Dallas could be really bad. Dirk is 38 and was really good last year, but sometimes guys that age fall off a cliff.
Celtics 16/17 Roster and Assets
- Thread starter sox311
- Start date
Does much the top of the East moving backward (Atl, Mia, Chi) mean Boston has sneaked into the top 3 in the conference?
Or am I overestimating Atl's back-sliding and underestimating the lack of rebounding from (presumably) losing Sully?
Or am I overestimating Atl's back-sliding and underestimating the lack of rebounding from (presumably) losing Sully?
Inspired by this thread, here's Version 0.1 of my wins projections.
This extremely rough. All minutes projections are 100% automated, based on the rotoworld depth charts right now, so there are some whacky results. I've also got every player playing 82 games right now. Plus, free agency is only half over. Sullinger is still with the Celtics on this model for instance.
Many other things will change as well. This is little more than my player efficiency projections combined with a very simple minutes projection.
This extremely rough. All minutes projections are 100% automated, based on the rotoworld depth charts right now, so there are some whacky results. I've also got every player playing 82 games right now. Plus, free agency is only half over. Sullinger is still with the Celtics on this model for instance.
Many other things will change as well. This is little more than my player efficiency projections combined with a very simple minutes projection.
Well, they were tied for top 3 last year and given their additions....I don't think they've snuck in as much as solidified their place there. If they don't finish top 3 I think it will be a pretty big disappointment.
Looking at your projections from last year, I'm guessing your player efficiency projections really like George Hill a lot, based on the Utah and Indy numbers. Unless they really hate Teague, Thad Young, Al Jefferson and Trey Burke and really like Mahinmi, Exum and second-year Hood?
Teague vs. George Hill is a big part of it, yeah. George Hill grades out pretty well (+1.5 per 100), while I have Jeff Teague projected as a -1.2 player.
I do want to reiterate that these are very rough projections, both on the player efficiency side, and the minutes projections side. This is more akin to the first set of projections I ran last year, here.
I do want to reiterate that these are very rough projections, both on the player efficiency side, and the minutes projections side. This is more akin to the first set of projections I ran last year, here.
And here's version 0.2, with a better aging curve projection, and incorporating additional efficiency metrics.
Bowiac,
Do you write on the NBA anywhere that I don't know about (i.e., anywhere not named "here?")
I love your posts and perspective.
Do you write on the NBA anywhere that I don't know about (i.e., anywhere not named "here?")
I love your posts and perspective.
Losing Lopez would help. Rondae Hollis-Jefferson likewise grades out as a pretty good player for them. Plus, Lin is a decent sized improvement over Jack.
24 wins is still a really low projection. It's more a fluke that the Lakers have managed to have an even worse projection. A lot of that is a function of their depth chart, where my automated minutes projections are giving Jordan Clarkson a ton of minutes, at a very poor level of play. Brooklyn doesn't have anything quite as negative. My minutes projections are just a mathematical formula right now - once free agency is over, I'll go through and correct a lot of these to match expected rotations more closely.
24 wins is still a really low projection. It's more a fluke that the Lakers have managed to have an even worse projection. A lot of that is a function of their depth chart, where my automated minutes projections are giving Jordan Clarkson a ton of minutes, at a very poor level of play. Brooklyn doesn't have anything quite as negative. My minutes projections are just a mathematical formula right now - once free agency is over, I'll go through and correct a lot of these to match expected rotations more closely.
I have a reasonably active twitter profile (linked above) where I post NBA stuff during the season. I also expect to be involved with Nylon Calculus this season, although time is going to determine the extent.
Interesting to see the lakers SO bad. I mean it's in their interests to be in the top 3 this year but they've been promising improvement and I'd have thought removing that Kobe would have helped alone!And here's version 0.2, with a better aging curve projection, and incorporating additional efficiency metrics.
The heat could drop sharply, with no Wade no Bosh potentially. And poor Davis. They are murdering his prime, if he gets hurt they could drop right down too.
