Do players develop?

[Nivek]

If you listen to John Thompson, Ewing didn't grow an offensive game from scratch. He didn't demonstrate his full offensive repertoire in college (according to Thompson) because of the nature of college defenses -- namely packed-in zones. I'm not sure I believe Thompson completely, but I've heard him say that.

Draft position is important in the NBA. I saw one study a couple years ago maybe showing that it's possible to predict the length of a player's career based on draft position. I know the All-Rookie team and Rookie Game team selections can be predicted almost exclusively by draft position.

[MDStar]

I had a thought. Not sure if it belongs here or in the draft thread but here goes. I looked into the top 20 end of the year PER's from last season to see where they were drafted. I'm sure it's probably obvious but i was curious. This information pretty much lends itself to Kev's info. That the players who tend to get better, are usually those who are picked in the top 5. If you are expecting anything out of late lottery or mid 1st rounders, you are setting yourself up to be disappointing. Now this doesn't mean that cant be "good" players but the likely hood that they become the best players, is very slim.

Back on topic, I guess my thought is that yes they can get better to an extent but unless they were seen to have some type of superstar ability, it is extremely unlikely that they would develop into that level of player and actually be someone that should be an integral part of your foundation.

The average draft position of these 20 players, with the exception of Manu who is an anomaly is 5.89. So basically, you gotta get that Top 5 pick in order to get you a star.


RK PLAYER GP PER Draft pos
1 LeBron James, MIA 79 27.34 1
2 Dwight Howard, ORL 78 26.13 1
3 Dwyane Wade, MIA 76 25.65 5
4 Kevin Love, MIN 73 24.39 5
5 Kobe Bryant, LAL 82 23.94 13
6 Chris Paul, NO 80 23.76 4
7 Kevin Durant, OKC 78 23.7 2
8 R. Westbrook, OKC 82 23.63 4
9 Derrick Rose, CHI 81 23.62 1
10 Dirk Nowitzki, DAL 73 23.52 9
11 Pau Gasol, LAL 82 23.33 3
12 Amare Stoudemire, NY 78 22.78 9
13 Zach Randolph, MEM 75 22.67 19
14 Tim Duncan, SA 76 21.94 1
15 Blake Griffin, LAC 82 21.93 1
16 C. Anthony, DEN/NY 77 21.82 3
17 Manu Ginobili, SA 80 21.78 (2) 28
18 L. Aldridge, POR 81 21.57 2
19 Kevin Martin, HOU 80 21.46 26
20 D. Williams, NJ/UTAH 65 21.19 3

[Ruzious]

Wow! Mind-blowing numbers there, though it's probably not quite as bad as Nivek's numbers indicate. If nothing else, rookies achieve their PER while playing limited minutes against backups in garbage time. 4 year players are presumably doing it for extended minutes against starters and the Scouting Report.

That's the thing - rookies minutes typically come against backups and in garbage time. Most of their minutes against starters are when the starters are more winded than they are. And there's a reason for that - rookies typically play like... rookies - they don't know the tricks of the trade. Take the minutes that the play in their rookie year vs the minutes they play in their 5th year. They aren't equal, and they aren't equivalent.

[keynote]

If you listen to John Thompson, Ewing didn't grow an offensive game from scratch. He didn't demonstrate his full offensive repertoire in college (according to Thompson) because of the nature of college defenses -- namely packed-in zones. I'm not sure I believe Thompson completely, but I've heard him say that.

Yeah, I call a teeny bit of BS on that. I could see it if Ewing's game was like Shaq's: completely based on the low block. But Ewing in his prime was a jump-shooting center, and I can't see any college zone stopping a true 7 footer who could catch and shoot out to 20 feet. The fact is: Ewing did'nt have much of a perimeter game at G'Town. If he had had one, he would've used it.

That being said: Ewing *did* average 20 ppg as a rookie. So, I'm overstating it a bit myself to claim that Ewing built his game from scratch at the NBA level. Instead, I'll say that he *rebuilt* his game from scratch.

[Nivek]

Wow! Mind-blowing numbers there, though it's probably not quite as bad as Nivek's numbers indicate. If nothing else, rookies achieve their PER while playing limited minutes against backups in garbage time. 4 year players are presumably doing it for extended minutes against starters and the Scouting Report.

