NBA Age Limit (Player Development,1 and Done,and HS players)

[colts18]

With all the talk of raising the age limit recently, I wanted to explore what kind of effect that would have on the players. I looked at players from the 1995 to 2014 season. The reason I chose 1995 as a marking point is because that is when KG became the 1st modern player to jump straight into the draft.


Players from 1995-2014:

Code: Select all

Rookie				2nd year				3rd year	
age	#	WS/48	PER		#	WS/48	PER		#	WS/48	PER
18	12	0.066	13.6		12	0.100	14.7		11	0.122	16.7
19	64	0.066	14.1		59	0.090	15.8		48	0.110	16.9
20	130	0.069	13.8		116	0.090	15.4		102	0.108	16.8
21	178	0.066	13.6		155	0.093	15.1		134	0.103	15.8
22	338	0.067	12.8		278	0.087	14.1		240	0.099	14.9

% at 15+ PER (League average):

Code: Select all

Age	#	%			     #	%			       #	%
18	2	16.7%		      4	33.3%			6	54.5%
19	18	28.1%			27	45.8%			23	47.9%
20	33	25.4%			47	40.5%			47	46.1%
21	36	20.2%			58	37.4%			63	47.0%
22	58	17.2%			67	24.1%			73	30.4%

All of the age groups have basically the same WS/48 for their rookie seasons. They range from .066 to .069. According to the pundits and commissioner, players would develop better if they were in college but the data doesn't show that. For the average 19 year old draft pick, his PER is 15.8 in his age 20 season, much higher compared to an Age 20 rookie who is at 13.8 PER. An age 21 year player is at 16.9 PER compared to an Age 21 2nd year player who is at 15.4 PER. The same patterns show up no matter what age group you compare.


Here is Improvement in PER and WS/48 from year 1 to Year 3.

Code: Select all

age	WS/48	PER
18	0.057	3.1
19	0.044	2.8
20	0.040	3.0
21	0.037	2.2
22	0.032	2.1

The younger the draft pick, the more he improves from year 1 to year 3.


One common criticism of the NBA is that players don't develop in the NBA if they are sitting on the bench. But when you look at the data, that doesn't show up. Players who play infrequently (<1000 MP) in their age 18-20 rookie seasons do improve quite a bit the next year.

Code: Select all

	      Rookie	Year 2		Improvement
MP	   364	  903		   539
PER	   11.1	13.2		   2.1
WS/48	0.040	0.070		0.030

Here are the stats for the high schoolers from 1995-2005 that were drafted and played a minute in the NBA. Despite the narrative, high schoolers actually come up well in this measurement. Their rookie year WS/48 is similar to the 18-23 age groups. Their improvement compares well too. They played less minutes than the age 20-22 age groups, but they were equally productive in those minutes.

Code: Select all

	      Rookie	2nd	3rd		Improvement
PER	   14.3	15.6	17.1		   2.8
WS/48	0.065	0.088	0.113		0.048

Here is a comparison between rookies from 1980-1994 and 1995-2014 rookies. The table is all rookies from age 18-23 in these time periods.

Code: Select all

	      Age	MPG	WS/48	PER	% w/15+ PER
80-94	22.3	18.2	0.064	13.2	16.6%
95-14	21.7	18.3	0.065	13.0	17.8%

There is literally no difference in the quality of rookies in modern times despite all the complaints of rookies being less ready these days. The average rookie is younger these days. From 80-94, there was only 1 rookie who was 19 years old (Cliff Robinson) and just 8 who were 20 years. Young guys rarely came out in those days.


Notes:
-Each age group performs similarly in their rookie years but the younger groups take over in years 2 and 3
-NBA players develop quite a bit more in the NBA than in college. Adam Silver's proposal to raise the age limit is robbing young stars of potential development in the NBA.
-LeBron led his team to the finals in his year 4 season. Would he have been able to accomplishment that if he stayed in college for 3 years and his 2007 season was his rookie year? I doubt it.
-The young guys aren't the problem in the NBA. Don't listen to that narrative. They are play just as well as older guys and players from back in the day.
-High schoolers did awesome in the NBA when you compare them to the age 19-23 groups. Lots of people said that for every Kobe there are 10 Kwame Browns, but the high school data doesn't prove that. For every Kwame there are guys like Rashard Lewis, Martell Webster, Andray Blatche, or Amir Johnson that came straight out of HS who were productive but stayed under the radar. In fact Kwame Brown had a pretty long career (10+ years). So did another HS bust (Eddy Curry).

