NBA Draft Philosophy
By Blogger
The NBA Draft, as executed these days, is an exercise in near-insanity. Except for the odd Tim Duncan no-brainer, GMs seem to think selecting basketball players for their teams is an in-depth affair that requires years of experience and training in Auerbachian mysticism to choose just the right player. Um, no.
The NBA GM’s job in the draft has been idiot-proofed. First round draft choices are assigned contracts based on their draft slots. Those players are tied to their teams for three years before the GM must make a decision on the player’s future, and even then the hard decision can be delayed a year by exercising a one year option. In other words, there will be no negotiation, and the only thing the GM has to figure out is which player will be best for his team. No hold outs. No signability issues. Simple, right?
Actually, though, NBA GMs make their jobs infinitely more difficult than they need to be. The primary way they do this is by selecting players in the draft based on “upside”, when such an attribute doesn’t measurably exist. It’s been explained elsewhere before, but when someone tries to assign a level of “upside” to a player, he or she is using imagination to picture what a player might be instead of basing that prediction on more objective fact. Other ways they do this is by ignoring players that could help, but are considered to have low ceilings, or by ignoring players that are blocked at their position by an entrenched starter. With only a little interrogation, though, one sees that neither of those conditions are real downsides.
After seeing bust after bust come out of the first round over the years, often because his future was wildly overrated by imaginative talent evaluators, Zach and I started talking about formulating a simple NBA Draft philosophy, perhaps in the same vein as Tangotiger’s baseball Marcel Projections, and I’ve consolidated mine into these guiding principles:
1)Every draft pick should result in a useful player, whether that means choosing a player or trading the pick for a player. After the blue chip prospects have gone off the board, teams start choosing project players and guys with major flaws in earnest. The star/bust mentality takes over, and guys like Luke Walton, who nobody expected to ever be an All Star but projected to be a solid seventh man in the league for years, get ignored. Those guys are infinitely more useful than the Maciej Lampes of the world. If a team turned every pick into a useful player, whether that guy is 18 years old and fresh out of high school, 22 years old and coming from Turkey, or 33 years old and coming from another team, they'll suddenly have a deep team.
2)Position only matters if the young player won’t get ANY playing time. In other words, picking Darko over Carmelo or Wade was utterly moronic. Joe Dumars may have thought Darko was better than Bosh, but there’s no way he thought he was a better overall player than Carmelo. He didn’t take Anthony because he already had Tayshaun Prince. The problem with that logic, though, is that it assumes that Anthony wouldn’t play because Prince was there. You think a good coach wouldn’t find a way to play both of them, either at the same time or in a rotation? That team is already a juggernaut, and would have been even more dominant had the Darko pick not bombed so spectacularly. However, on the flip side, the current Golden State Warriors would be stupid to select a point guard, since they already have both Baron Davis and Derek Fisher locking up the position for the foreseeable future.
3)The best player in the long term is likely the best player at the time of the draft. This applies more to blocs of players than it does to individual one to one player comparisons, but the best way to explain this principle is a one to one comparison. Aaron McKie was not as good as Jason Kidd at the time of their draft, and he still isn’t as good. That's just how basketball improvement works in the NBA. Even comparing high school guys to college guys to overseas guys, the best player at the time of the draft is likely to be the best in the future.
4)You shouldn't completely give up on a player until after his fourth season or his age-23 season, whichever comes first. At that point, he is what he is and you’ll know what he can do. The charts linked below bear this out. Except for a very few notable exceptions (Jermaine O’Neal), players do not improve dramatically over time. This ties in to number 3 above, too, in that the charts show that guys don’t leapfrog over everyone in their class.
On the charts, the players are listed in their draft order and by draft class. The stat I’ve chosen to use is PER. Bolded numbers indicate that that player led his class that season in PER. Seasons are left blank if the player did not play at least 500 minutes. The stat’s inventor, John Hollinger, has cautioned against its use as a catch-all number, so understand that while it’s a holistic number, it’s just one tool among many we use to evaluate players. In that spirit, keep in mind that while Kobe’s PER his rookie year was comparable to Tony Delk’s, nobody thought Delk was the player Kobe was.
The web site, Basketball-Reference.com, is invaluable, and if you haven’t bookmarked it yet, you’re missing out.
I hope that this column spurs more debate and study. I trust that someone can do more with available numbers than I’ve done in this modest endeavor.
View the charts HERE. (They looked terrible when I tried to fit them within the column.)
