The Official 2023 Draft Thread

[DCZards]

I see 9/15 years in which 9/16/28 was better than #2.

Really?



I see the #2 pick better than 9/16/28 in:
2019
2017
2016
2015
2013
2007
2006
2005

The 9/16/28 are better in:
2014
2012
2011
2010
2009
2008

2018 is a toss up.

So that's 8 to 6 in favor of the #2 pick (which, again is a fluke for how bad it was relative to #1 and #3).
It's an 10-3 win for the #3 pick, with 2006 and 2007 being toss ups. So on average a top 2 or 3 pick beats a basket of later picks about 2/3rds of the time. And the top picks are the only ones that give you a realistic chance at a franchise player.

Not to mention that the 2 & 3 picks have yielded all-NBA and MVP level players like Embiid, KD, Luka, Tatum, Harden and Aldridge, while the 45 picks at 9/16/28 have yielded only one such player--DeRozan.

In a star-driven league like the NBA, the Zards would be smart to keep a top 3 pick if they're lucky enough to get it.

[Kanyewest]

I see 9/15 years in which 9/16/28 was better than #2.

Really?



I see the #2 pick better than 9/16/28 in:
2019
2017
2016
2015
2013
2007
2006
2005

The 9/16/28 are better in:
2014
2012
2011
2010
2009
2008

2018 is a toss up.

So that's 8 to 6 in favor of the #2 pick (which, again is a fluke for how bad it was relative to #1 and #3).
It's an 10-3 win for the #3 pick, with 2006 and 2007 being toss ups. So on average a top 2 or 3 pick beats a basket of later picks about 2/3rds of the time. And the top picks are the only ones that give you a realistic chance at a franchise player.

Good chart btw.

Some bad GMs are really skewing the potential value of these #2 picks. In 2005, the Hawks could have selected CP3 or Derron Williams over Marvin Williams. In 2009, Thabeet over Harden, Curry, or even Evans. In 2012, the Hornets could have selected Beal (although I suppose were arguing for MKG at the time). Bagley over Doncic was jaw dropping even at the time.

IMO, it's pretty close to 50/50 when the 9/16/28 is better than #2. But if just feels like that the #2 pick when it hits is more likely to result in championship contending rosters.

The #9 pick really hasn't done that since 2002 with Amare. (although I suppose landing Demar/Poetl, both #9 picks did ultimately lead the Raptors into trading for Kawhi).

[nate33]

No Nate, you can’t just say that without doing the same analysis of pick #1, which is significantly worse results than #3.

The #1 pick would have been outstanding if not for the unfortunate injuries. Injuries robbed 3 guys of perennial All-NBA status (Zion, Rose and Oden). And injuries have played some role in the demise of Simmons as well. But injuries notwithstanding, of the 15 guys drafted, there are 8 All Stars (Zion, Simmons, KAT, Wiggins, Davis, Kyrie, Griffin, Rose), 4 multi-time All-NBA guys (KAT, Davis, Kyrie, Griffin) plus 3 more guys that made All-NBA one time (Simmons, Wall, Rose) plus Zion still has a real good chance. They have 1 MVP (Rose).

2019 Zion Williamson
2018 DeAndre Ayton
2017 Markelle Fultz
2016 Ben Simmons
2015 Karl-Anthony Towns
2014 Andrew Wiggins
2013 Anthony Bennett
2012 Anthony Davis
2011 Kyrie Irving
2010 John Wall
2009 Blake Griffin
2008 Derrick Rose
2007 Greg Oden
2006 Andrea Bargnani
2005 Andrew Bogut

By the criteria I used, the #1 pick looks about as good as the #3. They have the same number of All-Stars and the same number of multi-time All-NBA guys, though 1 less MVP. (In reality, the #3 is better because their best players were good for longer, like Harden versus Rose or Horford versus Simmons).

But I would also disagree, Poeltl has been much better than Ingram. Rozier much better than Russell. That’s 8 in my (biased) view.