That's the thing - rookies minutes typically come against backups and in garbage time. Most of their minutes against starters are when the starters are more winded than they are. And there's a reason for that - rookies typically play like... rookies - they don't know the tricks of the trade. Take the minutes that the play in their rookie year vs the minutes they play in their 5th year. They aren't equal, and they aren't equivalent.

It depends on which rookies you're talking about. Rookies averaging 30+ minutes a night -- your average top 5 pick -- are likely facing starters a good part of the time. Rooks getting less than 24 minutes a night are probably going against subs a fair amount of the time -- UNLESS they're starting games, in which case more of the minutes are probably against starters.

There's a way to estimate minutes vs. starters as a percentage, but it's a pain in the ass to gather all the data. Might be able to do it for rookie year, at least.

Edit: I don't see an easy way to get at the starter minutes estimate. Hmm, maybe I can ask one of the guys at APBRmetrics.

[7-Day Dray]

Players can develop, but players don't develop here because we're a dysfunctional organization.

[keynote]

Code: Select all

PICK    IMP
1       2.3
2       2.4
3       1.6
4       2.3
5       2.5
6       -1.1 <---
7       -0.7
8       1.2
9       1.5
10      0.8
11      1.0
12      0.2
13      -0.5
14      -1.0
15      -0.4
16      -0.9
17      -0.3
18      -2.3
19      -0.1
20      -0.5
21      0.8
22      -3.3
23      -1.3
24      -2.4
25      -1.5
26      -2.5
27      -1.3
28      -4.0
29      -1.1
30      -2.1

AVG     -0.4


We're screwed

It would be interesting to do a team by team analysis to look at the organization as a whole from either a scouting perspective or a player development perspective.

That would be nice, agreed. Although, I'm not sure how to sift through all of the variables to draw any worthwhile conclusions. At a minimum, we'd have to normalize for draft position. But, as has been discussed earlier in the thread, high picks are typically asked to do a different job than lower picks. A lower pick may be developed to do one thing (bang, play perimeter D, hit threes, etc); a higher pick will likely be asked to build a more varied skill set. Presumably, a player would get at setting picks and boxing out if that's all he worked on after practice than if he also ran through ball-handling drills, passing drills, and shot a Gilbertian amount of turn-around jumpers.

Still, sorting the data by team would be a nice first start. I can only hope that Ted & Co. have some way of tracking player development success by team - it sure would be nice for FO recruiting/poaching purposes.

[Nivek]

Not going to get to the "by team" thing right now. Took another look at the data using total PER (instead of the per 40 minute PER, which is how Hollinger (and b-r) presents it). As you'll see below, TOTAL PER production increases from year one to year four. That's because playing time increases pretty much across the board through the top 30 picks.

Possible explanations for the playing time increases even though per minute production has gone down:

1) Lack of choice -- older players retire, get cut/traded, and teams are forced to rely on these guys
2) Players actually have improved -- I mentioned this earlier that the players are "better" in the sense that they're more useful to helping a team win even though they may a lower PER.
3) Some other third thing -- perhaps coaches are just wrong. Maybe by their 4th season players have figured out that scoring will get them more playing time so they just shoot more.

Re: #1 -- We're seeing this to some extent in Washington. The team rid itself of veterans and have been forced to play Young and McGee (for example), pretty much whether they like it or not.

All this may actually connect back to Dan Rosenbaum's statistical +/- research. Basically (IIRC), Dan ran a bunch of regressions of box score data and +/- data. When he was finished, he could use box score data to estimate what a player's +/- SHOULD be. One of Dan's interesting findings was that younger players tended to have box score stats that "out-performed" their +/- scores. In other words, there was a disconnect between young players' stats and their ability to affect the scoreboard. That "lack of impact" seemed to disappear by a player's 3rd or 4th season, if I'm remembering right.

Also, there's a 0.7 correlation between total production improvement and minutes per game increase. Since the chart I put up earlier in the thread shows PER MINUTE production declining for many of these draft slots, it's pretty clear that more playing time is the cause for increased TOTAL production.

