FiveThirtyEight's CARMELO NBA Projections - July 7, 2016 Update

[max215]

LOL Smart vs Bradley.  Smart's 5 year value 156.5 million, Bradley's 35.9 million.  While I agree that there is no comparison in their on court value that does seem a little extreme.

It's projecting improvement for Smart. But yeah, this is a little ridiculous.

I'm very happy the model still likes Smart after his worse-than-expected season.

I think it speaks to how poor Bradley's advanced numbers are. We all know Bradley passes the eye test with flying colors, but his advanced numbers (specifically on defense) are shockingly bad.

[loco_91]

Lol, did anyone see Steph's 5 year market value?!

$370.7 MILLION

[The Oracle]

LOL Smart vs Bradley.  Smart's 5 year value 156.5 million, Bradley's 35.9 million.  While I agree that there is no comparison in their on court value that does seem a little extreme.

It's projecting improvement for Smart. But yeah, this is a little ridiculous.

I'm very happy the model still likes Smart after his worse-than-expected season.

I think it speaks to how poor Bradley's advanced numbers are. We all know Bradley passes the eye test with flying colors, but his advanced numbers (specifically on defense) are shockingly bad.

Bradley has never passed my eye test.  He can be a very annoying defender to ball handlers on the perimeter but as soon as a screen comes he becomes much less effective.  He is a poor help defender.  His steal numbers have never been what you would think they could be with his speed and length.  On offense he has a ton of limitations, poor shot selection, ball handling and passing etc.  He did improve some in a few of those areas last year.

[nickagneta]

LOL Smart vs Bradley.  Smart's 5 year value 156.5 million, Bradley's 35.9 million.  While I agree that there is no comparison in their on court value that does seem a little extreme.

It's projecting improvement for Smart. But yeah, this is a little ridiculous.

I'm very happy the model still likes Smart after his worse-than-expected season.

I think it speaks to how poor Bradley's advanced numbers are. We all know Bradley passes the eye test with flying colors, but his advanced numbers (specifically on defense) are shockingly bad.

Bradley has never passed my eye test.  He can be a very annoying defender to ball handlers on the perimeter but as soon as a screen comes he becomes much less effective.  He is a poor help defender.  His steal numbers have never been what you would think they could be with his speed and length.  On offense he has a ton of limitations, poor shot selection, ball handling and passing etc.  He did improve some in a few of those areas last year.

He seemed to pass the eye test of the 123 people who vote for the NBA All-Defense team. Those people are writers and broadcasters who have been in the industry for a long time. He was voted best defensive guard by that group.

Bradley passes my test. That's for sure. And his team mates and opposition, both coaches and players agree with me.

[The Oracle]

LOL Smart vs Bradley.  Smart's 5 year value 156.5 million, Bradley's 35.9 million.  While I agree that there is no comparison in their on court value that does seem a little extreme.

It's projecting improvement for Smart. But yeah, this is a little ridiculous.

I'm very happy the model still likes Smart after his worse-than-expected season.

I think it speaks to how poor Bradley's advanced numbers are. We all know Bradley passes the eye test with flying colors, but his advanced numbers (specifically on defense) are shockingly bad.

Bradley has never passed my eye test.  He can be a very annoying defender to ball handlers on the perimeter but as soon as a screen comes he becomes much less effective.  He is a poor help defender.  His steal numbers have never been what you would think they could be with his speed and length.  On offense he has a ton of limitations, poor shot selection, ball handling and passing etc.  He did improve some in a few of those areas last year.

He seemed to pass the eye test of the 123 people who vote for the NBA All-Defense team. Those people are writers and broadcasters who have been in the industry for a long time. He was voted best defensive guard by that group.

Bradley passes my test. That's for sure. And his team mates and opposition, both coaches and players agree with me.

Those awards are popularity contests and Bradley received a ton of publicity from the Portland and G.S. games at the end of the year that was unwarranted.  Those that actually know what they are looking at would never place Bradley as an all NBA level defender.

