Hi All,
Latest post is up where I look at how to identify and project a Top 10 draft pick bust at what seems to be a crazily early period in their career.
http://www.zigzaganalytics.com/home/-fi ... catches-on
Identifying NBA Draft Busts
Re: Identifying NBA Draft Busts
The # of busts is really small. If you accept the findings you'd want to take almost anything for them, but you might not have gotten much offered, very often. Probably should have taken them.
Most got better than their summer league performance. Not good to deserve keeping but the improvement likely would have GMs holding, hoping for further improvement or improved trade value.
Mudiay deserves the Exum treatment. Hope has lasted too long.
If you expanded the zone to summer league PER under 15, do you still get high rate of early NBA meh or worse careers or do a notable share of guys bounce back from less bad summer leagues to average or better NBA starts?
What are the correlations for efg%, ts%, assists, rebounding, usage, etc.?
Is there a combo of summer league and college / overseas and / or physical & combine performance marks that does even better?
I think they were all taken by GMs who didn't play in NBA except Rivers by the team employing his Dad. Of course they are the lion share of GMs, but something to consider.
SAC and PHX are the only teams to draft 2 busts or headed to busts. Presti, I think, is only GM to acquire 2 and failed at both attempted rehabs.
It is throwing out several nets at once to catch fish but a large majority of these are real young, foreign born or perhaps not academically gifted. The few others might be afflicted with delusions of their greatness.
Most got better than their summer league performance. Not good to deserve keeping but the improvement likely would have GMs holding, hoping for further improvement or improved trade value.
Mudiay deserves the Exum treatment. Hope has lasted too long.
If you expanded the zone to summer league PER under 15, do you still get high rate of early NBA meh or worse careers or do a notable share of guys bounce back from less bad summer leagues to average or better NBA starts?
What are the correlations for efg%, ts%, assists, rebounding, usage, etc.?
Is there a combo of summer league and college / overseas and / or physical & combine performance marks that does even better?
I think they were all taken by GMs who didn't play in NBA except Rivers by the team employing his Dad. Of course they are the lion share of GMs, but something to consider.
SAC and PHX are the only teams to draft 2 busts or headed to busts. Presti, I think, is only GM to acquire 2 and failed at both attempted rehabs.
It is throwing out several nets at once to catch fish but a large majority of these are real young, foreign born or perhaps not academically gifted. The few others might be afflicted with delusions of their greatness.
Re: Identifying NBA Draft Busts
"Most got better than their summer league performance. Not good to deserve keeping but the improvement likely would have GMs holding, hoping for further improvement or improved trade value." -
It's almost worse if they have a gradual improvement and are just outside the bust zone. GM's like you said would probably hold on in hope and 3-4 years later they have an asset that is worthless and not contributing. You would rather them in this bust zone so you can make the call, IF you believed the bust zone was correct.
If you expanded the zone to summer league PER under 15, do you still get high rate of early NBA meh or worse careers or do a notable share of guys bounce back from less bad summer leagues to average or better NBA starts?
A summer league PER of 15 is actually right on the average for a top 10 pick, so a little 2 high. Guys around the 15 PER in summer league.....Ben Simmons, Towns, Jabari Parker.
If i extend the bust zone from PER of 11 to say a max of 13. We get McLemore and Waiters (which i already mentioned as being just outside bust zone), above them but still under 13 PER in Summer League are Buddy Hield, Terrance Ross. Ross has a 3 year PER of 11.2, so there is a bust. Hield in his first season is at 11.1 which is borderline.
DeMarcus Cousins had a 12.9 Summer league PER, so there is the first All-star.
PER of 13 in summer league according to the model translates to just about a 3 year PER avg of 13. DeRozan had a 3 year PER of 13.3, CJ McCollum had a 13.3, Thomas Robinson 13.4, Dion Waiters 12.9. From the data i have from 2009 a 3 year PER of 13 becomes a 50/50 chance a player will go onto become a perennial All-star or bust. Whereas under 12 and most definitely under 10 there isn't anything close to a perennial all-stars
What are the correlations for efg%, ts%, assists, rebounding, usage, etc.?
I'll take a look at ts% and get back to you
It's almost worse if they have a gradual improvement and are just outside the bust zone. GM's like you said would probably hold on in hope and 3-4 years later they have an asset that is worthless and not contributing. You would rather them in this bust zone so you can make the call, IF you believed the bust zone was correct.
