Let’s say your team is a loser. If your team played in a European league that would be a problem. European sports leagues have a system of promotion and relegation, which means the losers each season get demoted into the sports minor leagues and the best of the minor leagues are promoted to the majors. In other words, in Europe – where socialism is not exactly a bad word – losers are punished. In the United States – where winning is supposed to be everything – losers in professional sports are rewarded with high draft choices.
Now if you are a Detroit Lions fan – such as I – the reverse-order draft is a good thing. Every April the Lions get a fair amount of attention from the national media, as people wonder whose promising career the Lions are going to seek to ruin next.
In basketball the story is a bit different. In the NBA one outstanding talent can have a dramatic impact on a team’s fortunes. Consequently the reverse-order draft gives teams a clear incentive to avoid winning when the playoffs are no longer possible. The latest team to follow this strategy is the Minnesota Timberwolves, who are sitting Kevin Garnett now that the road to the playoffs had hit a dead end.
Talking about who a team will take or why losers do not do their best at the end of a season is certainly fun. But in all this talk we tend to ignore the larger issue. Why do North American sports leagues employ a draft? Let me spend a moment on how the leagues justify this institution, and then I will discuss why I think this institution was actually enacted.
The Competitive Balance Argument
According to the leagues, the draft is designed to encourage parity. As such, it’s part of an arsenal leagues employ to promote competitive balance. In the NBA’s arsenal we not only see a reverse-order draft, but also a cap on payrolls, a cap on individual salaries, and revenue sharing. All of these institutions are designed to promote competitive balance in the NBA. Unfortunately, the data suggests that these institutions don’t really work very well.
To see this, consider the level of competitive balance in the Association. Last year the standard deviation of winning percentage in the NBA was 0.136. The idealized standard deviation for the NBA – given its schedule of 82 games – is 0.055. So the ratio of actual to ideal – a measure we call the Noll-Scully (named after Roger Noll and Gerald Scully, the two economists who originated this metric in the 1980s) — was 2.47 for the NBA. Prior to last night’s games the standard deviation of winning percentage in the NBA was 0.134. So the ratio this year currently stands at 2.42.
For the NBA to have an actual standard deviation that is more than twice the ideal is hardly unusual. For the past 27 years, the NBA’s Noll-Scully has been more than two every single season.
To put these numbers in perspective we looked at the level of competitive balance in a collection of sports leagues in soccer, football, hockey, baseball, and basketball. The average ratio across all leagues is 1.9, but in recent years most sports have offered a ratio below this mark. For example, the NHL has only had 3 years with a ratio at 1.9 or higher in the last 20 years. In baseball, the American League has been at 1.9 or higher nine times in the last 20 years while the National League has equaled or exceeded the 1.9 mark only five times across the same time period. And in the NFL, we have not seen a ratio equal to 1.9 since 1944.
In sum, despite a reverse order draft, caps on payrolls and salaries, and revenue sharing, competitive balance in the NBA tends to be much worse relative to other sports leagues. This is not only seen in the disparity in regular season winning percentages, but also in the distribution of championships (as detailed in The Wages of Wins).
The Short Supply of Tall People
The problem for the NBA – as we said again and again in The Wages of Wins – is the short supply of tall people. The NBA takes its athletes from a very tiny population, specifically very tall people. As Stephen Jay Gould noted twenty years ago, when your population of talent is small, the variation in athletic talent will tend to be higher. Consequently, when baseball began to integrate in 1947, the talent pool expanded and the level of competitive balance improved. Unfortunately for the NBA, the talent pool cannot be easily expanded. There are just so many really tall people who are good at basketball. Hence some teams have access to really good tall people – like Michael Jordan, Shaq, Tim Duncan, Dirk Nowitzski – and others are left with less talented big men. And when the talented play the less talented, the results tend to somewhat predictable.
Why Have a Draft?
A draft is not going to manufacture additional Greg Odens and Kevin Durants. Two teams are going to get these players and probably improve. Others will get less talented players and not do much better. The draft – as well as caps and revenue sharing – cannot change the number of truly talented basketball players that exist in the world. And consequently, these institutions do not have much impact on competitive balance.
