Around the NBA: 2017—2018

[slick_watts]

how does one justify hill > ball?

[thekaoswithin]

how does one justify hill > ball?

If you need 3+D vs Playmaking+D on your team due to roster construction, I could see you picking Hill just from a basketball perspective.

Given age, contracts, etc, I'm not sure there is a scenario where it would be justified however.

[Kizz Fastfists]

how does one justify hill > ball?

In a game of HORSE I'd take Hill. If I want ball handling, play making, rebounding, defense and everything other than shooting I'd take Lonzo. Lonzo is the better player and it isn't close unless you are trying to argue that shooting is all that matters. By any criteria in which you can argue that Hill is better than Lonzo you can also argue that Hill is better than Russ.

[bondom34]

how does one justify hill > ball?

Hill had a down year and offense is a thing.

Hill has years of being an above average starter and one or 2 as a below average one, Ball has half a year as a below average one.

TBH Hill played like he didn't care in Sacramento, and if you exclude last year's numbers if you think they're an abberation:

http://bkref.com/tiny/hDxwe

Also, the guy on sub .450 TS isn't convincing me when his team is better when he is off court offensively.

[slick_watts]

how does one justify hill > ball?

Hill had a down year and offense is a thing.

Hill has years of being an above average starter and one or 2 as a below average one, Ball has half a year as a below average one.

TBH Hill played like he didn't care in Sacramento, and if you exclude last year's numbers if you think they're an abberation:

http://bkref.com/tiny/hDxwe

Also, the guy on sub .450 TS isn't convincing me when his team is better when he is off court offensively.

why would they be an aberration? he's been injury prone for years and is 32.

every metric i can see has ball as a rookie comfortably ahead of hill, and they have opposite trajectories heading into next year. i do think it's interesting how you do not care about ts% in certain cases but all of a sudden it's important in this one. because the trash lakers team offense was slightly worse without him? okay.

ball as a below average starter last year? according to what? he's right around average among starting pg last year on rpm and slightly above average among starting pg on bpm. his CARMELO projection has him as a future all-star and a sizable leap in year two. can you guess what 32 year old george hill's trajectory is?

tbh i think you get too focused on players on / off. making cross team comparisons based on on / off is measuring the quality of the other team's reserves as much as it is the player themselves. this is also the issue you encounter with westbrook relative to other elite guards btw.

[bondom34]

how does one justify hill > ball?

Hill had a down year and offense is a thing.

Hill has years of being an above average starter and one or 2 as a below average one, Ball has half a year as a below average one.

TBH Hill played like he didn't care in Sacramento, and if you exclude last year's numbers if you think they're an abberation:

http://bkref.com/tiny/hDxwe

Also, the guy on sub .450 TS isn't convincing me when his team is better when he is off court offensively.

why would they be an aberration? he's been injury prone for years and is 32.

every metric i can see has ball as a rookie comfortably ahead of hill, and they have opposite trajectories heading into next year. i do think it's interesting how you do not care about ts% in certain cases but all of a sudden it's important in this one. because the trash lakers team offense was slightly worse without him? okay.

ball as a below average starter last year? according to what? he's right around average among starting pg last year on rpm and slightly above average among starting pg on bpm. his CARMELO projection has him as a future all-star and a sizable leap in year two. can you guess what 32 year old george hill's trajectory is?

tbh i think you get too focused on players on / off. making cross team comparisons based on on / off is measuring the quality of the other team's reserves as much as it is the player themselves. this is also the issue you encounter with westbrook relative to other elite guards btw.

They were an abberation because in a decade long career, he posted worse numbers only as a rookie and showed no other indications he'd decline like that. Its not a "trajectory" if there's a single data point.

Also, I don't think I've once said I don't care about TS. I've said it's not the same thing as offensive impact, and you can impact offense in other ways. In which case you can, as is shown by a ton of players. Ball however hasn't done that. And not really sure I'd rely that much on CARMELO rankings. They're nice, and a cool frivolity, but nothing to put a ton into (of note, Ball ranks in these just behind Curry in terms of 5 year value, which I highly doubt even more and in terms of 7 year ranks ahead of AD and Embiid).