If you expanded the zone to summer league PER under 15, do you still get high rate of early NBA meh or worse careers or do a notable share of guys bounce back from less bad summer leagues to average or better NBA starts?
A summer league PER of 15 is actually right on the average for a top 10 pick, so a little 2 high. Guys around the 15 PER in summer league.....Ben Simmons, Towns, Jabari Parker.
If i extend the bust zone from PER of 11 to say a max of 13. We get McLemore and Waiters (which i already mentioned as being just outside bust zone), above them but still under 13 PER in Summer League are Buddy Hield, Terrance Ross. Ross has a 3 year PER of 11.2, so there is a bust. Hield in his first season is at 11.1 which is borderline.
DeMarcus Cousins had a 12.9 Summer league PER, so there is the first All-star.
PER of 13 in summer league according to the model translates to just about a 3 year PER avg of 13. DeRozan had a 3 year PER of 13.3, CJ McCollum had a 13.3, Thomas Robinson 13.4, Dion Waiters 12.9. From the data i have from 2009 a 3 year PER of 13 becomes a 50/50 chance a player will go onto become a perennial All-star or bust. Whereas under 12 and most definitely under 10 there isn't anything close to a perennial all-stars
What are the correlations for efg%, ts%, assists, rebounding, usage, etc.?
I'll take a look at ts% and get back to you
Re: Identifying NBA Draft Busts
If we look at the remaining 4 lottery pick positions (picks 11-14) with data from 2009...keeping in mind that the later the pick in the draft the lower the expectations a front office is going to have on that pick. Picks 11-14 you are not expecting an all-star player, more hoping. A solid starting 5 player is more realistic.
Players with a summer league PER of > 11.........
Kendal Marshall (PHX 2012)...Summer league PER of 9.8..........3 year avg 10.9.........bust.
Shabazz Muhammad (MIN 2013)....Summer league PER of 6...........3 year avg 16.1.........Not a starting 5 player, but not horrible. Getting an average player here, although his past 2 season have seen his PER drop each year. Lets call him a rotational player. For pick 14, you are pretty much breaking even here.
George Papagiannis (Sac 2016).....Summer league PER of 1.1....1.1!!!!! This makes him the illustrious winner of the worst lottery pick summer league player since 2009 that played enough minutes to qualify. Model says he is a lock to be a bust.
Players with a summer league PER of > 11.........
Kendal Marshall (PHX 2012)...Summer league PER of 9.8..........3 year avg 10.9.........bust.
Shabazz Muhammad (MIN 2013)....Summer league PER of 6...........3 year avg 16.1.........Not a starting 5 player, but not horrible. Getting an average player here, although his past 2 season have seen his PER drop each year. Lets call him a rotational player. For pick 14, you are pretty much breaking even here.
George Papagiannis (Sac 2016).....Summer league PER of 1.1....1.1!!!!! This makes him the illustrious winner of the worst lottery pick summer league player since 2009 that played enough minutes to qualify. Model says he is a lock to be a bust.
Re: Identifying NBA Draft Busts
Probably almost none to none at all GMs who would make a big decision based on summer league even with knowing the bust analysis. But what about first half of rookie season? Is the evidence similar, stronger or less clear?
Re: Identifying NBA Draft Busts
Including age and using the metric DRE (which can be done with boxscore stats) could be interesting.
Re: Identifying NBA Draft Busts
The reason I looked at summer league was not only because of the opportunity to trade early and get as much value as possible if you thought they were a bust, but also because it puts the top 10 draft picks typically in a position where they are the go-to player on the summer league team or at least a lot more than they would get as a limited minutes /role playing rookie in the regular season. Summer league gives you a glimpse of how they potentially play if they had a higher usage/"main man" role.
First half of a rookie season, you are gonna get a lot of very limited minutes in many cases, which I know many metrics try and allow for but it's very hard to get a good indication of a rookie if they are getting junk time minutes for instance.
Something I can also take a look at as well (first half of season) as well as the true shooting/effective fg % and see if that translates from summer league.
Age and metric dre...again I'll put it on my list, interesting. Got another post i'm working on now, so will try and find some time.
First half of a rookie season, you are gonna get a lot of very limited minutes in many cases, which I know many metrics try and allow for but it's very hard to get a good indication of a rookie if they are getting junk time minutes for instance.
Something I can also take a look at as well (first half of season) as well as the true shooting/effective fg % and see if that translates from summer league.
Age and metric dre...again I'll put it on my list, interesting. Got another post i'm working on now, so will try and find some time.