Code: Select all

PICK    TotImp  MPG diff
1       133     2.8
2       262     3.2
3       122     0.8
4       224     0.4
5       143     2.9
6       -54     -3.9
7       172     1.4
8       77      1.8
9       184     2.5
10      178     5.9
11      266     5.5
12      100     3.6
13      238     3.3
14      45      1.2
15      145     4.7
16      58      1.5
17      149     1.2
18      96      0.8
19      211     5.2
20      78      3.3
21      186     2.9
22      93      -0.2
23      230     5.2
24      -34     -2.0
25      104     1.7
26      185     3.5
27      178     4.1
28      55      0.8
29      86      1.3
30      173     1.6

AVG     136     2.2


[pancakes3]

Players can develop, but players don't develop here because we're a dysfunctional organization.

don't you think it's more on the players that they aren't working on their games than the organization for being "dysfunctional".

i want to also direct attention to nivek's sig. "a lot of what we call talent is actually the desire to practice."

[fugop]

A couple of questions/observations:

1. Is it agreed that players develop in college? I think this is pretty clear, though my sense is based on anecdote and conjecture. 1st, there are a couple of guys who couldn't do much when they started college, but over the course of their careers became significant contributors -- Roy Hibbert is the best example that comes to mind. 2nd, the games between mid-majors and their typically more mature rosters versus the younger elite-recruiting colleges demonstrate something that I would characterize as development.

2. The time frame from 1985->2012 is interesting as it captures all of the changes in the average aage of a draftee. Guys drafted from 1985-1993 or so, were probably a year or two older, on average, than guys drafted from c. 1993->2007. Guys drafted since the 1 & done rule are probably somewhere in the middle of the age spectrum. I would like to see if there is a significant difference in post-draft development based on the age of the draftee.

3. This probably does vary, at the margins, based on club. I'd guess that it's as much peer influence as coach or organizational direction, though.

[doclinkin]

By observation at least in the NCAA two things seem to reliably improve with experience and maturity: fouls and turnovers. Generally I'm looking at draft pick prospects so I haven't tracked the stats as closely with the NBAers, but it seems to ring true there as well. For younger players anyway.

[fishercob]

Very interesting discussion indeed. It's a question of degrees to me. At it's core -- can guys who can't help you win learn to help you win? Is there any historical data that would inform us as to whether drafting a self-check like Vesely (even if he;s good at other things) can ever grow into a winning player.

The draft is all about potential. But should it be? Should we just be drafting guys who can definitely help, even if their ceiling isn't that high. Ernie swung for the fences with McGee, Pecherov, Blatche, Vesely, Nick, etc -- all guys who had very obvious warts coming in to the league, and we have very littel to show for it.

I have more thoughts on the subject but really am not feeling well. WIll post more another time.

[JonathanJoseph]

Agree that this is an interesting thread.

I think the answer is yes, but with some caveats and limitations. Jared Jeffries seemed to have a credible offensive game in college but it's been absent in the pros. Meanwhile, Antawn Jamision made himself into a long range shooter through hard work.

You can't overcome some limitations (like Kwame's hands) but other things can be improved within reason (like Vesely's jump shot, given that he seems to have some touch) and Seraphin's offensive game which needs experience and mental and physical reps as much as anything.

There's no quick answer to this question, but an interesting one to ponder.

[closg00]

[youtube]http://www.youtube.com/watch?v=ZAamJml1VLQ&feature=related[/youtube]

http://hoopshype.com/articles/sierra/mc ... a-can-wait
Yes, happens all the time, players mature overseas or in the NBDL and develop their games.

[dangermouse]

Why dont they seem to develop for us? Or wait until they get to another team before they blow up?

I think we need a new GM, talent scouts, coaches... the works. But guys/gals who are already proven.

Is there a Presti out there on another team who wants a challenge?

[Illuminaire]

Well, for one thing, it seems like the players who improve the most are the ones with a) a great work ethic and b) enough BBIQ to understand their weaknesses.

We have precious few players who match that description.

[Nivek]

I'm in the midst of managing a "player development" process, albeit not one related to basketball. Teachers/coaches make a big difference when the player/student is talented, hard-working, enthusiastic and receptive. Remove any of those factors and the value of the teacher/coach diminishes quickly.

