know, it’s very hard for most rookies to be sound enough defensively and mentally for their coaches to trust them on the floor in the playoffs. Especially for someone like Stevens who tends to not have a very long leash with rookies.
In terms of next year’s team I think the development of Langford and Grant is more important than any draft pick honestly

JakeRae · Aug 27, 2020

tbb345 said:
I don’t know, it’s very hard for most rookies to be sound enough defensively and mentally for their coaches to trust them on the floor in the playoffs. Especially for someone like Stevens who tends to not have a very long leash with rookies.
In terms of next year’s team I think the development of Langford and Grant is more important than any draft pick honestly

Agreed. This years draft picks are about helping the team 2+ years from now. It is a rare rookie that isn’t a net negative on the court. Romeo and Grant Williams have both shown NBA roleplayer potential as rookies. One or both could be averagish players next year and positives by year 3. Doing well in this draft will potentially set us up to have Toronto level depth in 2 years.

tmracht · Aug 27, 2020

Just dropping in to say bravo this was a hell of a read during my lunch break. Thanks!

wade boggs chicken dinner · Aug 27, 2020

InstaFace said:
So in Romeo's case, -0.3 VORP in short minutes AND a PER of only 5.4, in combination, suggests that he sucked out loud on the court this year and if he's seeing non-blowout playoff minutes it's because either he's had a total Eureka, or more likely, we're desperate and he's the best body available to throw at the problem.

Or, option C - IIRC, Romeo missed a good part of the beginning of the season rehabbing from his injury; realized that playing time is predicated on defense because there aren't many shots to go around, and has been really quite good on that end.

Again IIRC, Romeo was first wing off the bench in at least a couple of games. If this is correct, Brad's not trying to steal minutes with Romeo, he thinks he's going to get something positive out of him. Note that Romeo had a net rating of 15.6 v PHI (114.3 ORtgv 98.7 DRtg).

And I think Romeo matches up well v TOR.

Jimbodandy · Aug 27, 2020

This is a great thread. Thanks IF for kicking it off so well and for everyone's contributions.

I think that the Celtics approach for bench guys this year is interesting for two different reasons.

1. It is clearly a bridge year for role players here. Romeo & Grant are projects (high upside IMO, but YMMV). Edwards and Waters are even bigger projects (although I think that Waters is probably NBA PG3 already, it made sense to let him work kinks out in Portland). The Cs kept BW and SO (low cash) and brought in EK (low enough cash) to fill slots that their megadraft 2019 guys weren't seasoned enough for yet. If Grant, Tremont, et al. can supplant one of the placeholder guys next year, great. But clearly DA is taking many bites at the apple for role guys and will continue to do so. So it's understandable that there aren't guys "in the high minors" so to speak, because they're all on the big club now (JT/JB/DT), but the 2019 guys are the AA guys who might take a job next year, if you'll forgive my hamfisted analogy. And the 2020 guys will be the Short-A ball guys next season, etc. It's a bridge year 7-13.

2. Because the Celtics are loaded with shot generators and smart offensive guys (5 of the top 6), DA can and did get away with having defensive minded guys at 7-13. Skipping past the fact that JT/JB/MS are no defensive slouches themselves, it's entirely undertstandable that GW/TW/SO/RW/RL are defensive guys with no current or short-term offensive upside. If Lou Will is your starting PG, you want Pat Bev coming off the bench. If Pat Bev is your starting PG, you want Lou Will coming off the bench. In our case, we have five plus offensive players (or facilitators in MS case) in the top 6. Not having bench scoring really isn't that relevant. Stagger the minutes, and you have access to both offense and defense at any time. Sometimes Brad goes Tatum plus 4 octopus bricklayers, and our lead goes up because the other team can't do jack.

Apologies if some of this is redundant to above posts. And I do realize that even with this team construction, losing Hayward is introducing guys who are holes at one end of the floor and getting them more minutes. However, what I would say is that what's unusual about our team is that our offensive depth is basically isolated to the first group, and our bench group is largely defensive specialists (who hopefully have more upside someday). This is unusual team construction to not have bench scoring, but it's not inherently problematic.

p.s. there are no good stats for properly measuring hoop defense IMO. I encourage the enterprise and everyone's efforts to continue searching for one and building variations, but the best that we have currently are directional, not conclusive. And box score based ones (steals/blocks) just doesn't get us close to measuring guys' defensive contributions.

