The accuracy of NBA drafting

[mdenny]

https://medium.com/@burakcankoc/what-are-the-odds-to-become-an-all-star-for-each-draft-pick-2d113d6b82e5

It's interesting you brought this graph up.

You will note that this data was accumulated during covid and his starting point was 1989. Back when all the nba drafts were far more accurate. When prospects played multiple years in college.

How would this graph look like if you eliminated all the data from the 1990s?

The conclusions of this graph simply don't meet the eye test. Just look at the top 5 players drafted 1 through 5 from 2000 on. It's not EVEN CLOSE to say 46% of them became allstar players.

And we can go through all that if you want. But you should be able to tell just by looking.

Covid? Did Covid infect 30-40 years of data? What evidence to support your claim that drafts were more accurate back in the 1990's? Were GM's smarter or luckier back then? The so-called eye test is clear, you're just trying to spin it.

Even if you accept the 'drafts were more accurate" idea as true, then since Masai is proven to be more accurate in the draft, then it's better to have him draft as high as possible to get the best possible player in the draft.

There is a very simple answer to your question: players were drafted at an older age....after more seasons in college and more evaluation time.

I don't know why you are bringing up covid.

[ConSarnit]

It's interesting you brought this graph up.

You will note that this data was accumulated during covid and his starting point was 1989. Back when all the nba drafts were far more accurate. When prospects played multiple years in college.

How would this graph look like if you eliminated all the data from the 1990s?

The conclusions of this graph simply don't meet the eye test. Just look at the top 5 players drafted 1 through 5 from 2000 on. It's not EVEN CLOSE to say 46% of them became allstar players.

And we can go through all that if you want. But you should be able to tell just by looking.

Covid? Did Covid infect 30-40 years of data? What evidence to support your claim that drafts were more accurate back in the 1990's? Were GM's smarter or luckier back then? The so-called eye test is clear, you're just trying to spin it.

Even if you accept the 'drafts were more accurate" idea as true, then since Masai is proven to be more accurate in the draft, then it's better to have him draft as high as possible to get the best possible player in the draft.

There is a very simple answer to your question: players were drafted at an older age....after more seasons in college and more evaluation time.

I don't know why you are bringing up covid.

Is your contention that teams were better at drafting because they had more information (ie. greater sample size to judge players due to longer college careers)?

[Shakril]

Covid? Did Covid infect 30-40 years of data? What evidence to support your claim that drafts were more accurate back in the 1990's? Were GM's smarter or luckier back then? The so-called eye test is clear, you're just trying to spin it.

Even if you accept the 'drafts were more accurate" idea as true, then since Masai is proven to be more accurate in the draft, then it's better to have him draft as high as possible to get the best possible player in the draft.

There is a very simple answer to your question: players were drafted at an older age....after more seasons in college and more evaluation time.

I don't know why you are bringing up covid.

Is your contention that teams were better at drafting because they had more information (ie. greater sample size to judge players due to longer college careers)?

Why are we still talking about it?

In this very thread it was already proofen, that a higher Draft pick has only a slight better chance at Drafting a Star player - and even smaller when it comes to winning a chip. Dont get distracted by misleading graphs from DelAbbot, he is known for using misleading information and sometimes straight out lying. (like the graph were he uses the All-Star metric instead of actual Star players - an not even differantiate if they made the All-Star with their original team or with a different one.)

If you are willing to ruin a good Team Culture for 1.2% better chance at Drafting a Starplayer, well then your priorities are wrong.

[Thaddy]

I get that the stats don't look great but Masai has two reasonable pathways to making this happen. That's through trades and drafting. He's been hitting his picks late in the draft where the risk is low. Most late picks don't work out and those aren't black marks against a GM.

We should be looking to tank. This means having a fire sale and selling pieces for any draft capital we can get. If we tank and we can get a player like Edgecombe that's a lot easier to develop than the "middle of the 2nd round talent" we are accustom to seeing.