[Mike G]

Your 18- and 19-yr old rookies are generally better at that age than their peers who stay in college. That's why they are better at age 23; not because they came out earlier.

There's a study around here of players who reached certain levels of PER by a given age. These blocks of similarly-skilled players were compared by their age of entry into the NBA. One thing they didn't gain was any extra time in the league. Antoine Walker, Marbury, and others were about done by age 30.

It's impossible to compare the career of a player who enters at age 19 to himself if he enters later. Some guys leave college because they aren't smart enough to make it through; they might also not be smart enough to be adult millionaires.

There was this thread exactly 3 years ago - viewtopic.php?f=2&t=125

[Mike G]

OK, I think I have found my spreadsheet that breaks down careers that started after 1990 and had ended by 2010.
I split the careers into those that started before age 22; at age 22; and later than 22.
These groups were broken into those who by age 24 had career PER of: >16; 14.5 to 16; 13.5 to 14.5. Lower PER were not considered.

Showing: avg rookie age; sample size; avg PER and Minutes thru age 24 (defined at b-r.com); avg age in last season, years in the league, career games and minutes.

Code: Select all

Age-R    #   PER-24  Min-24   Age-L    yrs     G      Min
20.5    12    18.6    9493    31.9    12.4    839    26077
22.0    17    17.7    6936    32.4    11.4    691    20960
23.1     9    18.3    4736    34.8    12.7    814    25861
                                
20.7    12    15.1    6654    30.3    10.7    676    17620
22.0    20    15.3    6514    32.8    11.8    790    23682
23.0     3    14.7    3970    34.7    12.7    876    21832
                                
20.5     8    14.1    8505    30.3    10.8    654    18453
22.0    26    13.9    4843    31.8    10.8    716    18214
23.0     5    14.0    4107    29.0     7.0    354     8911                          

The better players (top 2 groups) seemed to gain longevity by waiting longer to declare. Lesser players did not; but sample size should be noted.

[colts18]

OK, I think I have found my spreadsheet that breaks down careers that started after 1990 and had ended by 2010.
I split the careers into those that started before age 22; at age 22; and later than 22.
These groups were broken into those who by age 24 had career PER of: >16; 14.5 to 16; 13.5 to 14.5. Lower PER were not considered.

Showing: avg rookie age; sample size; avg PER and Minutes thru age 24 (defined at b-r.com); avg age in last season, years in the league, career games and minutes.

Code: Select all

Age-R    #   PER-24  Min-24   Age-L    yrs     G      Min
20.5    12    18.6    9493    31.9    12.4    839    26077
22.0    17    17.7    6936    32.4    11.4    691    20960
23.1     9    18.3    4736    34.8    12.7    814    25861
                                
20.7    12    15.1    6654    30.3    10.7    676    17620
22.0    20    15.3    6514    32.8    11.8    790    23682
23.0     3    14.7    3970    34.7    12.7    876    21832
                                
20.5     8    14.1    8505    30.3    10.8    654    18453
22.0    26    13.9    4843    31.8    10.8    716    18214
23.0     5    14.0    4107    29.0     7.0    354     8911                          

The better players (top 2 groups) seemed to gain longevity by waiting longer to declare. Lesser players did not; but sample size should be noted.

There is a huge selection bias in your data. As I mentioned in the OP, young players didn't come out in the early 90's. They started coming out in the mid to late 90's. So if there career was already finished by 2010, then they had a short career because more than likely they came out in the late 90's or early 2000's. Your study won't pick up people like Kevin Garnett or Rasheed Wallace.

[Mike G]

Well, given that the idea was to study the career length of early vs later entries, it's hard to avoid having a sample that is limited to finished careers.

But I'm not sure there's an inherent bias in it. Extra-long careers are under-represented from both early and later entries. KG and 'Sheed aren't included, and neither are Ginobili, Andre Miller, and others who started late.

