I'm not suggesting getting rid of Zach. I want Zach to play and play a lot. Hopefully everything we imagine with our rose colored glasses comes true. What I don't like doing is ditching ACTUAL good players based on fantasies about how someone MIGHT pan out.
Enes Kanter's game is garbage (3 Viewers)
- Thread starter Blazer4ever
- Start date
Welcome to our community
Be a part of something great, join today!
I'm not suggesting getting rid of Zach. I want Zach to play and play a lot. Hopefully everything we imagine with our rose colored glasses comes true. What I don't like doing is ditching ACTUAL good players based on fantasies about how someone MIGHT pan out.
Ok, well here you go. I literally spent all day on this! While I was kind of right, that missed outside shots & 3-pointers hurt less than inside shots, and that layups are the worst, it's not to the degree I expected. I especially expected a worse result than this for missed layups. My guess explanation is that perhaps teams commit extra turnovers rushing to exploit the 5 on 4 after a missed layup, which partially offsets the advantage.
TorturedBlazerFan
Well-Known Member
- Joined
- Jan 31, 2018
- Messages
- 19,824
- Likes
- 23,357
- Points
- 113
Whelp I feel bad. Interesting stats though. Thanks for spending your time on it.
I'm not sure what you mean. The whole point was to see which missed shots are most detrimental to your defense. So having bad set defense due to a missed layup (shooter under the basket) should be included. Are you saying, eliminate fast-break missed layups but include half-court missed layups? I'm not sure I can detect as fast-break layup vs a half-court layup. Maybe I can by time of rebound vs time of layup.
Nah, I love doing this stuff. Plus now I have raw play by play data again. It's kind of a nightmare though. The NBA stats API sucks and it likes to punish you every now and then by disabling you for 10 minutes if you throttle their API too much.
TorturedBlazerFan
Well-Known Member
- Joined
- Jan 31, 2018
- Messages
- 19,824
- Likes
- 23,357
- Points
- 113
Oh were you using the nba-api on PyPi.org?
I like basketball reference for how easy it is to scrape and parse all the info. I havent really used the NBA’s api before.
BonesJones
https://www.youtube.com/c/blazersuprise
- Joined
- May 7, 2015
- Messages
- 44,580
- Likes
- 38,679
- Points
- 113
Fans are more negative on him than they should be. Look at his stats this year. Hes been solid on both ends.
TorturedBlazerFan
Well-Known Member
- Joined
- Jan 31, 2018
- Messages
- 19,824
- Likes
- 23,357
- Points
- 113
Im not sure if he’ll ever be a foundational piece like he was sold as. Im kind of dissapointed in that the Kanter trade seems to mean we wont see him much. I dont know what that does for his developement or his confidence.
Its kind of weird altogether how this season has gone for him.
handiman
Well-Known Member
- Joined
- Oct 15, 2008
- Messages
- 5,881
- Likes
- 3,916
- Points
- 113
Correct. I assume the same as you, that they're tough to separate out, which is why I asked.
I figure they're grouped together in the data and may be producing misleading averages. I would think a missed fast-break layup has a much higher likelihood of being run back on you for a score than does any other attempt.
BonesJones
https://www.youtube.com/c/blazersuprise
- Joined
- May 7, 2015
- Messages
- 44,580
- Likes
- 38,679
- Points
- 113
I've seen so many people say "Kanter trade" haha.
blazerkor
Well-Known Member
- Joined
- Sep 15, 2014
- Messages
- 16,686
- Likes
- 18,197
- Points
- 113
Yeah it's important to remember we got him off the buyout pile for nothing. I doubt he's back next season but he's an alright backup C. Not complete garbage.
I’m using the same approach they are but I’m a C# guy so I rolled my own. I don’t know python very well. Can you get individual play by play events from basketball reference? The NBA stats api is not very good for play by play. They don’t even give you who is on the floor for each event or who has possession or even the current score and much of the info is embedded in a text field that has to be parsed.