And I don't think it's a good sign Ball showed a - offensive on/off when his primary backups were Alex Caruso and Tyler Ennis. But maybe they're really good on offense?

And finally I do love trying to drag Westbrook's name into it and take him down, because as a Durant fan who decided to follow him, you should know He has zero to do with this conversation.

[slick_watts]

They were an abberation because in a decade long career, he posted worse numbers only as a rookie and showed no other indications he'd decline like that. Its not a "trajectory" if there's a single data point.

sure there is. age is the data point. he's 32. you expect decline, especially for a player who has historically been injury prone. this doesn't mean that he's going to for sure be worse next year than he was this year. he might even be better. but the overall progression for him looks like it's going to go down. and we know for sure lonzo's is probably going to go up.

lets put it this way. would you be more surprised if ball got better next year or if george hill got worse?

Also, I don't think I've once said I don't care about TS. I've said it's not the same thing as offensive impact, and you can impact offense in other ways.

i agree. would you say ts% would be a more pertinent relative measure for these two players, or two super-high usage scorers?

my problem is the way you tend to measure offensive impact is just pp100 on / off which is so noisy and imprecise. i think you're measuring the performance difference of two players' teammates as much as you're measuring the performance of the individual. you also will use rpm and bpm and other things too sometimes but not in this case apparently because it does not favor your argument.

In which case you can, as is shown by a ton of players. Ball however hasn't done that. And not really sure I'd rely that much on CARMELO rankings. They're nice, and a cool frivolity, but nothing to put a ton into (of note, Ball ranks in these just behind Curry in terms of 5 year value, which I highly doubt even more and in terms of 7 year ranks ahead of AD and Embiid).

doesn't CARMELO have strong performance in the apbr contests? what should we put 'a ton into', then? the letter grades being put on twitter by the guy who is trying to build a scam site like harabolos?

And I don't think it's a good sign Ball showed a - offensive on/off when his primary backups were Alex Caruso and Tyler Ennis. But maybe they're really good on offense?

who did lonzo primarily play with? who did his his replacements play with when he was injured? what were the lakers rotations like? if only we had some sort of metric available that compensated for these things to some extent.

And finally I do love trying to drag Westbrook's name into it and take him down, because as a Durant fan who decided to follow him, you should know He has zero to do with this conversation.

he has plenty to do with it, i think, because you freely swing rpm and bpm around like a battering ram when it is convenient in those discussions. but lonzo ball doesn't count i guess, and CARMELO is a frivolity.

i was a fan of this franchise before kevin durant was even born, kid.

[bondom34]

They were an abberation because in a decade long career, he posted worse numbers only as a rookie and showed no other indications he'd decline like that. Its not a "trajectory" if there's a single data point.

sure there is. age is the data point. he's 32. you expect decline, especially for a player who has historically been injury prone. this doesn't mean that he's going to for sure be worse next year than he was this year. he might even be better. but the overall progression for him looks like it's going to go down. and we know for sure lonzo's is probably going to go up.

lets put it this way. would you be more surprised if ball got better next year or if george hill got worse?

Also, I don't think I've once said I don't care about TS. I've said it's not the same thing as offensive impact, and you can impact offense in other ways.

i agree. would you say ts% would be a more pertinent relative measure for these two players, or two super-high usage scorers?

my problem is the way you tend to measure offensive impact is just pp100 on / off which is so noisy and imprecise. i think you're measuring the performance difference of two players' teammates as much as you're measuring the performance of the individual. you also will use rpm and bpm and other things too sometimes but not in this case apparently because it does not favor your argument.

In which case you can, as is shown by a ton of players. Ball however hasn't done that. And not really sure I'd rely that much on CARMELO rankings. They're nice, and a cool frivolity, but nothing to put a ton into (of note, Ball ranks in these just behind Curry in terms of 5 year value, which I highly doubt even more and in terms of 7 year ranks ahead of AD and Embiid).

doesn't CARMELO have strong performance in the apbr contests? what should we put 'a ton into', then? the letter grades being put on twitter by the guy who is trying to build a scam site like harabolos?