Fernando / Olynyk
Barnes / Mogbo
Barrett / Walter
Dick / Agbaji
IQ / Shead

The line up above should secure a top 8 pick worse case scenario. The draft capital and assets we get back from Brown, Boucher, Poeltl, Mitchell, and others should be enough to get us a solid player down the road to fill in the holes we have.

[raptorforlife88]

It's a good point about earlier drafts being easier to evaluate from a scouting standpoint. I think taking a very quick glance at drafts that took place in the early to late 90s, it does look like there's less mystery about the guys being drafted. Very few complete busts in the top 5 like you've seen in the last two decades.

In 90, Coleman and Payton went 1-2 and the next three guys have various arguments for being and not being in the rest of the top 5 and none are really busts relative to the rest of the draft. Maybe Abdul-Rouf but there were reasons there.

In 91, Dikembe, Steve Smith and Larry Johnson are the top 5 picks who are absolutely top 5 guys in their draft, and then Kenny Anderson is a maybe. Billy Owens is not, and looking at him he's a relative bust.

In 92, Shaq, Alonzo and Laettner are definitely all in the top 5 of their draft. The other two guys are not but also would not be considered busts. They had steady careers.

In 93, you've got Webber and Hardaway for sure and then possibly Mashburn in the top 5. Isiah Rider is the only one you can really consider a "bust".

In 94 you have an argument that all five of the top 5 picks would land in the top 6 when looking back at the draft. None were busts, all had long good careers.

In 95 Garnett and Sheed are for sure in the top 5 and then McDyess and Stackhouse have arguments as well. Joe Smith is the only one outside of it, and he's not really a bust either.

In 96, Ray Allen and Iverson are for sure in the top 5. Camby, Marbury and Abdur Raheem are the other three and are all really good players who would probably not be in the top 5 because this was possibly the best draft of all time.

In 97, there's Billups, Duncan and Van Horn would all still be top 5.

In 98, you'd only have Carter and Jamison, Bibby would land just outside. I think mostly from this point on it gets less and less consistent in the top 5 and you see more and more outright busts.

There's like an 8 year run there where maybe scouting had combined with the usual 2-4 year college career and led to a really consistent drafting record in those top picks. 1989 which is also in that sample does not conform at all to this and was a mess of a draft.

I don't know how much it would skew the average overall but it might skew it a little and I'd be interested to see how it might change the overall number in the sample to instead look at the 2000-2020 period for pick values. Maybe it's still pretty similar.

[rapz**7**]

Double post

[rapz**7**]

I get both sides of the coin on this debate and like others have mentioned there is a lot more context than the premise of the debate however I think teams in the lottery will predominantly have bad GMs. That is usually why they are tanking and over an extended period of years.
Masai and co. are also great at talent evaluation. We can feel pretty comfortable that more often than not they're taking the BPA.
Keeping those two things in mind I don't think we're a franchise that will be repeat offenders for tanking & additionally we have semi talented squad.
That's why I feel a draft pick in the top 10s isn't so bad for us as a franchise. We can be competitive and it's likely as other years have determined the teams drafting high likely won't be making the best picks . We are still very likely to get a player to take us to what we're trying to achieve.

[PhilBlackson]

It’s funny how some posters talk outta both sides of their mouth…

They somehow fully trust our FO to draft a good player later in the draft but when we discuss getting a top 4 pick suddenly they act like it’s all just luck lol make up your f**king minds what you believe.

[ATLTimekeeper]

It’s funny how some posters talk outta both sides of their mouth…

They somehow fully trust our FO to draft a good player later in the draft but when we discuss getting a top 4 pick suddenly they act like it’s all just luck lol make up your f**king minds what you believe.

I think there's a distinction here.

1) A team over a long period of time that is good at drafted is more likely to find NBA players in the draft.
2) Since the fail rate of even the best draft picks is high, there is no confidence that a good drafting team will nail that high draft pick.