Correct. I assume the same as you, that they're tough to separate out, which is why I asked.
Yeah, fast break layups are definitely in there. I could locate them by time difference between gaining possession and time of shot.
Before I do something like that I need to make sure I’m grouping events into possessions correctly. The api doesn’t tell you who has the damn ball. It tells you who an event pertains to and that can be home, away, or neutral. So while home still has the ball, away can have many events like timeout, substitution, lane violation, personal foul, technical foul shot. So correctly keeping track of who has the ball is a bit tricky and I need to check that thoroughly.
Yeah he’s doing ok and so young he’s gonna get much stronger. It’s just that fan expectations are so high.
TorturedBlazerFan
Well-Known Member
- Joined
- Jan 31, 2018
- Messages
- 19,824
- Likes
- 23,357
- Points
- 113
I know they have something called play index, I haven't used it but I assumed that did play by play. C# is a fun language, really good for application development, I've used it for a few projects. My current project is more HW related and we're using C++, I love how powerful C++ is but it's so verbose and it's so easy to screw up, anywho. I have this idea for using the hololens for sports applications where on the fly a coach or assistant coach could have so much visual data on the actual court, football field etc. I only say that because that will be a C# application, using Unity and the MRTK. Oh gosh I'm geeking... Have a good one.
tester551
Well-Known Member
- Joined
- Jan 23, 2009
- Messages
- 4,037
- Likes
- 3,850
- Points
- 113
That you are...
wizenheimer
Well-Known Member
- Joined
- Oct 19, 2008
- Messages
- 25,102
- Likes
- 38,205
- Points
- 113
the problem is (or was if you go back to when Kanter was signed), Portland is in a battle for playoff seeding and HCA; and, they all know that big changes could be coming if they fail in the playoffs again...
it's all win-now dynamic, and if you go by the numbers, Zach wasn't helping much in win-now. He is 9th on the team in on-off plus/minus. Portland is +5.9 points better offensively when he's not on the floor, but only 0.4 points better defensively with him on the floor (both per 100 possessions). That's a difference of +5.6 points and Portland's MOV is only +3.3 points
further, Portland's offensive rebounding rate is 26% better when Zach sits; and their overall rebounding rate is 50.8% with Zach on the floor, but 53.4% when he's sitting. Portland even has a better blocked shot rate when Zach is off the floor. In other words, Zach isn't helping nearly enough on defense to offset his offensive and rebounding deficiencies....if you credit the numbers
of course, the issue is, while Kanter has made a big difference in rebounding, for all those other numbers I gave for Zach, Kanter has been spectacularly worse. For instance, while Portland is a net 5.6 points worse with Zach on the floor, they are 17 points worse with Kanter on the floor. Yikes!
Why was Kanter signed? well, I think Olshey's seat is at least warm, if not downright hot. He needed to do something major, but he was apparently hoping that 3 minor deals would make for a major deal. And Kanter seems like a Stotts C insofar as he can get his own offense, and we know that one-on-one players seem to be like catnip for Stotts
Kanter has only played 8 games for Portland and only 18.5 minutes a game, so the statistical sample size is extremely small. We do know what his statistical history is though, and that's not encouraging
Stotts still has the option of going back to Zach in the rotation, and may if the Kanter experiment becomes and obvious failure.
TorturedBlazerFan
Well-Known Member
- Joined
- Jan 31, 2018
- Messages
- 19,824
- Likes
- 23,357
- Points
- 113
lol good call, good call.