Ingram is a $30M player and an All-Star. Poeltl is a serviceable starting center who would get played off the floor in a playoff game. It's not even close. No team would trade Ingram for Poeltl, and every team would trade Poeltl for Ingram. I'm sorry, but it's just absurd to call Poeltl the better player.

Likewise, Russell was an All-Star and best player on his team for a season, and had several more seasons as a top 2 player on his team. Rozier was never more than a 3rd/4th option role player. Objectively, at least when looking at salaries, the league has regarded Russell higher. Ranking Rozier higher is defensible, but I'd say it's a minority view. I was trying hard to be objective with my criteria.

[NatP4]

I certainly would not use the same criteria. It’s a discussion of value/on court impact, not how many guys squeezed into an all star game.

If your criteria classifies Ben Simmons as a “hit”, then the discussion ends here. That same draft that produced Simmons and Ingram 1&2, you have Jamal Murray at 7, Poeltl at 9, Sabonis at 11, Siakam at 27, Dejounte Murray at 29.

And to reiterate, Poeltl has been significantly better than Ingram in his career. It’s not even close. Really not even comparable. Your logic centers around salary/trade value which does not always (or even oftentimes) line up with actual on court value. Just wait for the Kuzma extension.

[NatP4]

There’s really 6 “hits” out of 15 on that list. Maybe 7 I guess if you wanna include Derrick Rose. I suppose that’s fair as he was a MVP player for a few years before the injuries.

7/15 at #1

Durant, Aldridge, Morant at #2, that’s 3/15, and we will exclude 2020 with James Wiseman.

That’s 10/30 players at that 1/2 picks over a 15 year span that were really hits. And of course we will argue about what constitutes a success at those slots, but you know that an all star appearance should not be the criteria.

IMO your chart shows that it’s pretty close to 50/50 on 2 Vs 9/16/28, but if you add 2020+2021 which are both wins for 9/16/28 IMO, you can definitely conclude that it’s 50/50. Which is not evidence for making the trade, but simply evidence that it’s close.

I make the trade because of my opinion on specific players available at those slots in this draft 2023. But to conclude again, #9&#28 are uniquely bad relative to the picks surrounding them in that range. #3 is uniquely great. I think it would be useful to expand the sample to the picks surrounding those picks.

But trading 2 for specifically 9/16/28 based on the results over the last 15 years, on average, would be a bad idea.

[nate33]

I certainly would not use the same criteria. It’s a discussion of value/on court impact, not how many guys squeezed into an all star game.

If your criteria classifies Ben Simmons as a “hit”, then the discussion ends here. That same draft that produced Simmons and Ingram 1&2, you have Jamal Murray at 7, Poeltl at 9, Sabonis at 11, Siakam at 27, Dejounte Murray at 29.

And to reiterate, Poeltl has been significantly better than Ingram in his career. It’s not even close. Really not even comparable. Your logic centers around salary/trade value which does not always (or even oftentimes) line up with actual on court value. Just wait for the Kuzma extension.

Go propose an Ingram for Poeltl trade on the Trade Board and gauge the results. I don't know how you can even argue this. Guys like Poeltl (defensive rim protectors who can't score and can't switch) just aren't that hard to find. He's basically Daniel Gafford (who only cost us Troy Brown). A couple of years ago, we plucked Robin Lopez off the scrap heap to give us similar production. I know you like his on/off numbers, but he has done nothing in his career. As a starter, he hasn't even led his team to 37 wins.

As I said, Rozier over Russell is defensible, but it's close.

Simmons was actually a really good player for a few years. It's easy to forget that given his recent play. You don't make 3 consecutive all star games and an All-NBA team if you suck. Simmons' teams won 50+ games in each of those 3 seasons (adjusting for shortened schedule). And made the 2nd round twice. Of those other guys from the 2016 draft that you mentioned, none of them have done squat except Murray. And we're talking about Simmons versus the entire field of the 2016 draft.