If, among those whose careers have finished, late starters are also late finishers, should we expect current careers -- once they have finished -- to be distributed differently?
I'd agree that it's time to re-do the study. In 2010, it seemed we had just reached the point where we could see meaningful results. In 4 years, a lot more careers could be included.

EDIT - Rasheed Wallace was in the study, as I believed he was finished in 2010.

[colts18]

I looked at the 37 high school players who were drafted in the NBA and played a minute in the league. You have to remember that a lot of these players are still active

Average:
10.0 seasons played (10.1 if you count Livingston and Bynum's injured seasons)
600.0 Games played
16785.7 MP

24 out of 37 still active (64.9%) (All of the HS draftees from 2005 are still active)
6 of the inactive played last season (81.1% played through 2013)

28 played 9+ seasons
32 played 7+ seasons
12 played 12+ seasons (7 still active)
5 played 5 or less seasons

High school players have killed it so far in the NBA. KG is in his 19th season, Jermaine O'Neal in his 18th. So is Kobe. There might be 3 or 4 players who will end up with 20+ seasons from the High school list

[Mike G]

Amir Johnson is 'still in the league' and 'killing it', after a 39-minute rookie season, then 124, etc. He did get some serious D-league time, but not what he'd have gotten in college.

If subsequent studies concur that an earlier NBA entry also predicts an earlier career demise, then the alternatives become more stark: At age 30, you might be out of basketball with no college degree; or you might still be playing AND have a degree for later.

Drafting players who aren't ready for the NBA just so that nobody else can get them is a crappy use of human resources, IMO. And I'd rather see these guys playing in college and making those games more fun to watch.

Average:
600.0 Games played
16785.7 MP

That's right around 28 mpg for their careers to date.
When I look at the 44 players who got 2 NBA years by age 20, between 1996 and 2006, they averaged 22 mpg for their first 2 seasons. That's in 53.5 games and 1170 minutes.
http://bkref.com/tiny/248N2

Continuing to the present season -- the parameters include Anthony Davis, Andre Drummond, Kyrie Irving -- the averages are the same.
Minutes-weighted avg PER has been right around 15, avg WS/48 = .081, both before and after 2005.

If we don't let the LeBrons and Dwights dominate the weightings, then players' avg mpg = 18.6, PER = 13.0, WS/48 = .060, before 2006.
From 2006 on, they're 20.6, 14.0, and .069 -- no older (?) but with a year of college.

Again, these are players who either believed they were ready for the bigtime, or they were counseled to go for it, or they just couldn't cut college. A mixed bag, and doesn't include those who never played a minute.

[Mike G]

A mini-glimpse at rookie age and career development: Age at NBA entry vs Age at Peak.
From the top 100 Win Share seasons in this Century; top 100 spots are taken by 35 different players. Peak is defined as 2 of 3 best: WS, WS/48, and PER. If it's a single entry for a player, it's just WS.
http://bkref.com/tiny/LwRSZ
By age in rookie season, showing avg age of Peak season:

Code: Select all

AgeR  #    Peak   players
18    2    25.0   Kobe, McGrady
19    4    25.3   Garnett, LeBron, Dwight, Durant
20   10    25.0   Amar'e, Paul, Rose, Dirk, Wade, Brand, Arenas, Harden, Love, Shaq
21   10    26.7   Griffin, Billups, Gasol, Pierce, Peja, Ray, Marion, Curry, Duncan
22    6    27.2   Mourning, Roy, Payton, Francis, Nash, Vince
24    2    32.0   B Barry, Cassell

Pierce is counted twice. He may have peaked at age 24 (2002) or at 30 (2008). He doesn't change the 21YO avg of 27.
Players who are judged to have peaked this season: Durant, Harden, Love, Griffin, Curry. I wouldn't bet on any of them to improve on this season, though 1 or 2 might.
Players who peaked in the '90s -- Robinson, Malone -- are excluded.

[colts18]

538 did a study on this




http://fivethirtyeight.com/features/dra ... nba-teams/

[Mike G]

Thanks for the link. More info from that article:

Code: Select all

.     4-yr Win Shares / Others drafted  / Relative
. Class           Avg / at same pos. avg / Value
Freshmen & HS    12.0        12.8          - 6%
Sophomores       13.1        11.8          +11%
Jr & Sr           9.6         9.3          + 3%
International     7.3         8.5          -15%

Nate Silver suggests their may be some value in drafting players after 2 years of college. And upperclassmen may also be a better value at a given draft position, relative to freshmen and internationals.