Saints Rest · Aug 27, 2020

Some of this thread, especially those posts that note the depth of the Celtics top-6, reminds me of those times when a game will be very close or tied and the TV producers will throw up a stat like “Bench Scoring: Team A - 36; Team B - 12.”
I don’t care if the Celts 7-9 guys only contribute 5 PPG when their top 3 are averaging 76 PPG.

oumbi · Aug 27, 2020

wade boggs chicken dinner said:
Thanks for doing this. A couple of things.

(1) Instead of MPG, I think you should minutes played. For example, you have Romeo at 11.6 mpg. However, he only played 320 minutes this year. Compare to Terence Davis, who you have at 16.8 mpg - but he played 1209 minutes, or almost 4x what Romeo played. Which brings me to (2)

(2) Aren't WS and VORP counting stats? Sorry I am not a facile with these stats as I would like to be either. But if they are, that would explain Romeo's poor counting stat numbers. Also, other teams rely way more on their benches than BOS does. For example, BOS's bench would never score 100 points, even if you gave them two games to do it. They just don't get the minutes.

Finally, as far as the overall evaluation of the various benches, I guess I don't see a lot of daylight between what BOS has and what other teams have - except, of course, that DA was never going to sign anyone like AI, Jae, or Reggie Jackson because BOS needs to develop players on rookie deals, not try to add a one-year player for a title run. Just as one example, maybe Connaughton is a better overall basketball player than Semi but I think BOS needs what Semi provides - ability to switch; ability to guard specific players like Giannis; and he's shooting 38% from 3P this year - way more than what Connaughton might provide. YMMV.

As for Romeo, I'm a big fan of his and think he's going to have a long career in the NBA. I posted this in the playoff thread but I'll post again here: a deep dive into his defense, particularly his on-ball defense, which may be as good as anyone on the Cs.

Perhaps I am missing something, but the use of the word "facile" here is not what I would think it would be.

I learned the word to mean: appearing neat and comprehensive only by ignoring the true complexities of an issue; superficial. So, wishing to be MORE facile would not seem to be a desirable goal. But again, maybe language has passed me by on this one.

wade boggs chicken dinner · Aug 27, 2020

oumbi said:
Perhaps I am missing something, but the use of the word "facile" here is not what I would think it would be.

I learned the word to mean: appearing neat and comprehensive only by ignoring the true complexities of an issue; superficial. So, wishing to be MORE facile would not seem to be a desirable goal. But again, maybe language has passed me by on this one.

It's a pandemic, we can make words mean whatever you want.

I'll try to think up a better word and edit it later.

chilidawg · Aug 27, 2020

InstaFace said:
OK so I went and looked it up, VORP is built off of taking Box Plus-Minus (a rate stat), and then adjusting for playing time and pro-ration and pace etc in order to turn it into a counting stat. BPM itself takes into account the defensive metrics tracked in a box score - steals, blocks and rebounds - but the author acknowledges that that's only a small fraction of what's important in playing good defense.

Do we have a better stat available on the defensive front? I know the era of player-tracking and play-by-play data is pretty recently dawned.

FWIW, the +/- per 100 and On-Off numbers for the Celtics Bench:
Wanamaker +5.5, -1.6
GW. +7.2, +1.0
Ojeleye -.3, -9.5
Green. -1.4, -9.1
Kanter +9.6,+4.3
RW +5.1, -1.6
Langford +10.6, 4.6.

So there's a number that should represent both offensive and defensive contributions. They're small sample sizes for some of these guys, and it's not a personal stat but what the team has done when they're on and off the floor. On-Off means that the team was 4.3 points better per 100 possessions when Kanter was on the floor compared to when he was off. Only Semi and Green come off as below average.

wade boggs chicken dinner · Aug 28, 2020

According to NBA math, GW has been very good compared to other rookies taking into account offensive points added and defensive points saved.

