Raw PER Improvement Graph created by Dan Sinclair of Brew Hoop and Where55Happens.wordpress.com
Hopefully you have enjoyed the NBA Draft Lottery study Mitchell Maurer and I have published over the last week or so here at Brew Hoop. We have had a lot of fun in the process, and hope we have added some interesting and important knowledge to the NBA draft discussion. Everyone at Brew Hoop would like to encourage you to take another look at Part 1, Part 2, Part 3, and then use the data set from Part 3 (and the data set from this study) to make further insights and add to the draft knowledge of the Brew Hoop community using the FanPost feature. In the spirit of expanding our draft-related content, I did a bit of additional research into the issue of player improvement among the lottery picks over the last decade. Specifically, I looked at the PER of the first qualifying season a player produced in the NBA and compared it with the PER of the best qualifying season in a player's career. Recall that we define a qualifying season in this way:
By our standards, a qualifying season is one in which the player plays in at least 50 games and averages at least 12.0 minutes played per game. If one of those two conditions is not met, that player's season is rendered non-qualifying for the purposes of this study.
The idea is to explore the issue of how much weight we should give to the first season a player gets significant playing time, and whether these young players with so much of their value tied up in "potential" do in fact become more productive over time. I kept this study limited to players drafted within picks #1-16 during the past decade that have three or more qualifying seasons (50+games played, 12+ min/gm), so obviously the 2009 and 2010 drafts are automatically excluded for not meeting qualifying season requirements. Just for reference, 112 of the 144 eligible players (78%) met my parameters. Here is the link to the data set, and after the jump you can see what I found:
(1) Don't judge a lottery player too harshly based on their first qualifying season, because most players improve. Of the 112 players who met the parameters, 102 of them (91%) had a high qualifying season better than their first qualifying season. Of the ten players who did not show some level of improvement, five of those players posted above average PER in their first qualifying season: Etan Thomas- 16.4, Darko Milicic- 15.2, Nick Collison- 15.0, Channing Frye- 18.1, Micheal Beasley- 17.2. As you can see, these are pretty valuable seasons right off the bat, so in these cases the lack of improvement can't really be held against the players. The intuitive and often regurgitated concept that young players need time to adjust to the league is usually referenced in tandem with data that merely shows a positive relationship between increased minutes played and the almost inevitable increases in raw per game numbers. This takes the analysis one step deeper by focusing on efficiency. Interestingly, the basic idea holds up beyond a simple more minutes-more production paradigm, as the data shows that most players also become more efficient performers over time. Here is a table that shows the raw improvement in efficiency by draft position:
*Note: Chart is color-coded with the highest values in green and the lowest values in red. Bold values represent players who have a 1st qualifying season PER of 15 or greater. Underline values represent players who have a 1st qualifying year below 15, but have a career best qualifying PER of 15 or greater. Bold-border values represent players who have a career best qualifying per of 18 or greater.
(2) Very few players start their careers as above-average performers (these players are bold in the table above).Only 34 players (30% of the player pool) had a first qualifying season PER above 15. GMs and scouting departments do a great job of drafting these more complete prospects with the highest picks. This finding is in line with what we generally know about the draft process, as the players with more flaws or material deficiencies drop further in the draft all the time. The interesting thing to note is that the scouts and GMs are fairly accurate in identifying who the more complete players are before they step on an NBA court. As top picks typically go to the worst teams with incentives to play these picks big minutes early in their career, the fact that the most productive players still come earlier in the draft is a testament to GMs actually grabbing the players with the most advanced skill level. Here is the breakdown:
47.1% (16/34) were taken in picks #1-4
23.5% (8/34) were taken in picks #5-8
20.6% (7/34) were taken in picks #9-12
8.8% (3/34) were taken in picks #13-16
(3) However, there is still hope for these other players to become above-average performers at some point in their career. Of the remaining 78 players who had a first qualifying season PER below-average, 42 of those players (54%) had a career-high PER above-average (these players are underlined in the table above). Again, GMs and scouting departments do a great job of drafting the players that become above-average performers with the highest picks.