The reality of player development is something I've been saying for awhile now: Players develop by going to a gym on their own and putting in the work. Having the right coach/teacher/tutor can help, but the way humans develop skills is by repetition. The coach can help, but ultimately the player has to get his ass in the gym and do it. Over and over and over again -- until it's second nature. Until it's automatic. The greats are the ones who obsess over the little details -- who do it perfectly for hours until they KNOW they have it down.

Magic taught himself the sky hook, how to shoot free throws at a 90% rate, how to shoot jumpers. Bird taught himself how to use his left hand almost as well as he did his right hand.

Player development isn't some mystical thing. It's not a mystery. It's a lot of long, boring hours in the gym doing the same thing over and over again. Guys willing to do that will improve. Guys not willing to do that might get a little better here and there, but won't "develop" in the way we all mean.

[jivelikenice]

Players can develop. I think the degree of development is where the debate lies. I think the issue with development with players on this roster is we're taking a lot of players who were never really supposed to be more than roles players and hoping that they develop into bonafide starters.

[fugop]

Why dont they seem to develop for us? Or wait until they get to another team before they blow up?

I think we need a new GM, talent scouts, coaches... the works. But guys/gals who are already proven.

Is there a Presti out there on another team who wants a challenge?

A few Wizards whom I would say "developed" during their time on the roster:

1. Caron Butler. Sub-15 PER before arriving in Washington. >17 while here. Statistical improvement across the board, from shooting percentages to rebounding and assist percentages.

2. Larry Hughes. Hughes' years here were massive anomalies, performance outliers.

3. Brendan Haywood. His ORtg is a straight line improvement.

4. Gilbert Arenas. Gilbert's PER his four years with the Wizards: 16, 21.3, 23.8, 24.

Other guys like Juan Dixon showed steady statistical improvement. It seems to me that the combination of players/coaches in the mid-2000s had success developing guards. Whether that's system or personality, I have no idea.

[Zonkerbl]

Just went back to my draft pick analysis spread sheet and compared rookie PER to 4th year PER for the top 30 picks in every draft (where players have completed at least 4 seasons) since 1985. That's a pool of 690 players. A full table of average improvement by pick is below.

The AVERAGE "improvement" in PER from year one to year 4 was -0.4. In other words, on average this group of 690 players got slightly less productive as measured by PER. That number is affected by some players who weren't in the league by a 4th season.

As you'll see below, biggest "average" improvements were made by top 5 picks. This may be in part a selection bias. Guys taken in the top 5 are perceived on draft day as having the most ability. Teams therefore invest more in those players in terms of coaching, dollars, and playing time.

I could do the same analysis using total PER, which would account for playing time, but the result would be pretty similar. I could also go back and run the numbers for years 2 and 3. Can't go beyond year 4, though because that's where I stopped in my draft pick analysis and I don't have the time to go back and compile a new data set.

Code: Select all

PICK    IMP
1       2.3
2       2.4
3       1.6
4       2.3
5       2.5
6       -1.1
7       -0.7
8       1.2
9       1.5
10      0.8
11      1.0
12      0.2
13      -0.5
14      -1.0
15      -0.4
16      -0.9
17      -0.3
18      -2.3
19      -0.1
20      -0.5
21      0.8
22      -3.3
23      -1.3
24      -2.4
25      -1.5
26      -2.5
27      -1.3
28      -4.0
29      -1.1
30      -2.1

AVG     -0.4


My one quibble with this analysis is that a large proportion of your sample is from years when players stayed in college for four years and are therefore popping into the NBA more or less fully developed. Kevin Garnett opened the floodgates in 1995 and it wasn't really until Kwame Brown in 2001 or so that you regularly had all the best players leaving after one year (unless you start counting from Elton Brand).

I would argue that your 1985-2000 sample is fundamentally different from your 2001-2007 sample. I would be interested to see if the "improvement" numbers in the 2001 to 2007 subsample are different, because a lot more of the impact players are coming in with virtually no college-level instruction.

Or to ask the same question differently, how much do players that came in out of HS vs came out after Freshman yr vs came out after sophomore year improve? A lot of our players are in that "no or very little formal bball instruction" category. Vesely is arguably our most experienced pick and he's still super young.