[Robb]

If I'm reading this right from last year (and I may not - the way the data is presented is unclear but I'm flipping back and forth between last year's projections and this year's):

Thomas Projection: 4
Thomas Actual: 7.3

Bradley Projection: 1.8
Bradley Actual: 3.1

Crowder Projection: 1.9
Crowder Actual: 6.1

Yeah...so if this misses that badly on our top 3 players from last year then I'm not likely to give it much consideration this year.

It may have missed that much individually, but was one of a few projections to put the Cs at near 50 wins.

[Snakehead]

I think Smart will improve a lot and I think his defense is better than Bradley's.  All NBA Defense is nice but I don't put much stock in it (guys like Rondo and Kobe on it for years when they didn't belong).

[Robb]

Those awards are popularity contests and Bradley received a ton of publicity from the Portland and G.S. games at the end of the year that was unwarranted.  Those that actually know what they are looking at would never place Bradley as an all NBA level defender.

You're going to need to back that up. What I've read regarded the advanced metrics only reflect negatively on Bradley primarily because he's not a great rebounder, but no one whom I've read or listened to has had anything negative to say about his defense.

[nickagneta]

LOL Smart vs Bradley.  Smart's 5 year value 156.5 million, Bradley's 35.9 million.  While I agree that there is no comparison in their on court value that does seem a little extreme.

It's projecting improvement for Smart. But yeah, this is a little ridiculous.

I'm very happy the model still likes Smart after his worse-than-expected season.

I think it speaks to how poor Bradley's advanced numbers are. We all know Bradley passes the eye test with flying colors, but his advanced numbers (specifically on defense) are shockingly bad.

Bradley has never passed my eye test.  He can be a very annoying defender to ball handlers on the perimeter but as soon as a screen comes he becomes much less effective.  He is a poor help defender.  His steal numbers have never been what you would think they could be with his speed and length.  On offense he has a ton of limitations, poor shot selection, ball handling and passing etc.  He did improve some in a few of those areas last year.

He seemed to pass the eye test of the 123 people who vote for the NBA All-Defense team. Those people are writers and broadcasters who have been in the industry for a long time. He was voted best defensive guard by that group.

Bradley passes my test. That's for sure. And his team mates and opposition, both coaches and players agree with me.

Those awards are popularity contests and Bradley received a ton of publicity from the Portland and G.S. games at the end of the year that was unwarranted.  Those that actually know what they are looking at would never place Bradley as an all NBA level defender.

In other words, you know better than everyone. Sure. Whatever.

[coffee425]

This is so confusing.
I love Smart, and I do see all-start potential in his project... but then why is there only 1 all-star in his comparable top 10

[PhoSita]

This CARMELO projection defies what seems to be common sense too often for me to credit it much.

[Robb]

This is so confusing.
I love Smart, and I do see all-start potential in his project... but then why is there only 1 all-star in his comparable top 10

They're not career comparables, I don't think. Just a comparable stage to a stage in that persons career.

[Bobshot]

I don't know if I should take these projections very seriously. FWIW, they predict Dunn was a much better pick than Brown, Noel is much better than Okafor, and Zeller should be odd man out. They also aren't thrilled with Horford over the next few years.

[byennie]

I don't know if I should take these projections very seriously. FWIW, they predict Dunn was a much better pick than Brown, Noel is much better than Okafor, and Zeller should be odd man out. They also aren't thrilled with Horford over the next few years.

That's just the limitation of any statistical model. Dunn was a much better player than Brown in college. There's no input for athleticism, or intelligence, or work ethic, of college system, etc. Even as a freshman, Brown's advanced metrics weren't that great. He was an inefficient player at Cal. The projections become more accurate over time, with more data.

[byennie]

If I'm reading this right from last year (and I may not - the way the data is presented is unclear but I'm flipping back and forth between last year's projections and this year's):

Thomas Projection: 4
Thomas Actual: 7.3

Bradley Projection: 1.8
Bradley Actual: 3.1

Crowder Projection: 1.9
Crowder Actual: 6.1

Yeah...so if this misses that badly on our top 3 players from last year then I'm not likely to give it much consideration this year.

Compared to what better set of predictions? Saying that e.g. Crowder outperformed his projection is basically just saying that CARMELO isn't a crystal ball. The fact that they had him that high I'd say is a good sign. No system had the Cs putting up 48 wins, and +/- is pretty tightly linked to winning.