[payitforward]

I see 9/15 years in which 9/16/28 was better than #2.

That is without the “god-mode GM/hindsight” picks.

Again, #3 has been uniquely great slot over the 15 years compared to 1/2/4/5.

And of course, picking higher is better, but if you did this same exercise with #1&#2 instead of #2 and #3, and #27 instead of #28, and #11 instead of #9, it’s drastically different.

And this is why I always viewed this as a nonsensical exercise and complete waste of time.

What's the benefit of drafting #1 or #2? You have nearly every draftable player at your disposal!

Who the hell is pigeon holed into one specific player? I can't speak for Memphis drafting Thabeet at #2 & Mayo #3. Or Cleveland deciding on Bennett at #1. Its an imperfect science and there are alot of bad GMs but historically your star quality players, your 1st/2nd option types are going to be found higher in the draft. You can find quality role players later but that's what they usually are, low usage role players. Of course there ocassional diamonds in the rough who usually are undervalued due to being outside the box (Jokic/Giannis) or guys who's work ethic was unparalleled (Jimmy Butler/Middleton) but that's the exception, not the rule.

Sigh...
As I've demonstrated pretty clearly, in about 1/2 the drafts from 2008 to 2020 the very best player on the board at #2 (not the guy taken then but the guy who should have been taken then) has produced more value than the 3 very best players on the board at 9,16 & 28 (again, not the guys taken at those spots but the 3 best players still on the board at 9, 16, & 28) about 1/2 of the time. In 2008, for example.

The other half of the drafts, the opposite was true. E.g. in 2011 the best player in the whole draft was available at # 2 -- & he was still on the board at #9. Obviously, you'd rather have Kawhi plus the guys whom, being a genius, you'd pick at 16 & 28 than Kawhi without those guys.

Thus, even if you are a genius, you only do better in 1/2 the drafts. If you are short of being a genius, you can't expect to do that well. Since, yeah, you are definitely short of being a genius (whoever the word "you" is referring to), no more need be said about this idea.

Does this mean you should never trade down? Of course it doesn't mean that! E.g. in most drafts, for example, one would trade the #2 pick for the #3 pick plus the #4 pick. So, whether you do or don't trade down is a function of the particular opportunity in the particular draft -- what you have to give with what set of players on the board in return for what you get.

Wherever you pick, chance is involved. That doesn't mean the draft is 100% chance -- purely "a crapshoot" -- only that there is no way to eliminate entirely the effect of chance on the draft & draft picks. Period.

The best weapon against the effect of chance is multiple tries. Obviously. & that is one of the motivations for trading down. It's a good one. Which doesn't mean that it always produces a benefit.

Of course, another motivation might be that you are oh so smart. You know better than the other GMs. You know that a few guys -- guys you can name -- are going to be available further down (at the spots you want to trade down to) & will be better than any single player available at the spot you want to trade out of. Because you are that much smarter than everyone else.

Since that's a billshut motivation, because no you are not smarter than everyone else, trading down for that reason will shorten your career as a GM.

Of course, you might still be right on one occasion or another. But that doesn't make you a genius.

[Jay81]

old article but...

https://nbadraftroom.com/thoughts-on-the-2022-mcdonalds-all-american-game/

[payitforward]

Overall, it's easy to make an incontrovertible argument for trading down from absolutely any pick after the first 3 (on the assumption that the first 3 picks will often include at least 2 of the best players in the draft).

1. In no draft in NBA history, have picks 4-30 been made in an order that closely mirrors the order of how good those 27 players turn out to be. Never. In fact, it's not usually close. (Feel free to pause here & take a look at any actual draft to confirm.)

2. Thus, it is virtually certain that some number of players will be available in the range I might trade down to who turn out to be better than most of the guys I am considering taking if I don't make the trade.