Kings are also very one man dependent. Any team like that can go from bad to awful in a night.
Crazy to think that if things fall in various ways the Sixers could have two top 5 picks and the Celtics have a top 3 as well. Frankly it's pretty likely.
I feel kind of bad for the Nets, I mean that trade. That Nets Celts trade and the worst case for the Kings from the Sixers Kings trade are just two of the most amazingly lopsided potential trades.
Wade signing with the Bulls is great news for the Celtics, both for the Nets pick and potential trades.
For one, the Heat are expected to match the Tyler Johnson contract now. Two, Calderon was traded to the Lakers (Brooklyn was interested).
In terms of trades, Butler Wade and Rondo are all such strong personalities and are such a bad fit together on the floor, I think that one of them will be traded mid-season. Since Rondo and DWade don't have that much value, I imagine the person traded will be Butler.
For one, the Heat are expected to match the Tyler Johnson contract now. Two, Calderon was traded to the Lakers (Brooklyn was interested).
In terms of trades, Butler Wade and Rondo are all such strong personalities and are such a bad fit together on the floor, I think that one of them will be traded mid-season. Since Rondo and DWade don't have that much value, I imagine the person traded will be Butler.
This is interesting. Bowiac, who do the Nets have that grade out as positive players other than Brook Lopez? I assume Jeremy lin but I can't imagine he's graded out that highly.And here's version 0.2, with a better aging curve projection, and incorporating additional efficiency metrics.
Personally, I think the Lakers will eclipse that predicted win total by a decent amount. Remove Byron Scott and Kobe Bryant add in a good coach (or at least not a historically inept one), some competent veterans and the young talent getting better and I think they are more in the 23-27 win range.
Didn't these projections of yours have us at about 50 last year, too?And here's version 0.2, with a better aging curve projection, and incorporating additional efficiency metrics.
What does that say about the Horford signing? (Actually asking, as opposed to rhetorically asking).
Here's what I have for the Nets roster right now:
The player ratings are marked xKPM+, and are on the same scale as RPM or BPM. Zero is average, most players are between -2 and +2. The Nets benefit from a "floor" I've put in for player ratings of -2.5. There's a lot of research showing replacement level (i.e., D league) players are about -2 points per 100 possessions. There are some players worse than this in the NBA, but it's usually hard to keep your job and play that badly, so I don't allow any player rating projection to go below -2.5. All the Nets players with a -2.5 ratings there are guys who are being lifted up to the floor. All rookies are given a -2 rating right now, although this is something that will be changed in time.
Basically, they have 5 decent players: Lin, RHJ, Booker, Lopez, and Reed.
I had the Celtics between 48 and 50 wins in the various iterations I ran last year, yeah.
The same as with the Nets above, here's what I have for the Celtics right now:
As a reminder, the minutes distribution here comes entirely from the Rotoworld depth charts, divvying up 48 minutes per possession by a formula based on player quality, and the number of players listed at a position. There are some obviously wrong results here (Amir with the same playing time as James Young..., Sullinger still on the team, etc...). Once free agency is over, the minutes projections will be replaced with more realistic numbers, but I don't want to go through that process for every team with so many players still in flux. This is a decent proxy for now however.
Woah. Why does that metric love Amir Johnson so much? Or, more to the point, what am I missing about Amir Johnson? Also, why is Olynyk so close to Horford? Am I just bad at reading that metric?
And Bradley so low? What am I missing?
And Bradley so low? What am I missing?
You're reading it all correctly.
Bradley has a big differential between his BPM (box score metric) and his RAPM (plus/minus metric). There's a lot of research showing that the best way to predict future plus/minus results is a weighted blend of hbox score metrics and pure adjusted plus/minus metrics. I've done my own research to that effect, and the ratings above reflect that. Bradley is being dragged down by his poor BPM here. But this too is a work in progress fwiw, and will change when I have time to do more work on the split between offense and defense. I expect Bradley to improve by a decent amount once I do that.