So what then is the point of the draft? What the draft does is restrict the salaries paid to newcomers into the league. Initially it restricted who was allowed to bid on a player’s services. The past tendency of high draft choices to hold out led the NBA to institute a rookie salary scale, which eliminated the rookies bargaining power completely. And what did the elimination of bargaining power do for the NBA? It allowed teams to pay far less for the wins and revenue the top talents create early in their careers.
In sum, people have argued that the draft as evidence of socialism creeping onto professional sports. In reality, though, the draft is really part of the darker side of capitalism. The owners of capital (if I can get all Marxist for a moment) have created an institution that allows them to exploit their workers (i.e. pay the workers a wage that is below the amount of revenue the workers generate for the firm).
So should Greg Oden turn pro and begin his exploitation by an NBA team? Well, right now he is being exploited even more by the NCAA. So he might as well join the Association where at least he can be exploited less.
Yes, that’s my advice to Oden. “Turn pro and reduce your level of exploitation.” Not exactly a slogan destined for a bumper sticker some day.
Update: Doug Drinen at Pro Football Reference (JC Bradbury’s primary co-author) has authored a great post on why the NFL Draft should be abolished. Drinen makes a very good argument for why ending the draft is a good idea, although he fails to tell us what Lions fans are supposed to do in April. Given that my favorite time of the year as a Lions fan is the month of April, you can see why personally I am opposed to Drinen’s idea. As an economist, though, I can’t see a flaw in his logic. Thanks to Katie Gold for pointing out Drinen’s comments.
– DJ
Wages of Wins Journal 
Donald A. Coffin
April 17, 2007
It remains a mystery to me why no one has challenged the amateur player draft system in the courts. The players in these drafts are not represented by the pro sports unions, so it’s ridiculous to claim that the draft system has been negotiated on behalf of the new entrants. Clearly, drafts reduce the negotiating leverage of the draftees (particularly the most talented among them), reducing their compensation, in most cases, for years to come. (In large part, this comes through reductions in signing bonuses.)
But it does make the owners wealthier, and, I suppose, that’s in the national interest.
Owen
April 17, 2007
Members of the players union also benefit from the current draft system . The labor agreement represents a division of the spoils between current players and management. You can’t just blame the owners, veteran players are also looking to maximize their pay at the expense of rookies who are not yet members of the club. Paying Oden would mean taking the money out of the pocket of other big men in the league. Or at least that is how it seems to me…
Dave -In sum: competitive imbalance exists in basketball more than in any other sport. This is an unavoidable byproduct of the “short supply of tall people.” It has been this way forever. And the current arrangements don’t effectively address the problem as they are putatively meant to.
I buy all that. If you were the supreme dictator of the NBA isn’t there some measure you could implement that might improve matters. There seem to be about 90 players who have played over 1000 minutes this year and are over .150 in WP48. If there were distributed a bit more evenly couldnt you at least have a situation where you have ten teams at the top of roughly equal quality rather than three?
Basically, is competitive balance totally a lost cause, or is there something you think could be done about it?
Jason
April 17, 2007
League rules may have little to do with competitive balance as assessed by variation between the winning percentage of various teams in any given season, but I do not think that this is what everyone is talking about when they talk about competitive balance. I think for many, it is the perception (whether or not it is based in reality) that the same teams are always good and the same teams are always bad with far less concern with how many standard deviations around the mean it takes to be good or bad. I suspect that the angry fan who complains that there’s no balance in the game doesn’t mean that the winning teams are winning 80% of the time in a season, but means that the same teams are in the playoffs seemingly every year.
I do not think that any league wants a situation where in any given game, the prior probability of either team winning is equal. While that may be the extreme of competitive balance, that’s not what the leagues are aiming for. Rather, a reverse order draft serves to change which teams will be better and which teams will be not as good in subsequent seasons.