View: https://twitter.com/NBA_Math/status/1298635059100975104

Eddie Jurak · Aug 28, 2020

JakeRae said:
I think the non-big formulation makes little sense because different teams run different rotations. At the end of the day, the Celtics are a bit hamstrung on depth because they only use one big and 3 of their 8 players who contribute positively are bigs. I’d also add that cutting Wanamaker out of the list of guys compared masks just how terrible our bench is because he is a big part of why it is bad. Finally, I would avoid PER as a metric as a general rule and would suggest looking to impact metrics like those compiled by Basketball Index.

This is really important, especially when you consider how the Celtics match up against the Raptors.

In addition to not having Gordon Hayward, Enes Kanter is likely to have limited use in this series. As far as bigs go, the Celtics are going to need a lot of quality minutes from Theis, Grant, and TL; and maybe even a few from Ojeleye. The Raptors will likely exploit the crap out of any Kanter minutes. So, as a practical matter, the Celtics will essentially be without 3 of their top 11 regular season minutes guys (Hayward, Kanter, Green) including one key starter.

Assming some combination of Theis, Grant, and TL can cover the center position, what do they have for wing/point depeh behind their starting unit of Tatum, Brown, Walker, and Smart?

Wanamaker is their top bench wing, a role he is overmatched in. Then it is Langford and Ojeleye. The Celtics are going to have to ride Tatum, Brown, and Smart hard, because the cupboard is pretty bare after that. I think Langford is eventually going to be at least a good all-around player, but he has yet to develop an offensive game at the NBA level. Can Ojeleye hit enough from 3 to make the Raptors pay attention to him? He's streaky.

The Celtics got 2/3 of their points vs the Sixers from the trio of Tatum, Brown, and Kemba. That percentage likely goes up against the Raptors, win or lose.

BaseballJones · Aug 28, 2020

Well for a long time people have said that NBA success is built around the stars. Let's see if that holds true for Boston now. They've got three absolute studs on the team right now, all capable of going off for 30 any given night, and all provide leadership and much more. If they want to beat the defending champs, Tatum, Brown, and Kemba are going to need to really show up big.

I hope now is when we see that treating Kemba's knee with kid gloves will pay off. They're going to need him to play a lot in this series.

DeJesus Built My Hotrod · Aug 28, 2020

Thank you. Do you happen to have Semi',

chilidawg said:
FWIW, the +/- per 100 and On-Off numbers for the Celtics Bench:
Wanamaker +5.5, -1.6
GW. +7.2, +1.0
Ojeleye -.3, -9.5
Green. -1.4, -9.1
Kanter +9.6,+4.3
RW +5.1, -1.6
Langford +10.6, 4.6.

So there's a number that should represent both offensive and defensive contributions. They're small sample sizes for some of these guys, and it's not a personal stat but what the team has done when they're on and off the floor. On-Off means that the team was 4.3 points better per 100 possessions when Kanter was on the floor compared to when he was off. Only Semi and Green come off as below average.

Do you happen to have Semi's defensive +/- per 100 handy? No worries if not but just curious. He is one of the players whom I am think shows up differently on the Celtics internal metrics/grading than what is publicly available. Put another way, unless he has pictures of Stevens with his collar unbuttoned, the "traditional" advanced metrics suggests that Semi should be getting far less run than he does. Instead his minutes per game have increased this year though some of that is due to injury, its not all of it.

It will be telling what they do with his option during the offseason. Given he knows his role, is relatively cheap and the scheduled offseason/preseason is compressed, he may well be back.

wade boggs chicken dinner · Aug 28, 2020

chilidawg said:
FWIW, the +/- per 100 and On-Off numbers for the Celtics Bench:
Wanamaker +5.5, -1.6
GW. +7.2, +1.0
Ojeleye -.3, -9.5
Green. -1.4, -9.1
Kanter +9.6,+4.3
RW +5.1, -1.6
Langford +10.6, 4.6.

So there's a number that should represent both offensive and defensive contributions. They're small sample sizes for some of these guys, and it's not a personal stat but what the team has done when they're on and off the floor. On-Off means that the team was 4.3 points better per 100 possessions when Kanter was on the floor compared to when he was off. Only Semi and Green come off as below average.

Thanks for posting. Are these playoff or regular season stats? Also, did you get these from NBA.com or some other source? Just curious.