We see this frequently, but you know the best way to look at it is to see drafts where the same team drafted multiple players and the later selection ended up much better.

A good example is Poeltl, Siakam and (undrafted Fred). All three are hits. Siakam and VanVleet ended up being much better than the lottery pick. Had we tanked for Poeltl, how would you feel about that process? And imo Poeltl ended up an appropriate value at 8.

Same deal with Jalen Suggs and Franz Wagner. If the result was just Suggs, Orlando would have whiffed that draft. You have to think of it that way. Is it worth it to burn a season and blow the pick? What happens after that?

[raptorforlife88]

It’s funny how some posters talk outta both sides of their mouth…

They somehow fully trust our FO to draft a good player later in the draft but when we discuss getting a top 4 pick suddenly they act like it’s all just luck lol make up your f**king minds what you believe.

Depends on what's being argued. Conflating the two things doesn't make sense for sure if you're saying it's all luck while saying we make good picks late.

But you can make arguments about the relative value of pick slots while noting that our FO has been good with picks regardless of spots and that's perfectly coherent. I think it's interesting to note how those values may have changed the past twenty years relative to years prior (with longer college careers and track records for drafting). If those values have changed than the difference between the top five and next five picks may be smaller (but still have a gap) relative to decades prior.

It might even make you feel better about it if the Raptors drop out of the top five picks. Which there was a decent change of happening if they finished 4th or 5th last as well.

[dohboy_24]

I feel like myself and others have attempted to make a specific point over and over on this board about the accuracy and expected results of top 5 picks in the NBA draft so I want to make a thread here to actually dig into the analysis via the 'redraft' exercise.

Everyone is free to expand the sample set. But from these 5 years.....a top 5 pick ends up being a top 5 player from his draft only 20% of the time. I have looked over all the years....and that approximate figure seems to generally hold true.

Of course it does. The probability that the first 5 picks of the draft will ultimately become the top 5 players in the draft class is as statistically likely as the next set of 5 picks (#6 to #10) doing so.

5 picks out of 30 possible first round selections = 16.67% probability that all of the top 5 picks (in terms of draft order) in a draft year will become the top 5 players (in terms of statistical performance) in their draft class

25 picks out of 30 possible first round selections = 83.33% probability that any one of the remaining 25 picks (in terms of draft order) in a draft year will become a top 5 player (in terms of statistical performance and/or accolades) in their draft class

So QED, all this hype about 'top 5 picks' being 'essential to talent accumulation' is essentially bunk. There is no steep falloff in quality after the first top 5 picks in the nba draft historically. The steep falloff is in the initial hype that the fanbase propagates for the first year or two of the top 5 pick's career.

Given the history....It's just foolishness to be overly upset and melodramatic about 'falling from 4 to 7' or whatever in the draft lottery. Good drafting is FAR more important than draft order.

Yes, there is quite a difference between the #1 and #5 pick in the draft but the real drop-off doesn't occur until after the first 10 picks.

Beyond those top 10 selections the trends in performance tend to flatten out as there isn't as much of a difference between the #11 pick and the #21 pick as there is between those selected among the first 10 picks in the draft.









SOURCE: https://wsb.wharton.upenn.edu/wp-content/uploads/2021/11/2018_Basketball_Sherman_Tanking.pdf

[mdenny]

I feel like myself and others have attempted to make a specific point over and over on this board about the accuracy and expected results of top 5 picks in the NBA draft so I want to make a thread here to actually dig into the analysis via the 'redraft' exercise.

Everyone is free to expand the sample set. But from these 5 years.....a top 5 pick ends up being a top 5 player from his draft only 20% of the time. I have looked over all the years....and that approximate figure seems to generally hold true.

Of course it does. The probability that the first 5 picks of the draft will ultimately become the top 5 players in the draft class is as statistically likely as the next set of 5 picks (#6 to #10) doing so.

