1st rd draft selection / performance by VORP to date

Home for all your discussion of basketball statistical analysis.
Post Reply
Crow
Posts: 5690
Joined: Thu Apr 14, 2011 11:10 pm

1st rd draft selection / performance by VORP to date

Post by Crow » Tue Sep 18, 2018 5:11 pm

For 2017 first round the correlation between draft pick with VORP to date (with correlation score reversed to recognize "higher"draft picks have a lower draft number) was -.085. For 2016 draft picks .27. For 2015 picks, .25. For 2014, -.30. For 2013, .05. For 2012, .48. For 2011, .05. For 2010, .56. For 2009, .43. For 2008, .21.

That doesn't seem impressive for the insiders. At all.
If you weight the correlation by the number of years of performance data, the average correlation is 0.25.

By that seemingly low .25 standard, the insiders have had 3 much better than average years, 3 near average and 4 more awful than usual draft classes.

I'd think I could beat that. Perhaps by a lot. May have to test that. I've been content in past to make a handful of draft calls. I think I've done pretty well with that. But I'll reflect on that as best I can later.

Crow
Posts: 5690
Joined: Thu Apr 14, 2011 11:10 pm

Re: Team 1st round draft performance by VORP to date

Post by Crow » Tue Sep 18, 2018 5:41 pm

Let's see how this assembled in 10 minutes ranking does:

Rk Pk Tm Player
1 4 MEM Jaren Jackson
2 3 ATL Luka Doncic
3 7 CHI Wendell Carter
4 1 PHO Deandre Ayton
5 10 PHI Mikal Bridges
6 11 CHO Shai Gilgeous-Alexander
7 2 SAC Marvin Bagley
8 5 DAL Trae Young
9 8 CLE Collin Sexton
10 9 NYK Kevin Knox
11 12 LAC Miles Bridges
12 16 PHO Zhaire Smith
13 6 ORL Mohamed Bamba
14 26 PHI Landry Shamet
15 30 ATL Omari Spellman
16 14 DEN Michael Porter
17 23 IND Aaron Holiday
18 19 ATL Kevin Huerter
19 15 WAS Troy Brown
20 17 MIL Donte DiVincenzo
21 13 LAC Jerome Robinson
22 18 SAS Lonnie Walker
23 25 LAL Moritz Wagner
24 20 MIN Josh Okogie
25 27 BOS Robert Williams
26 29 BRK Dzanan Musa
27 21 UTA Grayson Allen
28 22 CHI Chandler Hutchison
29 28 GSW Jacob Evans
30 24 POR Anfernee Simons

Not quite the same as doing this in June but there is no information that I've gained from the summer that I consider an especially valuable or unfair edge. One could protest that the draft pick # is an unfair advantage but the actual pick number isn't much different that what was available to everybody before the draft in the form of projections / expectations of picks and I don't consider the insider's actual picks that reliable as evidenced above. To the extent that those consensus projected and actual picks influenced me they give me no advantage against the actual picks. I didn't pick the same as the insiders in any case and average difference was 4.3. If I spent more time on this it would probably be more different and probably better.

I could try a 10 hour, 100 hour or 1,000 hour ranking. But I'll start with the 10 minute version.

eminence
Posts: 97
Joined: Sun Sep 10, 2017 8:20 pm

Re: Team 1st round draft performance by VORP to date

Post by eminence » Tue Sep 18, 2018 6:39 pm

Yeah, from my looks draft performance has been very meh for a very long time, and hasn't made any discernible improvements at any point in time.

eminence
Posts: 97
Joined: Sun Sep 10, 2017 8:20 pm

Re: Team 1st round draft performance by VORP to date

Post by eminence » Tue Sep 18, 2018 7:05 pm

Image

It's not a perfect measure by any means, but that's Allstar draft position vs draft class and kind of supports the point made.

Crow
Posts: 5690
Joined: Thu Apr 14, 2011 11:10 pm

Re: Team 1st round draft performance by VORP to date

Post by Crow » Tue Sep 18, 2018 9:04 pm

Even this seemingly lousy level of pick to performance correlation using VORP is aided by the tendency to play 1st rounders more {maybe more than they deserve) and to give them more opportunities to have impact.

No one in last 10 years has been allowed to accumulate much negative VORP (Mudiay may be worst at -3 total) but low to meh VORP is not sufficiently suppressed or replaced with alternatives.

Crow
Posts: 5690
Joined: Thu Apr 14, 2011 11:10 pm

Re: Team 1st round draft performance by VORP to date

Post by Crow » Tue Sep 18, 2018 9:25 pm

Last season 18 players had a a negative VORP while playing 1500 plus minutes. Only 3 played in playoffs, none won a series. Jamal Crawford, Bryn Forbes and Andrew Wiggins.

Plus 1 was about the median for these main rotation players.