And I don't think it's a good sign Ball showed a - offensive on/off when his primary backups were Alex Caruso and Tyler Ennis. But maybe they're really good on offense?

who did lonzo primarily play with? who did his his replacements play with when he was injured? what were the lakers rotations like? if only we had some sort of metric available that compensated for these things to some extent.

And finally I do love trying to drag Westbrook's name into it and take him down, because as a Durant fan who decided to follow him, you should know He has zero to do with this conversation.

he has plenty to do with it, i think, because you freely swing rpm and bpm around like a battering ram when it is convenient in those discussions. but lonzo ball doesn't count i guess, and CARMELO is a frivolity.

i was a fan of this franchise before kevin durant was even born, kid.

1. 1 data point. I would expect Ball to get better. I'd also expect Hill to be better next year than last.

2. I've never used just on/off as a measure of anything. RPM, RAPM, a bunch of things, but never just one. Balls' RPM and RAPM indicate the same, but I don't really care to drop a laundry list of everything saying he's bad at offense every time I post. I don't find the need nor do I really care enough. Ball was 421 of 540 in RAPM offensively. Hill 54. And I'd value offense from a PG way way over defense.

3. It came in 2nd actually, to a poster from the PC board here. RPM does, and RAPM does, but CARMELO is a blend with BPM involved. And I don't weight BPM that heavily.

4. We have their roster, and his most frequent 2 man lineups were with KCP, Kuzma, Lopez, and Ingram. Ennis's were Kuzma, Randle, Hart, and KCP. And honestly I'd argue KCP is the best of those players.

5. I don't think I've swung BPM around once, because its not something I use that often. DBPM is flat garbage and OBPM is ok but flawed too. RPM I do, and not "like a battering ram". I use RAPM, and RPM, and on/off, and box score measures. Feels like you're either not reading when I post or selectively remembering. Also, I don't actually post my entire thought process when I evaluate players. I don't really have the time to.

6. It was a jab.

Edit: To be clear I like 538s system, but the whole contract evaluation isn't what I mean here. Its who I want for this year only. And it is Hill. Lebron is getting to a point where I wouldn't punt a season. And that Lakers roster is doing that. CARMELO is a fine tool, I shouldn't have used the term frivolity but didn't mean it quite that way.

[slick_watts]

lonzo ball and ben simmons last year are the only two rookie starting pg in nba history to start their careers with a 2% blk / 2% stl season. lonzo obviously isn't simmons overall but discrediting his impact defensively imo is overlooking how much of an outlier it is for his position at the moment.

i know how CARMELO did in apbr and has done over the years which is why i mentioned it. and it does project lonzo to be more valuable on average than hill next season.

obviously we won't agree on this. i still think you give too much credence to on / off. nobody takes individual raw +/- seriously over season long samples even for cross-team comparisons, and controlling for pace doesn't really make it much better imo.

[Pillendreher]

What I'm wondering is: How are you projecting a player based on a single season? At that point, you're just looking at historical comparisons, right?

[spearsy23]

What I'm wondering is: How are you projecting a player based on a single season? At that point, you're just looking at historical comparisons, right?

It's obviously a shot in the dark, but almost every player gets better in year two especially at 20. Ball could be MCW I guess, but that seems like a worst case scenario while expecting Hill to be markedly better than last year seems like a best case scenario.

[slick_watts]

What I'm wondering is: How are you projecting a player based on a single season? At that point, you're just looking at historical comparisons, right?

this may not be true anymore, but my understanding is that rapm uses a prior somewhere between -1.0 and -2.0 for rookies based on draft position. ball and simmons priors will be higher than most rookies for this reason but still negative. this is why when a rookie looks impressive on an rapm-based metric especially in a starting role heads turn.

but i think rpm might have modified this somewhat by adding their own nuance.