5 picks out of 30 possible first round selections = 16.67% probability that all of the top 5 picks (in terms of draft order) in a draft year will become the top 5 players (in terms of statistical performance) in their draft class

25 picks out of 30 possible first round selections = 83.33% probability that any one of the remaining 25 picks (in terms of draft order) in a draft year will become a top 5 player (in terms of statistical performance and/or accolades) in their draft class

So QED, all this hype about 'top 5 picks' being 'essential to talent accumulation' is essentially bunk. There is no steep falloff in quality after the first top 5 picks in the nba draft historically. The steep falloff is in the initial hype that the fanbase propagates for the first year or two of the top 5 pick's career.

Given the history....It's just foolishness to be overly upset and melodramatic about 'falling from 4 to 7' or whatever in the draft lottery. Good drafting is FAR more important than draft order.

Yes, there is quite a difference between the #1 and #5 pick in the draft but the real drop-off doesn't occur until after the first 10 picks.

Beyond those top 10 selections the trends in performance tend to flatten out as there isn't as much of a difference between the #11 pick and the #21 pick as there is between those selected among the first 10 picks in the draft.









SOURCE: https://wsb.wharton.upenn.edu/wp-content/uploads/2021/11/2018_Basketball_Sherman_Tanking.pdf

Great post and good evidence.

So if the REAL major drop-off is after 10....that certainly constitutes a decent argument against all-out tanking.

[mdenny]

Here is a an off-topic point that is also a major factor in the tanking debate. The problem is that it would be exceedingly difficult to measure:

It seems likely that tanking hinders player development....so the draft picks you DO make while tanking don't become the players they would otherwise become.

I recently saw someone say this about Joel embiid's career. Can't rememeber who said it....but they opined something like "perhaps his apathetic approach to the game and winning grew out of the tanking environment he was drafted in?".

Suppose Embiid was drafted onto a team that had a mix of young players and vets that was focused on winning since day one? Would he be a different player? I think that's highly plausible. (Even if not for embiid particularly...then definitely for SOME players).

I suppose there IS a way to measure this. One could attempt to measure the relative success of pick outcomes as a function of the team that drafts them?

What if what we call "good team drafting" is moreso a function of the environment that a pick is brought into then it is "the right selection"?

I'm convinced that a bunch of the failed draft picks for perennially tanking teams would've ended up being completely different players if they had entered the league on a different team. Perhaps the Miami heat are a good example of this? Ie heat culture and all the success stories that their team seemed to have for several years.

There's alot of factors in player development that are hindered by a tanking environment. It could be reasonably argued that all the garbage that comes with a 4 year tank creates psychological habits that are alot harder to break when the team is "supposed" to start winning.

[bballsparkin]

I'm not sure of some of the OP's re-rankings. Cousins wan't a top 5 pick that year? ROY, two times 2 NBA, 4 times all star. Eric Bledsoe or Lance Stephenson goes before him in what world? Victor Oladipo was definitely a top 5 pick. Injuries derailed his career. That's using hindsight. It can happen to any player. 2014 is definitely an interesting one. No one really stands out taken in the first round other than Embiid. Wiggins and Gordon have had fine careers. Andrew Bogut doesn't make the top 5? Channing Frye before Bogut?

Bogut was ROY, and All-NBA player who recieved multiple DPOY votes. Led the league in blocks one season. There's a reason he went first over shorter proven commodities like Chris Paul and D Williams. Unfortunately injuries cut his career short. Even then he was an important player for GSW 2015 championship. Although not as impressive as before injuries. Corey Brewer over Mike Conley? Did I read that right? Heck I wouldn't even take Noah over Mike Conley. Kevin Love is questionable, NOT. Blake Griffin doesn't survive? Steph, Harden, Holiday, DD, and then who if not Blake? Dude was a 5 time All NBA. 4 years straight and came in 3rd once for MVP voting.

But yes, there's no guarantee with drafting carry on.