Plus/minus metrics are big fans of Amir and Olynyk. It's not that Horford ranks low, it's that the Celtics' other bigs already rank well. Amir in particular has been an RAPM star for his whole career. Neither of them have big splits between their plus/minus metrics and box score metrics fwiw. They grade out well regardless.
Bradley has a big differential between his BPM (box score metric) and his RAPM (plus/minus metric). There's a lot of research showing that the best way to predict future plus/minus results is a weighted blend of hbox score metrics and pure adjusted plus/minus metrics. I've done my own research to that effect, and the ratings above reflect that. Bradley is being dragged down by his poor BPM here. But this too is a work in progress fwiw, and will change when I have time to do more work on the split between offense and defense. I expect Bradley to improve by a decent amount once I do that.
Plus/minus metrics are big fans of Amir and Olynyk. It's not that Horford ranks low, it's that the Celtics' other bigs already rank well. Amir in particular has been an RAPM star for his whole career. Neither of them have big splits between their plus/minus metrics and box score metrics fwiw. They grade out well regardless.
Olynyk measuring out this well really, really seems to contrast what a lot of casual fans say about him (just a stiff, crap, Danny blew the pick, etc.)
I knew he graded out well, but at #16 overall, he was far and away the biggest surprise for me from the link you posted the other day on RAPM since 2001.
I do think there are some oddities to Olynyk's ratings - he grades out better defensively than offensively by some metrics for instance. I think he's an underrated defender by the "eye test" because he moves better laterally than vertically, but that's a bit of a head scratcher all the same.
Amir's rating I basically take at face value however. He's been good for a long time, across multiple teams, coaches, teammates, and even positions. His rating is fairly "robust" as far as plus/minus metrics go.
Amir's rating I basically take at face value however. He's been good for a long time, across multiple teams, coaches, teammates, and even positions. His rating is fairly "robust" as far as plus/minus metrics go.
I've generally been really impressed with Olynyk's defense except when he has to guard strong post players in the post. He lacks the strength to really man up against guys who know how to use their bodies on the inside. But, there are not a lot of those guys in the league and Stevens didn't ask him to do that very much. Against stretch 4's and 5's, Olynyk plays good man defense and does an excellent job of switching and closing down angles. His defense isn't flashy, but it is legitimately very good.
Interesting conversation about Olynyk. I notice plus/minus stats a lot. While in any one game they are largely noise, over a longer period they usually match up pretty well with what I perceive to be a player's worth. KO just looks goofy out there on the floor, though, and he fouls a lot. So his +/- figures, which always look good, are a head-scratcher. I look in vain for reasons to explain them and I guess these posts are a good starting point.
The lane is never clogged when he is on the floor and he can stay in front of guys. He gets killed by traditional bigs in the post and I think that sticks out when watching him on D.
Nate Silver at 538 just published his CARMELO* rankings, which include this really interesting projection on other players with similar career trajectories to Al Horford:
You can play around with the tool here: http://projects.fivethirtyeight.com/carmelo/al-horford/
*Career-Arc Regression Model Estimator with Local Optimization
You can play around with the tool here: http://projects.fivethirtyeight.com/carmelo/al-horford/
*Career-Arc Regression Model Estimator with Local Optimization
Fwiw, 538 appears to have stopped using RPM or any plus/minus metric in CARMELO. Instead, they've moved to a purely BPM based system, in order to generate historic comps from before RPM was available. BPM is a purely box score stat, so it's available much further back (no play by play data needed).
I understand the appeal, since players comps are fun, but the result is that I'd take their projections a lot less seriously. Losing plus/minus data really hurts your defensive metrics, and I'm not sure how much you gain by being able to compare players today to guys from before the three point line existed for instance.
I understand the appeal, since players comps are fun, but the result is that I'd take their projections a lot less seriously. Losing plus/minus data really hurts your defensive metrics, and I'm not sure how much you gain by being able to compare players today to guys from before the three point line existed for instance.