33.3% (14/42) were taken in picks #1-4
35.7% (15/42) were taken in picks #5-8
16.7% (7/42) were taken in picks #9-12
14.2% (6/42) were taken in picks #13-16
(4) While most players become more productive over the course of their career, very few players become drastically more efficient. Only 38 players (34%) posted a career-high PER at least 5 points higher than their first qualifying year. If you recall the PER value reference guide in Part One, an increase of 5 or more PER points represents a serious improvement. It can mean anything from an end-of-the-bench player becoming an above-average performer, to a solid 2nd banana becoming a bona fide All-Star, to a borderline All-Star becoming an MVP candidate. Here is the breakdown by draft slot:
34.2% (13/38) were taken in picks #1-4
29.0% (11/38) were taken in picks #5-8
26.3% (10/38) were taken in picks #9-12
10.5% (4/38) were taken in picks #13-16
- *Note: If we restrict the previous point to players who had a career high PER above 18 (these values are bold-bordered in the table above), which equates to a baseline of a Solid 2nd Option and eliminates all the players who set very low standards their first qualifying season, the pool is reduced to 29 players, and the breakdown is as follows:
44.8% (13/29) were taken in picks #1-4
27.6% (8/29) were taken in picks #5-8
20.7% (6/29) were taken in picks #9-12
6.9% (2/29) were taken in picks #13-16
(5) The eight players who have made the largest improvements from their first qualifying season are not high-risk projects, and instead they are eight of the biggest stars in the NBA.
#1 LeBron James PER improvement of 13.4 (from 18.3 to 31.7)
#2 Dwyane Wade PER improvement of 12.8 (from 17.6 to 30.4)
#3 Amare Stoudemire PER improvement of 11.4 (from 16.2 to 27.6)
#4 Kevin Durant PER improvement of 10.4 (from 15.8 to 26.2)
#5 Chris Bosh PER improvement of 9.9 (from 15.1 to 25.0)
#6 Dwight Howard PER improvement of 8.8 (from 17.2 to 26.0)
#7 Russell Westbrook PER improvement of 8.4 (from 15.2 to 23.6)
#8 Chris Paul PER improvement of 7.9 (from 22.1 to 30.0)
Conclusion: I come away from this additional study with a few thoughts to pass along. Regardless of his level of production during his first season of significant play (50+ games, 12+ min/gm), you can reasonably expect a player selected in the lottery to improve his productivity during the course of his career. Whether due to the quality of coaching, adjustment to the NBA game, the hard work of players to improve aspects of their game, or working their way into roles that accentuate their talents and minimize their faults, lottery players do generally become more productive as they gain NBA experience. This is a comforting thought for franchises in need of quality players, especially when considering so few players start their career as above-average performers. However, the top 8 picks are still the most valuable spots to draft these project players because few players drafted outside the top 8 ever transform themselves from below-average to above-average players. Especially when drafting in the late lottery, it seems to be a better strategy to target players with more experience and well-defined talents and skills (meaning guys with higher initial baseline production levels, but possibly lower ceilings), because very few late lottery players ever make substantial improvements over the course of their career. The red-flags that push high-upside players into the late lottery are likely the same limitations and flaws (whether character-related, injury-related, or skill-related) that prevent them from ever making the leap to become good or great players. Finally, it is interesting to note that the players who have shown the most improvement over their career are some of the best players in the NBA. Sometimes we think of these players as being great from the moment they step on the floor but nearly all of the players listed above actually turned in a season barely above-average before rising to NBA stardom. Hope springs eternal when it comes to the NBA Draft, and while we have tempered some of the pie-in-the-sky projections with hard evidence, I think this supplementary study shows there are certainly still reasons to remain hopeful as the draft approaches.
Looking at the roster for Bucks draft picks from past seasons, I expect the biggest improvement in PER for the upcoming year to be from...
Larry Sanders (55 votes)
Brandon Jennings (26 votes)
Luc Richard Mbah a Moute (0 votes)
Ersan Ilyasova (4 votes)
Andrew Bogut (25 votes)
110 total votes