Andres Alvarez is a software engineer. Arturo Galletti is an electrical engineer and statistician. Both are avid NBA fans that have been looking over draft data from the last thirty years. The following is an article that started as a long e-mail conversation and hopefully will be the first of several interesting articles on the NBA draft.
Some quick background
This article uses Wins Produced and WP48 [Wins Produced per 48 minutes] to evaluate player’s performance.* This measure uses three key components to evaluate a player:
- The player’s per minute box score statistics
- The player’s team’s per minute box score statistics
- The average performance at the player’s position (PG, SG, SF, PF or C)
A full explanation can be found here. To give a general scale, an average player has a WP48 score of 0.100. The very best players in the league usually have a WP48 over 0.300. To put this in perspective; an average player who plays a full season at 24 minutes a game would generate around four wins for their team. In contrast, a player posting a 0.300 WP48 would generate more than twelve wins in this time on the court.
Two for one special in the draft lottery
With measure in hand, let’s start with the story of two fans (who we will call Fan A and Fan B). Both of their teams did poorly last year and both fans want hope for the upcoming season. Fan A finds out that their team managed to draft a rookie who is believed to be a hot prospect. Fan B finds out that their GM has signed two average role players (i.e. two players who can post a 0.100 WP48]. Which fan should feel happier about their team’s upcoming season?
Many analysts and fans will of course point to Fan A. After all, the draft is a magical place that can change the fortune of any team. The bounce of a ping pong ball can turn a team from a bottom dweller to a contender, or so we are lead to believe.
However, the reality is much more sobering. Each year many rookies are considered “hot prospects”. But how many of these “hot prospects” actually become elite rookies? Again, an average player (with a WP48 of 0.100) that plays 24 minutes a game will generate around four wins a year for their team. What are the odds that our “hot prospect” can be twice as productive as an average player and generate eight wins? The answer from the data – revealed in the following tables — won’t make Fan A very happy.
In the past 33 years, only 64 rookies have produced more than eight wins his rookie season. In other words, only about two rookies per year reach this level. And that means the vast majority of rookies really don’t make a substantial impact on their new NBA team during their initial season in the Association.
Table 3: Elite Rookies from 1990-1999
Okay, the big story is that finding an impact rookie is difficult. Here are a few more stories we see as we look at the few players who actually were able to produce eight wins his rookie season.
The 1970s: The Start of the Draft Myth
As we do not have data for the full 1970s it is hard to give a full analysis of this decade. What we can note is that the current image of the draft may be drawn from this period. In our small sample of three years a total of twelve rookies that played twice as good as an average player were seen. Additionally, Magic Johnson and Larry Bird entered the league the same year and immediately managed to turn their respective franchises around. The experience with Magic and Bird might have led people in the NBA to believe that it was not that difficult to find a rookie who could turn a team into a championship contender.
1980s to 1990s: A Widening Market
The 1970s and 1980s were an amazing time for elite rookies. From 1977 to 1989, eight rookies entered the league and produced a WP48 mark that was three times the mark of an average player. In contrast, this only happened twice after 1990 (Shaquille O’Neal and Chris Paul).
To explain such differences we can look at what happened to the population of players the NBA could draw upon. In the 1970s and 1980s, the NBA primarily took its talent from the United States. In the 1990s, though, talent was taken from around the world. As the population of players available increased, the ability of any one player to substantially differentiate himself from the league average diminished (because there are now more and more good players in the league).
Given this observation, perhaps we should not be surprised to see that the quality of the elite rookies declined in the 1990s. Even with 130 more minutes per season, rookies generated almost two fewer wins for their teams, going from an average of 13.5 wins each to 11.1 wins each.
1990s to 2000s: A Youth Movement
In the 2000s we notice another trend. The number of elite rookies drops from eighteen or nineteen (what we saw in the 1980s and 1990s) to only twelve. And an elite rookie only appears in six years in the 2000s.
Beyond the impact of population, there is also the issue of age. Specifically, the age of the few elite rookies we see in the 2000s had decline. Prior to 2000, the elite rookies joined the league near their 22nd birthday. As a player’s peak is at age 24 or 25, it would seem older would be better. In other words, a player at 19 or 20 is not expected to produce as much as a player at 21 or 22.
Advice to Management
Consequently, given the impact of population and age, finding an impact rookie has become even more difficult. Of course, even when we look at the entire data set this is not an easy task. And it is still true if we just look at the top 10 picks.
Of the 64 elite rookies, 47 were drafted using a top ten pick. At first glance, it seems the draft makers may actually be able to spot good rookies. There are two observations, though, to be made.
First of all, there have been 330 rookies selected in the top 10 since 1977. So less than 15% of these rookies – who were generally considered “hot prospects” — have made substantial impact his rookie season (and only 13 of the 33 players chosen with the first pick). Furthermore – and perhaps more importantly – there is a clear bias when we look at the link between draft position and overall production. Published research (as detailed in Stumbling on Wins) has shown that there is a clear link between playing time and draft position (even after you control for player performance). Given this bias, we should not look at the aggregate performance of the top draft picks and conclude that these players were clearly better.
All in all, elite rookies are not commonly found in the draft. This is problematic for GMs of troubled franchises, who clearly would like to deliver quick results to their fans. The advice for GMs looking for quick results would be two-fold. First, a GM should lose illusions about the draft holding the immediate answer. Second, the GM should focus on other decisions that are more likely to help the team in the short-term, such as free agent signings and keeping any existing talent (Ahem! Minnesota and New York).
– Andres Alvarez and Arturo Galletti
* – the Wins Produced numbers discussed come from the calculations of Dave Berri. These will be posted on-line in the near future.
The WoW Journal Comments Policy