Jimbodandy · Aug 28, 2020

DeJesus Built My Hotrod said:
Thank you. Do you happen to have Semi',

Do you happen to have Semi's defensive +/- per 100 handy? No worries if not but just curious. He is one of the players whom I am think shows up differently on the Celtics internal metrics/grading than what is publicly available. Put another way, unless he has pictures of Stevens with his collar unbuttoned, the "traditional" advanced metrics suggests that Semi should be getting far less run than he does.

I think that you're onto something with their internals.

At the risk of beating a dead horse, given our team makeup at the top, I think that the focus on the bench guys should always be on how well they function defensively.

Of course it matters if Semi/Grant can nail open threes, but I still think that we're not depending on bench guys to score. When I'm pondering go forward rotation for this season at least, what I care about is what situations suit EK, SO, RW, and GW. Knowing that 3 of those guys suck offensively isn't material.

chilidawg · Aug 28, 2020

wade boggs chicken dinner said:
Thanks for posting. Are these playoff or regular season stats? Also, did you get these from NBA.com or some other source? Just curious.

Regular season. They're from Basketball Reference, https://www.basketball-reference.com/teams/BOS/2020.html. Playoff numbers are there too, the bench has very strong numbers, in large part due to the great game 2 performance. Kanter, GW, Wanamaker and Langford were all a part of that big run that turned a 14 point deficit into a blowout win, and all have very strong +/- as a result.

DeJesus Built My Hotrod · Aug 28, 2020

chilidawg said:
Regular season. They're from Basketball Reference, https://www.basketball-reference.com/teams/BOS/2020.html. Playoff numbers are there too, the bench has very strong numbers, in large part due to the great game 2 performance. Kanter, GW, Wanamaker and Langford were all a part of that big run that turned a 14 point deficit into a blowout win, and all have very strong +/- as a result.

Did you get the +/- per 100 possessions from B-Ref too? I can't seem to find the components for the numbers you posted and, again, I am looking for Semi's defensive +/- per 100. That should be pretty damn negative to get him to the minus 0.3 number posted above.

wade boggs chicken dinner · Aug 28, 2020

chilidawg said:
Regular season. They're from Basketball Reference, https://www.basketball-reference.com/teams/BOS/2020.html. Playoff numbers are there too, the bench has very strong numbers, in large part due to the great game 2 performance. Kanter, GW, Wanamaker and Langford were all a part of that big run that turned a 14 point deficit into a blowout win, and all have very strong +/- as a result.

OK. I was looking at NBA.com and they have different numbers but basically in the same ballpark. Thanks for doing that.

DannyDarwinism · Aug 28, 2020

Mainly to reiterate the issues JakeRae raised, this would be more telling if it was done using Playoff Minutes Played, and including bigs, and with PIPM, RAPTOR or Simple Average. As is, it maybe tells us something about the usage of the (loosely defined) wing players of various playoff teams during the regular season.

wade boggs chicken dinner · Aug 28, 2020

DannyDarwinism said:
Mainly to reiterate the issues JakeRae raised, this would be more telling if it was done using Playoff Minutes Played, and including bigs, and with PIPM, RAPTOR or Simple Average. As is, it maybe tells us something about the usage of the (loosely defined) wing players of various playoff teams during the regular season.

someone could parse RAPTOR here - https://projects.fivethirtyeight.com/2020-nba-player-ratings/ Though I have to admit I have to wonde about any system that has Korkmaz as #2 rated in defense (+13.1, behing Tony Bradley of the Jazz). I mean Korkmaz is no Shake Milton but he's been more or less unplayable v Cs.

DannyDarwinism · Aug 28, 2020

wade boggs chicken dinner said:
someone could parse RAPTOR here - https://projects.fivethirtyeight.com/2020-nba-player-ratings/ Though I have to admit I have to wonde about any system that has Korkmaz as #2 rated in defense (+13.1, behing Tony Bradley of the Jazz). I mean Korkmaz is no Shake Milton but he's been more or less unplayable v Cs.

Just imagine how bad he must be when he’s that much better offensively than what they’re rolling out!

All comprehensive stats have their outliers (let’s call them “Amir Johnson’s”), so take BPM or RPM, or DARKO, if that makes you feel better. The point is, PER and Win Shares are, to a way too large extent, usage based. They don’t tell you much at all about how a guy performs in his role.