[spearsy23]

What I'm wondering is: How are you projecting a player based on a single season? At that point, you're just looking at historical comparisons, right?

this may not be true anymore, but my understanding is that rapm uses a prior somewhere between -1.0 and -2.0 for rookies based on draft position. ball and simmons priors will be higher than most rookies for this reason but still negative. this is why when a rookie looks impressive on an rapm-based metric especially in a starting role heads turn.

but i think rpm might have modified this somewhat by adding their own nuance.

I could be completely wrong here, but I don't believe RPM is prior informed. If I remember correctly it was the first year and not subsequent years. I don't think I ever read a reason, but if I had to guess it's because espn isn't attempting to predict value, they're prefer a comparison tool, which generates more debate and has more randomness.

[slick_watts]

What I'm wondering is: How are you projecting a player based on a single season? At that point, you're just looking at historical comparisons, right?

this may not be true anymore, but my understanding is that rapm uses a prior somewhere between -1.0 and -2.0 for rookies based on draft position. ball and simmons priors will be higher than most rookies for this reason but still negative. this is why when a rookie looks impressive on an rapm-based metric especially in a starting role heads turn.

but i think rpm might have modified this somewhat by adding their own nuance.

I could be completely wrong here, but I don't believe RPM is prior informed. If I remember correctly it was the first year and not subsequent years. I don't think I ever read a reason, but if I had to guess it's because espn isn't attempting to predict value, they're prefer a comparison tool, which generates more debate and has more randomness.

i believe it uses box score priors?

rpm seems to change every year, though.

[Andre Roberstan]

FWIW, re: the Basketball Index numbers—I talk with at least one well-respected Nylon guy who's on board with what they're doing. Goldstein knows his business for sure, and Tim's sharp, if Lakers-biased. Putting letter grades out there without methodology is a bad idea, though, and I think they're pulling some from Synergy player tracking stuff, which is...not great.

[spearsy23]

Putting letter grades out there without methodology is a bad idea, though,

That's entirely dependant on your goal, putting out the stuff that generates interest first helps if you're trying to monetize it. Very few people are going to care about the methodology, most fans will care about and spread the rankings rather they agree or disagree.

[bondom34]

On/off was about the last thing I used to look at Ball. Actually it was the last. I didn't even know it until just before that, because it came up on the trade forum. But everything else to me indicates that for this year only I'd rather Hill, and not very close.

[slick_watts]

FWIW, re: the Basketball Index numbers—I talk with at least one well-respected Nylon guy who's on board with what they're doing. Goldstein knows his business for sure, and Tim's sharp, if Lakers-biased. Putting letter grades out there without methodology is a bad idea, though, and I think they're pulling some from Synergy player tracking stuff, which is...not great.

i think it's a great idea from an accessibility // marketing perspective.

i tend to trust the hobbyists, or the people who design and implement models for teams or competition. reading the model testimonials makes it sound like a model for gamblers, which is an entirely different thing but probably more broadly lucrative. see: harabolos.

[Andre Roberstan]

FWIW, re: the Basketball Index numbers—I talk with at least one well-respected Nylon guy who's on board with what they're doing. Goldstein knows his business for sure, and Tim's sharp, if Lakers-biased. Putting letter grades out there without methodology is a bad idea, though, and I think they're pulling some from Synergy player tracking stuff, which is...not great.

i think it's a great idea from an accessibility // marketing perspective.

i tend to trust the hobbyists, or the people who design and implement models for teams or competition. reading the model testimonials makes it sound like a model for gamblers, which is an entirely different thing but probably more broadly lucrative. see: harabolos.

I'm unsure what Tim's long-term goal is. I think he wants to be a scout. The gambling model thing seems more tailored to doing quant work for Voulgaris types, though. IDK.

He's got a background in film study that most quants don't (and probably most analytics departments don't). Not sure I trust how dogmatic he is about a lot of this stuff, especially at his age, but no question that he's smart.

[Jstock12]

Speaking of Lonzo... How do you manage to go from 55/41/67 shooting splits (in college) to 36/30/45 in just a span of 1 year... Hasheem Thabeet is a better career FT shooter than Lonzo by a whopping 13% Lonzo literally makes Rajon Rondo look like Steve Nash in comparison.