The NBA GM’s job in the draft has been idiot-proofed. First round draft choices are assigned contracts based on their draft slots. Those players are tied to their teams for three years before the GM must make a decision on the player’s future, and even then the hard decision can be delayed a year by exercising a one year option. In other words, there will be no negotiation, and the only thing the GM has to figure out is which player will be best for his team. No hold outs. No signability issues. Simple, right?
Actually, though, NBA GMs make their jobs infinitely more difficult than they need to be. The primary way they do this is by selecting players in the draft based on “upside”, when such an attribute doesn’t measurably exist. It’s been explained elsewhere before, but when someone tries to assign a level of “upside” to a player, he or she is using imagination to picture what a player might be instead of basing that prediction on more objective fact. Other ways they do this is by ignoring players that could help, but are considered to have low ceilings, or by ignoring players that are blocked at their position by an entrenched starter. With only a little interrogation, though, one sees that neither of those conditions are real downsides.
After seeing bust after bust come out of the first round over the years, often because his future was wildly overrated by imaginative talent evaluators, Zach and I started talking about formulating a simple NBA Draft philosophy, perhaps in the same vein as Tangotiger’s baseball Marcel Projections, and I’ve consolidated mine into these guiding principles:
1)Every draft pick should result in a useful player, whether that means choosing a player or trading the pick for a player. After the blue chip prospects have gone off the board, teams start choosing project players and guys with major flaws in earnest. The star/bust mentality takes over, and guys like Luke Walton, who nobody expected to ever be an All Star but projected to be a solid seventh man in the league for years, get ignored. Those guys are infinitely more useful than the Maciej Lampes of the world. If a team turned every pick into a useful player, whether that guy is 18 years old and fresh out of high school, 22 years old and coming from Turkey, or 33 years old and coming from another team, they'll suddenly have a deep team.
2)Position only matters if the young player won’t get ANY playing time. In other words, picking Darko over Carmelo or Wade was utterly moronic. Joe Dumars may have thought Darko was better than Bosh, but there’s no way he thought he was a better overall player than Carmelo. He didn’t take Anthony because he already had Tayshaun Prince. The problem with that logic, though, is that it assumes that Anthony wouldn’t play because Prince was there. You think a good coach wouldn’t find a way to play both of them, either at the same time or in a rotation? That team is already a juggernaut, and would have been even more dominant had the Darko pick not bombed so spectacularly. However, on the flip side, the current Golden State Warriors would be stupid to select a point guard, since they already have both Baron Davis and Derek Fisher locking up the position for the foreseeable future.
3)The best player in the long term is likely the best player at the time of the draft. This applies more to blocs of players than it does to individual one to one player comparisons, but the best way to explain this principle is a one to one comparison. Aaron McKie was not as good as Jason Kidd at the time of their draft, and he still isn’t as good. That's just how basketball improvement works in the NBA. Even comparing high school guys to college guys to overseas guys, the best player at the time of the draft is likely to be the best in the future.
4)You shouldn't completely give up on a player until after his fourth season or his age-23 season, whichever comes first. At that point, he is what he is and you’ll know what he can do. The charts linked below bear this out. Except for a very few notable exceptions (Jermaine O’Neal), players do not improve dramatically over time. This ties in to number 3 above, too, in that the charts show that guys don’t leapfrog over everyone in their class.
On the charts, the players are listed in their draft order and by draft class. The stat I’ve chosen to use is PER. Bolded numbers indicate that that player led his class that season in PER. Seasons are left blank if the player did not play at least 500 minutes. The stat’s inventor, John Hollinger, has cautioned against its use as a catch-all number, so understand that while it’s a holistic number, it’s just one tool among many we use to evaluate players. In that spirit, keep in mind that while Kobe’s PER his rookie year was comparable to Tony Delk’s, nobody thought Delk was the player Kobe was.
The web site, Basketball-Reference.com, is invaluable, and if you haven’t bookmarked it yet, you’re missing out.
I hope that this column spurs more debate and study. I trust that someone can do more with available numbers than I’ve done in this modest endeavor.
View the charts HERE. (They looked terrible when I tried to fit them within the column.)
posted by Blogger at 12:05 AM
1 Comments:
You've laid out some good guidlines. I wish the team I root for (Jazz) had followed number 3. Chris Paul was a better player than Deron Williams in college yet the Jazz chose Williams based on his size and the dreaded "upside" potential. Williams could be a solid pro but Chris Paul was should have been their pick.
Post a Comment
<< Home