DannyDarwinism · Aug 28, 2020

Damn right I’ll always step up to Grant Williams slander.

benhogan · Aug 29, 2020

1. When looking at rookies/young players (ie Grant/Romeo) the situation is fluid. They are better/more aware of their role now than when they racked up some of their metrics last fall. So comparing them to other teams veteran bench players using a full season of PER/Raptor/etc can be a hair misleading.

2. Player use/role is different in the playoffs vs regular season. With Kemba out during the regular season and his minutes carefully managed Brad Wanamaker had to play more minutes, in situations that were not advantageous to play him.

3. Tatum/Kemba/Smart/Brown will play more high leverage minutes and take more shots in the playoffs which will minimize the impact of the bench.

4. And lastly, Brad has scouted the match-ups, playing the same team for 7 straight games will lead to specific bench match ups (ie Kanter vs Joel or Grant vs Horford in the 76ers series) as opposed to playing different teams nightly during the regular season.

All of this will de-emphasize the impact of the Celtics young/new bench in the playoffs vs the regular season.

Eddie Jurak · Aug 29, 2020

benhogan said:
1. When looking at rookies/young players (ie Grant/Romeo) the situation is fluid. They are better/more aware of their role now than when they racked up some of their metrics last fall. So comparing them to other teams veteran bench players using a full season of PER/Raptor/etc can be a hair misleading.

2. Player use/role is different in the playoffs vs regular season. With Kemba out during the regular season and his minutes carefully managed Brad Wanamaker had to play more minutes, in situations that were not advantageous to play him.

3. Tatum/Kemba/Smart/Brown will play more high leverage minutes and take more shots in the playoffs which will minimize the impact of the bench.

4. And lastly, Brad has scouted the match-ups, playing the same team for 7 straight games will lead to specific bench match ups (ie Kanter vs Joel or Grant vs Horford in the 76ers series) as opposed to playing different teams nightly during the regular season.

All of this will de-emphasize the impact of the Celtics young/new bench in the playoffs vs the regular season.

That's right. This is going to be on the starters to win or lose, and the bench's objective will be to get in there and compete for the few minutes the starters need to rest. However, they are missing one key starter and this is a series where they cannot really expect much help from one key reserve (Kanter). I think Toronto should be the favorites here and the difference will be whether Tatum rises to a new level. The good news, I guess, is if they can get by Toronto, maybe they get Hayward back at some point in the next series.

Saints Rest · Aug 29, 2020

benhogan said:
1. When looking at rookies/young players (ie Grant/Romeo) the situation is fluid. They are better/more aware of their role now than when they racked up some of their metrics last fall. So comparing them to other teams veteran bench players using a full season of PER/Raptor/etc can be a hair misleading.

2. Player use/role is different in the playoffs vs regular season. With Kemba out during the regular season and his minutes carefully managed Brad Wanamaker had to play more minutes, in situations that were not advantageous to play him.

3. Tatum/Kemba/Smart/Brown will play more high leverage minutes and take more shots in the playoffs which will minimize the impact of the bench.

4. And lastly, Brad has scouted the match-ups, playing the same team for 7 straight games will lead to specific bench match ups (ie Kanter vs Joel or Grant vs Horford in the 76ers series) as opposed to playing different teams nightly during the regular season.

All of this will de-emphasize the impact of the Celtics young/new bench in the playoffs vs the regular season.

Great points. I think a basketball team’s deep bench is like a baseball team’s bullpen. In the playoffs, a good coach can minimize usage to maximize effectiveness thru ideal matchups.
Franconia did this.
Cora did this.
Hopefully, Stevens will as well.

benhogan · Aug 29, 2020

Saints Rest said:
Great points. I think a basketball team’s deep bench is like a baseball team’s bullpen. In the playoffs, a good coach can minimize usage to maximize effectiveness thru ideal matchups.
Franconia did this.
Cora did this.
Hopefully, Stevens will as well.

Yes, middle relief/platoon situations, basically good coaching

One of Brad's more underrated strengths as a coach is his ability to get the best out of pretty average players by putting them in good situations/match-ups.

The Brad Machine has worked wonders for Centers over the years.

wade boggs chicken dinner · Aug 29, 2020

benhogan said:
The Brad Machine has worked wonders for Centers over the years.