This is incontrovertible, btw. Just look at any draft -- literally any one of the dozens that have taken place. Moreover, this is true for 2 closely related reasons that can never change:

a. The guys in the draft are not fully developed.
There's no reliable way to know how good they'll be -- you can be right sometimes, but you are certain to be wrong much more often.

b. You are not the great genius who knows better than everyone else.
Over time, your picks at various positions will tend to exhibit quite closely the inherent overall level of accuracy vs. inaccuracy.

Overall, the above demonstrates that chance is an inherent & significant factor in how well a draft pick turns out & how well you pick'em. Uncertainty can't be eliminated. Acting as if it were insignificant is stupid.

The one demonstrated way to lessen the influence of chance factors on success in any situation with inherent uncertainty, is to have more opportunities, to give yourself more than one shot: two shots give you a better chance of making a shot than one. QED.

[payitforward]

Here are a couple of mistaken objections to the above:

I. "Having 2 picks doesn't guarantee success either."
Easily countered: there are no guarantees, but a single success is more likely with 2 picks than 1.

II. "If you're right, then the same logic would have a team trade those 2 picks for 4! &, therefore, those 4 for 8!! & so on! "
Again, that's easily countered. There are reality constraints: e.g. you may not be able to afford the roster room for 4 rookies. Or for 2 rookies. Moreover, beware of conventional thinking, because you may have the mistaken idea that you don't have roster room for 4 rookies -- or 2.

[DCZards]

Overall, it's easy to make an incontrovertible argument for trading down from absolutely any pick after the first 3 (on the assumption that the first 3 picks will often include at least 2 of the best players in the draft).

1. In no draft in NBA history, have picks 4-30 been made in an order that closely mirrors the order of how good those 27 players turn out to be. Never. In fact, it's not usually close. (Feel free to pause here & take a look at any actual draft to confirm.)

2. Thus, it is virtually certain that some number of players will be available in the range I might trade down to who turn out to be better than most of the guys I am considering taking if I don't make the trade.

This is incontrovertible, btw. Just look at any draft -- literally any one of the dozens that have taken place. Moreover, this is true for 2 closely related reasons that can never change:

a. The guys in the draft are not fully developed.
There's no reliable way to know how good they'll be -- you can be right sometimes, but you are certain to be wrong much more often.

b. You are not the great genius who knows better than everyone else.
Over time, your picks at various positions will tend to exhibit quite closely the inherent overall level of accuracy vs. inaccuracy.

Overall, the above demonstrates that chance is an inherent & significant factor in how well a draft pick turns out & how well you pick'em. Uncertainty can't be eliminated. Acting as if it were insignificant is stupid.

The one demonstrated way to lessen the influence of chance factors on success in any situation with inherent uncertainty, is to have more opportunities, to give yourself more than one shot: two shots give you a better chance of making a shot than one. QED.

To quote my man PIF: “I’m not sure what these recent posts are about.”

Or how they differ from dozens I’ve already read.

[NatP4]

Here are a couple of mistaken objections to the above:

I. "Having 2 picks doesn't guarantee success either."
Easily countered: there are no guarantees, but a single success is more likely with 2 picks than 1.

II. "If you're right, then the same logic would have a team trade those 2 picks for 4! &, therefore, those 4 for 8!! & so on! "
Again, that's easily countered. There are reality constraints: e.g. you may not be able to afford the roster room for 4 rookies. Or for 2 rookies. Moreover, beware of conventional thinking, because you may have the mistaken idea that you don't have roster room for 4 rookies -- or 2.

All of what you are saying is obvious, but I don’t understand why you would conclude that pick 2 for 9/16/28 is a terrible idea.

I appreciate Nate throwing together the chart, but it was extremely flukey with 28 being the worst pick in the draft over a 15 year span, and 9 being the next worst pick in that span. Again, if you use 11/13/27, all of them have produced better results than the top overall pick.

For example: if I then have 9/16/28, should I trade back from 9 to 11 while also moving up from 28 to 27? Then is it a good idea?