Mark Murphy @Murf56 2m2 minutes agoSalt Lake City, UT
Ainge said options on Johnson and Jerebko being picked up. Ante Zizic will stay in Europe for a year. Team hasn't made that decision on Yabu
Ainge said options on Johnson and Jerebko being picked up. Ante Zizic will stay in Europe for a year. Team hasn't made that decision on Yabu
Brian Robb @CelticsHub 7m7 minutes agoSalt Lake City, UT
Ainge said roster is not completed. Team still looking for upgrades, especially in shooting department.
Ainge said roster is not completed. Team still looking for upgrades, especially in shooting department.
The Celt's D should be outstanding this season. Brad talked last year about how good Amir was at defending the 4, adding Horford to the interior D with the excellent perimeter defenders and the Celts D will be top 3 defensively.
This sounds like acquiring another playmaker to replace Turner is not on Ainge's priority list, so he must have confidence that Rozier and Smart can initiate the offense.
Copying amfox's post from the NBA thread:
I don't see any must-haves in that list for the Celtics, and certainly not if the signing would compromise their ability to offer a max contract during the next offseason.
I want crabbe but Portland would match. Feel like you could get Harkless but if it's only one year deal Portland prob matches? They just signed Ezeli tho and have Leonard to possibly Match an offer sheet too.
Thinking about the thread title - Celtic assets; the turn around accomplished by Ainge is remarkable. Arguably, when you look at the assets of every team, you could easily place the Celtics as #3 currently in total assets (obviously GSW and Cleveland have more total assets). It is quite possible that the Celtics are the #2 or 3 seed in the East for the next 2 years and also have top 3 draft picks for those both those years.
I would think at this point an upgrade would come via trade. I'd guess since Sullinger still isn't signed, he must have mutual interest with a team over the cap that would trade for him. The Clippers would be a fit since they desperately need a big to go in their rotation, and he fits the profile Doc Rivers looks for, has previously played for Doc Rivers. Too bad the Clippers have no assets outside of Blake Griffin and Chris Paul that Boston would want.
So what have we got for a rotation next year?
C Horford
PF Amir
SF Crowder
SG Bradley
PG IT4
Bench:
Smart
Rozier
Olynyk
Brown/Jonas
Deep bench:
Jonas/Brown
RJ
Big to be named later
Mickey
Portland:
Jackson
Bentil (autocorrects to lentil)
Europe:
Zizic
Yabusele
Gone:
Young
C Horford
PF Amir
SF Crowder
SG Bradley
PG IT4
Bench:
Smart
Rozier
Olynyk
Brown/Jonas
Deep bench:
Jonas/Brown
RJ
Big to be named later
Mickey
Portland:
Jackson
Bentil (autocorrects to lentil)
Europe:
Zizic
Yabusele
Gone:
Young
Supposedly the Kings are looking to move a couple of players headlined by Rudy Gay but the name that is interesting to me is Ben McLemore. I assume he's in a contract year but is he a good enough shooter/wing/defender to play 25 minutes at the 2 and small ball 3 for the Celts this year? I don't watch enough Kings games to know but I remember him being a top level athlete for that Kansas team.
This is great stuff. Thanks for sharing.I had the Celtics between 48 and 50 wins in the various iterations I ran last year, yeah.
The same as with the Nets above, here's what I have for the Celtics right now:
Bradley @ .1 really?
Where did some of the top drafted players grade out (ie Bender, Hield, Murray, Dunn)?
I'd take a shot on McLemore. Hasn't really put it all together yet but tough to judge anyone on that dumpster fire of a team. Would be a far better option off the bench than RJ Hunter or James Young. Not sure what we could offer to make it worth it for Sac but would gladly trade either or both for McLemore. He's an RFA next summer so they couldn't expect much.
On mobile so cant link but NBA transcation wire has the Cs pulling the QO to Sully and renouncing him.
Didnt see that one coming.
Didnt see that one coming.