Centers, college players, IT4, Evan Turner, KO, Jae Crowder, and Jonas Jerakbo, Who did I miss?

benhogan · Aug 29, 2020

wade boggs chicken dinner said:
Centers, college players, IT4, Evan Turner, KO, Jae Crowder, and Jonas Jerakbo, Who did I miss?

Avery's production dropped like a rock after he left, but that may be due to injuries?

I guess the better question is who has lept outside of Brad's grip and was suppressed by Stevens? crickets

wade boggs chicken dinner · Aug 29, 2020

benhogan said:
I guess the better question is who has lept outside of Brad's grip and was suppressed by Stevens? crickets

Jared Sullinger did pretty well in a couple of games in the Tournament.

bigq · Aug 29, 2020

wade boggs chicken dinner said:
Centers, college players, IT4, Evan Turner, KO, Jae Crowder, and Jonas Jerakbo, Who did I miss?

All of Australia comes to mind but he would be included in the center category. Terry Rozier is another.

DannyDarwinism · Aug 29, 2020

InstaFace said:
PER is a rate stat, how is it usage-based?

I left bigs aside because they were more confounding to the analysis to leave in. Players can be somewhat fungible 1-4 and can guard or play a range of those positions, but there's more of a categorical difference between your Jaylen Browns and a true center in terms of what they can do and why you'd have certain ones on your roster (leaving your rare swings like Horford, Adebayo and AD aside). But non-bigs can displace each other from a roster, almost without regard to position (though of course, not without regard to their particular skills and abilities and matchups), whereas bigs are kinda competing for roster spots within their own category. And I have fewer concerns about our big rotation than I do about our non-big rotation, which is the category that includes a few rookies being looked-to for growth and assuming larger roles.

I should have been more precise- PER tends to value volume regardless of efficiency. I'm not sure what the "break even" point is now, where more shooting will increase PER, but a couple years ago it was something like 30% on twos and 20% on threes. So a high-volume, low efficiency scorer, like say Brandon Jennings, has a career PER of 15.7 with a poor career TS% of 49.7%, whereas Marcus Smart stands with a PER of 12.1, with a similar 50.1 TS%. The difference is mainly attributable to Jennings having a much higher FGA rate. A good metric shouldn't reward players for inefficient volume. Brandon Jennings, an "above average player" according to PER, who's still in his prime, was last seen playing in Russia.

PER also doesn't do much to capture defense (it only accounts for blocks and steals), doesn't account for who a player is playing with or against, and, of course, does not capture any non-box score impact- screening, boxing out, good defense that doesn't lead to blocks/steals, ball movement that doesn't lead to an assist, etc... A lot of the stuff that's important to contributing to winning basketball games. It doesn't tell us anything about whether a team is better when a guy is on the court or off. Even a guy like Draymond, who's fantastic at generating blocks and steals, is an on/off beast, and even generally rates well in offensive RPM, has a 15.1 (dead average) career PER, lower than Brandon fickin' Jennings, despite being better at everything on a basketball court except chucking up shots.

There's a good reason nobody in the basketball analytics world uses PER. It's a very blunt instrument with little predictive value, and I just don't think it tells us much about the respective bench depth of play-off teams.

This is a pretty comprehensive breakdown and discussion of the pros and cons of PER: https://www.reddit.com/r/nba/comments/cmrkv3/basketball_stat_player_efficiency_rating_per/

I think Jeremias Engelmann's Multi-year RAPM (obviously not particularly useful for rookies) or Jacob Goldstein's PIPM. https://fansided.com/2018/01/11/nylon-calculus-introducing-player-impact-plus-minus/ would be much more useful tools for this kind of analysis.

On Assets and Liabilities: Celtics Playoff Bench vs their Peers

The Ultimate One

Well-Known Member

Eli apologist

The Ultimate One

Dope

Member

Member

Well-Known Member

The Ultimate One

Member

The Ultimate One

Member

The Ultimate One

Member

Member

The Ultimate One

Member

Member

used to be tbb

Member

Member

Member

Member

Well-Known Member

Member

Member

Member

Member

canderson-lite

ivanvamp

Well-Known Member

Member

Member

Member

Well-Known Member

Member

Member

Member

Member

Member

The Ultimate One

Granite Truther

canderson-lite

Well-Known Member

Granite Truther

Member

Granite Truther

Member

Member

Member