Now I am more likely to draft a Sabonis/Myles Turner/SGA/Gobert/Siakam?

Wouldn’t someone need to come up with an evaluation for every single pick in the sample period, come up with some kind of score and average out the pick scores within the range of the discussed picks?. Once you do that, it’s going to be obvious that a pick in 9ish range, 16ish range, 28ish range, all added together, is worth a significantly higher score than #2, no?

[payitforward]

But but but... Zards... repetition is the key to learning!!

Not to mention... it's the off season! & we don't even know who they're interviewing (if anyone!) to run the joint.

Here's what I'm thinking, Zards... if you'd just say that you agree, that you take my point. really... I feel sure that would really help a lot!

Not just Zards, btw -- the rest of you too. C'mon now -- you can do it.

I can't remember, btw, whether I've ever told the story of my attempt to become the GM of the Warriors... almost 50 years ago.

[payitforward]

... I don’t understand why you would conclude that pick 2 for 9/16/28 is a terrible idea....

I didn't conclude that it was a terrible idea.
What I concluded was that if you compared what a backward look tells us was the bpa at 2 with the same data for 9/16/28, then assuming those optimal picks are made at all 4 spots, it looked like the trade was a good one in about half the drafts between '08 & '20. I.e. in half of the drafts the best guys who were on the board at 9/16/28 have turned out to be worth more than the best guy who was on the board at 2.

In other words, with perfect knowledge of who was going to be on the board at 9/16/28, & with perfect knowledge of how good all the players in the draft were going to turn out to be, a GM would choose to make the trade about half the time.

But that perfect knowledge is unavailable on draft day!

If instead I skew things by comparing the guys who *actually* went at #2 in each of those drafts against the hindsight-enabled knowledge of the best players on the board that same year at 9, 11, & 28 -- why then of course I'm going to do better, a whole lot better, most years by making the trade! But, that's obviously an unfair comparison.

...if you use 11/13/27, all of them have produced better results than the top overall pick.

For example: if I then have 9/16/28, should I trade back from 9 to 11 while also moving up from 28 to 27? Then is it a good idea?...

There are no magic numbers in the draft. If, over a decade, 11 happens to get you better guys than 9, the fact means nothing. There will be another pair of numbers with the opposite result -- where 1 pick lower happens to get better players over a stretch of a few years. There's nothing to learn there.

Main point: if you compare a hindsight-enabled view with the results from the non-hindsight-enabled view available on draft day, you're going to get great results that have no real meaning. Two of the very best players in the game went at 15 & 30 in 2011. Your trade would have made it possible to take them at 11 & 27, while grabbing Tobias Harris at 13.

...Wouldn’t someone need to come up with an evaluation for every single pick in the sample period, come up with some kind of score and average out the pick scores within the range of the discussed picks?. Once you do that, it’s going to be obvious that a pick in 9ish range, 16ish range, 28ish range, all added together, is worth a significantly higher score than #2, no?

I think that's essentially what Pelton did to get the following results: http://nbasense.com/draft-pick-trade-value/2/kevin-pelton-2.

By the way, based on his pick-values, it looks like 2 wouldn't get you 9, 16 & 28. Not that his values are conclusive: far from it.

[DCZards]

Here's what I'm thinking, Zards... if you'd just say that you agree, that you take my point. really... I feel sure that would really help a lot!

PIF, if you're asking me to agree with the following...I do. I can't see trading out of a top 3 pick (unless it's maybe for a 4 or 5 pick) given the evidence that the best players (All-NBA types and MVPs) are typically at the top of the draft.

Overall, it's easy to make an incontrovertible argument for trading down from absolutely any pick after the first 3 (on the assumption that the first 3 picks will often include at least 2 of the best players in the draft).

[payitforward]

There you go!

Now, good buddy, who do you think we should be targeting for the GM job?

As I mentioned in that thread, I recall that back in '19 you were very positive on Troy Weaver. I was too....