Crow
Posts: 5690
Joined: Thu Apr 14, 2011 11:10 pm

Re: Team 1st round draft performance by VORP to date

Post by Crow » Sun Dec 02, 2018 7:48 pm

Crow wrote:
Tue Sep 18, 2018 5:41 pm
Let's see how this assembled in 10 minutes ranking does:

Rk Pk Tm Player
1 4 MEM Jaren Jackson
2 3 ATL Luka Doncic
3 7 CHI Wendell Carter
4 1 PHO Deandre Ayton
5 10 PHI Mikal Bridges
6 11 CHO Shai Gilgeous-Alexander
7 2 SAC Marvin Bagley
8 5 DAL Trae Young
9 8 CLE Collin Sexton
10 9 NYK Kevin Knox
11 12 LAC Miles Bridges
12 16 PHO Zhaire Smith
13 6 ORL Mohamed Bamba
14 26 PHI Landry Shamet
15 30 ATL Omari Spellman
16 14 DEN Michael Porter
17 23 IND Aaron Holiday
18 19 ATL Kevin Huerter
19 15 WAS Troy Brown
20 17 MIL Donte DiVincenzo
21 13 LAC Jerome Robinson
22 18 SAS Lonnie Walker
23 25 LAL Moritz Wagner
24 20 MIN Josh Okogie
25 27 BOS Robert Williams
26 29 BRK Dzanan Musa
27 21 UTA Grayson Allen
28 22 CHI Chandler Hutchison
29 28 GSW Jacob Evans
30 24 POR Anfernee Simons

Not quite the same as doing this in June but there is no information that I've gained from the summer that I consider an especially valuable or unfair edge. One could protest that the draft pick # is an unfair advantage but the actual pick number isn't much different that what was available to everybody before the draft in the form of projections / expectations of picks and I don't consider the insider's actual picks that reliable as evidenced above. To the extent that those consensus projected and actual picks influenced me they give me no advantage against the actual picks. I didn't pick the same as the insiders in any case and average difference was 4.3. If I spent more time on this it would probably be more different and probably better.

I could try a 10 hour, 100 hour or 1,000 hour ranking. But I'll start with the 10 minute version.

Status check Dec 2: Draft pick correlation to BPM, 0.39. My quick list, 0.56. 42% higher correlation.

Any draft junkies want to make same comparison? Fairly similar relative numbers on VORP (I beat pick #s by teams by 49%) but lower on each. Removing guys who haven't played doesn't change things.

How much is 49% better early performance to draft pick correlation worth to your team?

I should have been harsher on a few guys I didn't believe in.

Crow
Posts: 5690
Joined: Thu Apr 14, 2011 11:10 pm

Re: Recent 1st round draft performance by VORP to date

Post by Crow » Sun Dec 02, 2018 10:15 pm

Looks like my season date correlation is higher than for all but one draft pick order in last 11 years and about twice the average correlation. Good year for GMs relative to their average.

eminence
Posts: 97
Joined: Sun Sep 10, 2017 8:20 pm

Re: Recent 1st round draft performance by VORP to date

Post by eminence » Mon Dec 03, 2018 3:40 am

Here's how I had the first rounders ranked heading into the draft (had several 2nd rounders and an undrafted guy in my top 30, but leaving them out for now).

1. Luka
2. JJJ
3. Young
4. WCJ
5. Mikal
(Clark)
6. Bamba
7. Bagley
8. Porter (this is one I imagine insider info helps a ton with, at the time I submitted my final rankings I was under the impression he was on track to play this season)
9. Ayton
10. Shamet
(Okobo)
11. Evans
12. Brunson
(Edwards)
(J. Carter)
13. Walker IV
14. SGA
(Bates-Diop)
15. Huerter
16. Okogie
17. Zhaire
(Melton)
18. T. Brown
19. Miles
20. Sexton
(D. Hall)
(Milton)
21. Wagner
(Graham)

Then the 9 first rounders I didn't have in the top 30:
Williams
Musa
Knox
DiVincenzo
Hutchison
Holiday
Allen
Simons
Robinson

If you wouldn't mind running that through Crow, I'd love to see how I'm doing. Obviously very early, and some of the lower picks helped by simply not playing so far.

Crow
Posts: 5690
Joined: Thu Apr 14, 2011 11:10 pm

Re: Recent 1st round draft performance by VORP to date

Post by Crow » Mon Dec 03, 2018 5:10 am

I can't exactly compare your projections to the GMs top 30 or mine because of who is included and excluded, but comparing it to the draft pick #s including the 2nd rounders and Clark listed at 61, it appears your early performance to draft pick correlation is about 0.372. Not apples to apples but on surface about 11% than the GM's top 30 correlation.

... I guess I can run GM correlation for guys on your list...
you are still ahead of the GMs by 6% that way.

By GMs I mean th aggregate of them each making only one pick. Some unknown number of their big boards might be better than this average.

Crow
Posts: 5690
Joined: Thu Apr 14, 2011 11:10 pm

Re: 1st rd draft selection / performance by VORP to date

Post by Crow » Thu Jan 17, 2019 5:29 pm

My quick preseason draft rankings over GM draft picks on correlation of ranking / pick to VORP lead has grown to 66%. Both lost some correlation over early December but the GM picks lost way more. .45 correlation for me, just .27 for them.

The Stepien had a sizeable number of guys drafted not in their site's 2018 rankings and vice versa so it is a different list for correlation. It also had a far higher correlation to performance to date than GMs. Mine is slightly higher.

I could do this in 2019 with more rigor if some team wanted another opinion.

Post Reply