BonesJones
https://www.youtube.com/c/blazersuprise
- Joined
- May 7, 2015
- Messages
- 44,580
- Likes
- 38,679
- Points
- 113
In other words: Jusuf Nurkic is better than Zach Collins.
wizenheimer
Well-Known Member
- Joined
- Oct 19, 2008
- Messages
- 25,102
- Likes
- 38,205
- Points
- 113
if we're talking Zach at C, sure. For that matter, so is Ed Davis last year and Meyers Leonard this year
also, Zach is part of a below average bench
Zach at PF? Aminu is better, though frustrating to watch at times. And apparently, Stotts has decided that Aminu-Harkless-Layman at PF are better options
my point was that it sure looks like Blazer management, both front office and coaches, believe it's sink-or-swim time. There's no concern the rest of the season for player development. It's win-now and it's going with the rotations they believe give them the best chance....and by best chance, I mean preserve their jobs
blue9
Well-Known Member
- Joined
- Nov 25, 2012
- Messages
- 10,729
- Likes
- 7,169
- Points
- 113
I think the "foundational" aspect has been blown out of proportion. That generally means someone who the offense is going to be built around, or at least tailored to. And unless you've got a Tim Duncan, or maybe a Dirk Nowitski, you shouldn't be building an offense around your PF. So in this sense, Zach absolutely should NOT be a foundational piece.
But Zach can absolutely be the sort of player you want to lock in as your long-term PF. Because a PF should be a jack-of-all-trades that does all the little things that make the team better than the sum of its parts. I don't want Zach taking 15 FGA - I want him setting screens, playing hard-nosed defense, and having enough offense to keep defenses honest. I have no reason to think he can't excel in this role. What I'd really like to see is him playing alongside Nurk.
TorturedBlazerFan
Well-Known Member
- Joined
- Jan 31, 2018
- Messages
- 19,824
- Likes
- 23,357
- Points
- 113
Neil was the guy who said he would be a "foundational" player though. I think Zach can be an ok PF/C.
blue9
Well-Known Member
- Joined
- Nov 25, 2012
- Messages
- 10,729
- Likes
- 7,169
- Points
- 113
Well, Neil's an idiot.
wizenheimer
Well-Known Member
- Joined
- Oct 19, 2008
- Messages
- 25,102
- Likes
- 38,205
- Points
- 113
he probably did
he also said: “He is a franchise-level building block,’’...
that's absurd level over-selling a 10th pick. Olshey's ego probably has it's own bedroom in his mansion
https://www.nbcsports.com/northwest...ing-neil-olshey-was-sold-zach-collins-january
hoopsjock
Well-Known Member
- Joined
- Jan 31, 2018
- Messages
- 16,302
- Likes
- 26,283
- Points
- 113
It still baffles me why Nurk and Collins together hasn't gotten enough run this year to at least see if there is potential there. Why is it that Collins can play with Kanter and Leonard (and Davis last year) but Stotts seems to be unwilling to pair him with Nurk?
Strenuus
Well-Known Member
- Joined
- Aug 9, 2013
- Messages
- 50,990
- Likes
- 36,369
- Points
- 113
blue9
Well-Known Member
- Joined
- Nov 25, 2012
- Messages
- 10,729
- Likes
- 7,169
- Points
- 113
Absolutely baffling.
Sounds like a lot of fun. I watched something a while back where they are starting to use neural networks in sports. Like they want to input real time player movements so the neural networks can "watch" and learn what is going on... and then I don't know what they do... make suggestions I guess. If you've followed Alpha Zero (chess) or Alpha Go at all, these deep NN are getting absolutely amazing what they can do. In chess, Alpha Zero plays very different from normal chess engines. It plays much more human and non-materialistic. Anywho.
TorturedBlazerFan
Well-Known Member
- Joined
- Jan 31, 2018
- Messages
- 19,824
- Likes
- 23,357
- Points
- 113
Ive seen some Neural Nets in video games like starcraft2. Its kind of bad but Ive lost that like “awe” of technology. Havent seen anything in a long time that really made me go, “oh wow”. I think its just seeing it every day kind of makes me take it for granted.
Have you ever made Neural Net? Its actually pretty fun to screw around with. I built one last summer as just a random project for fun, I was just using it to see if I could make enemies learn player behaviors turns out I could.