[nate33]

I appreciate Nate throwing together the chart, but it was extremely flukey with 28 being the worst pick in the draft over a 15 year span, and 9 being the next worst pick in that span. Again, if you use 11/13/27, all of them have produced better results than the top overall pick.

The #9 pick was literally THE BEST pick in the draft relative to it's slot. It's flukey alright. But flukey in the other direction. It outperformed picks #4, #6, #7 and #8 despite being in the portion of the draft where the dropoff for each pick is precipitous.

Chart of the production of each pick since 1980 with win shares as the metric:



Utilizing All-Star appearances as the metric:



Data is from here:

https://thedatajocks.com/nba-draft-pick-values/

Note, this article was from 2021 so they probably ignored the couple of drafts just prior to that, but a handful of recent draft picks wouldn't seriously alter a chart based on data from the last 40 years.

[payitforward]

PIF, if you're asking me to agree with the following...I do. I can't see trading out of a top 3 pick (unless it's maybe for a 4 or 5 pick) given the evidence that the best players (All-NBA types and MVPs) are typically at the top of the draft.

Overall, it's easy to make an incontrovertible argument for trading down from absolutely any pick after the first 3 (on the assumption that the first 3 picks will often include at least 2 of the best players in the draft).

Off the top of my head, I can't recall anyone doing it, can you?

Naturally, it's possible to imagine a situation in which you'd trade out of 1, 2 or 3. But, it'd be really rare. &, naturally, you can think of occasions -- many occasions! -- where teams made such bad picks at those spots that it would have been smart to do so!

For that matter, there is always, I mean literally every year, at least 1 player, sometimes a handful of them, taken lower -- sometimes substantially lower -- than the top 3 who turns out to be a lot better than 1 or more of the top 3 -- sometimes a lot better than all three of them.

I just took a quick look, & the above is true in every single draft from 1999 to 2020! Wow... in all 22 of those drafts, at least one player, sometimes a handful of them, taken way lower than the top of the draft has been better than 1 or more (sometimes all) of the top 3. The most extreme case is 1999, but 2011, 2012, 2013, 2014 & 2015 are almost equally shocking.

I admit to being surprised.

[payitforward]

Take 2011 as an incredible illustration of this fact: 2 & 3 were Derrick Williams & Enes Kanter.

15 was Kawhi, 30 was Jimmy Butler, 60 was Isaiah Thomas.

For that matter 9 was Kemba, 11 was Klay Thompson, 16 was Vucevic, 19 was Tobias Harris, 24 was Reggie Jackson, 29 was Corey Joseph, 31 was Bojan Bogdanovic, 42 was Davis Bertans, & 55 was E'Twaun Moore -- a high-level journeyman who played over 12,500 minutes in the league.

Every one of those guys was a whole hell of a lot better than Derrick Williams! A number of them were also better than Enes. & both Kawhi & Butler were better than both of those guys and also better than Kyrie, who went #1.

Kind of a wow, don't you think?

[9 and 20]

I appreciate Nate throwing together the chart, but it was extremely flukey with 28 being the worst pick in the draft over a 15 year span, and 9 being the next worst pick in that span. Again, if you use 11/13/27, all of them have produced better results than the top overall pick.

The #9 pick was literally THE BEST pick in the draft relative to it's slot. It's flukey alright. But flukey in the other direction. It outperformed picks #4, #6, #7 and #8 despite being in the portion of the draft where the dropoff for each pick is precipitous.

Chart of the production of each pick since 1980 with win shares as the metric:



Utilizing All-Star appearances as the metric:



Data is from here:

https://thedatajocks.com/nba-draft-pick-values/

Note, this article was from 2021 so they probably ignored the couple of drafts just prior to that, but a handful of recent draft picks wouldn't seriously alter a chart based on data from the last 40 years.

Wiz have been trying their best to bring this back into line. This year we'll give it another go.