[repole]

Would be interesting to try and compare guys based on their high school recruiting rankings, but there's admittedly a lot of variance from year to year, especially at the top.

I have trouble getting behind the argument that these guys are ill prepared for the world if they don't go to college for a year or two and then end up out of the league. How many one and done guys are going back to get their degree? And how many guys are taking classes that have actual tangible real world value? It would seem likely to me that the guys who intend on getting to the NBA as soon as possible won't be taking college academics as seriously as one might hope.

That said, I understand why the league wants a higher age limit, it makes perfect sense from a product stand point. Get guys who are closer to their prime, meaning the value a player produces while on their cost controlled rookie deal is substantially higher, and because of it the actual quality of your product has increased. From a team standpoint, it has to hurt when you've spent 4 years developing a high school kid, they haven't done much yet, they leave, and eventually blossom for someone else.

I'd have no qualms with someone arguing that raising the age limit is a good thing for the NBA, or for the individual teams, but I don't see how taking away options from an 18 year old is a good thing for them.

[Mike G]

What happened before the age limit and HS+1-yr was that high school kids were being heavily scouted, and then drafted, in the event that they might grow into their elongated frames and be coordinated enough to play well.
So teams were drafting kids who were clearly not ready to contribute, just so other teams could not get them.

Kids who really, really wanted to start making a million bucks were able to do so. What they missed out on wasn't so much a serious or complete college education, but simply a year to demonstrate at least a bit of their potential -- or lack thereof.

There's no universal age cutoff separating the men from the boys. But HS kids are kids, almost 100%. Ages 17, 18, 19, a lot of psychosocial development is going on, as well as physical.

The options for a young player who is barely eligible (and not quite ready for NBA) are: (a) to blossom and star in college, vs (b) to warm a bench for money.

[Statman]

The options for a young player who is barely eligible (and not quite ready for NBA) are: (a) to blossom and star in college, vs (b) to warm a bench for money.

Baseball has it right here - the "star" players coming out of high school and college (except for VERY few exceptions) are just plain NOT ready for the big leagues. The holes in their game (or lack of physical development) would be so obvious. So, star young players know they need to pay their "dues".

NBA teams draft players that are FAR from ready (very few draftees are going to be a TRUE positive impact for their teams as rookies) - and then feel they HAVE to get them out there on the court to justify the draft pick. This is compounded by the fact that an NBA team is only 5 deep at any one time - and the team feels they HAVE to get their guy they gave that guaranteed contract some court time - even if honestly he'd be AT BEST AA if it were baseball (Zach Lavine for example).

In MLB - Kwame Brown may never have played a game in the majors, except for maybe some late September call up. Darko, maybe the same. It's too bad with the salary structure the way it is and how teams have historically pushed their big ego youngin's into the far so early that we may never really see a proper player development. VERY FEW players are ready to be actual POSITIVE contributors (ie, at least coming close to an average NBA player) before the age of 23. We have outliers, but honestly there are just maybe a couple a draft at most.

The Spurs, well, they draft guys and pretty much ALLOW them to develop (usually overseas) and get them on the court in San Antonio much closer to their prime. These more mature, more developed players understand how to "mesh" better - they paid their dues, they learned their trade and maybe how best to get the most from their games - while minimizing their flaws.

I don't understand how MANY NBA teams in the first round don't draft for BEST potential for positive contribution over the first 4 years. Guys like Zach Lavine - well there is a high likelihood he will be no better than a replacement player for 3 years, and the team will completely give up on him after "wasting" that guaranteed $$ and roster spot by the time he's 22. I'm worried Andrew Wiggins may be a little like this also - I think he'll be WAY in over his head as a rookie, and probably year 2 - he was nowhere near the dominant player at the same age college guys like Anthony Davis, Blake, Durant, Kyrie, Love, etc.

Kids not named LeBron need time. Truly ELITE kids sometimes need just one year. The rest need time.