I agree that the available talent does have much to do with why in any season, competitive balance is what it is, but suspect that in basketball, it’s more than *just* the supply of tall people. The nature of the game, where scoring happens regularly and the contest is a product of many, many encounters such that chance is minimized means that random results are less likely to overly influence the outcome. A very low scoring game like soccer or baseball might rest on a single scoring event. When fewer scoring opportunities are realized, luck comes into play. The parametric ability (something we can’t actually know, but just assume for a moment that there’s an absolute measure of a team’s ability represented by some parameter rather than an estimate derived from a sample mean) reflects the outcome better when luck is minimized as it is when there are more scoring opportunities. In this regard, tennis, another sport where the game never involves a single score but is a product of many, many scoring encounters, should have a lower ‘competitive balance.’
Of course, this isn’t contrary to your explanation, and in fact works in concert as the team with the best tall players will have an even greater advantage since a couple of fluke events are less likely to change things. Any given game is more likely to fall out according to parametric abilities of the two teams.
Katie Gold
April 17, 2007
The guys at pro-football-reference.com make an excellent argument for abolishing the draft. The is exactly what you say it is — the anti-market system. Given the popularity of free agency, its amazing that the draft still exists today, since it is really a form of the reserve clause (from baseball). I think it would be extremely interesting if the NFL got rid of the draft, and you would certainly see the Salary Cap gurus on each team earning their pay. Check out the post here: http://www.pro-football-reference.com/blog/wordpress/?p=287
Okapi
April 17, 2007
How is the “idealized” standard deviation for the Noll-Scully determined? Is it just a binomial model with 30 teams playing 82 games and every game set at 50%/50%? Thank you.
Owen
April 17, 2007
Jason – What you said. Competitive imbalance within one season isn’t such a bad thing. The major problem is the inability of bad teams to become good even over the course of several years, i.e. my Knicks. I feel like in football at least success, i.e. playoff appearances and winning records, is spread much more broadly. As a football fan, hope springs eternal, it’s always reasonable to hope that you will have a winning season next year. Miracle turnarounds are possible, except for the Lions. In New York, however, we know the next two years at least will be miserable.
dberri
April 17, 2007
Okapi,
The ideal standard deviation is simply the average winning percentage divided by the square root of schedule length. For the NBA, that is 0.500 / square root of 82, which equals 0.055.
Jason,
Explaining the differences in competitive balance across sports by referencing the nature of the games being played runs into a problem when you consider how competitive balance has changed in baseball across time. In the first half of the 20th century, baseball’s level of balance looked like basketball. The game of baseball, though, was still the same. If it is the nature of the game that tells the story, then competitive balance in baseball shouldn’t be different in the first half of the 20th century relative to the last half of the century.
That being said, if it is the nature of the game that drives the level of balance, that still keeps the basic story we tell in place. League policies do not seem to drive the level of balance. Rather, league policies are enacted to take money from the players and give it to the owners.
Owen,
Although the NBA doesn’t have much balance, it doesn’t seem to matter much either. The NBA has grown in popularity over the last three decades despite having little balance. Although economists like me spend a great deal of time writing about the importance of competitive balance (I have ten publications on this topic), I have come to think that maybe balance isn’t really all that important. Even with the level of balance the NBA has, outcomes are still somewhat uncertain. So these games are still good to watch.
Katie,
Haven’t read the post, yet, but I would be opposed to ending the draft in football. As a Lions fan, this is my favorite time of the year. If you take away the draft, what are we left to look forward to as Lions fans?
dberri
April 17, 2007
Katie,
Okay, I read the post. Your right, that’s a very good idea. Ending the draft in the NFL would force teams to decide how they are going to build their franchise. As it is now, teams are almost compelled to add several rookies each season. Without the draft some teams could opt out of this market and build entirely with veteran free agents. Others would choose to purchase more young players. All in all, no more draft would be a better system.
That being said, I still look forward to end of April and seeing who the Lions select. And I start looking forward to this event sometime in September.
Brian
April 18, 2007
I’m not sure that this post demonstrates that the NBA’s rules are ineffective in supporting competetive balance. In an absolute sense, yes, but that may be unavoidable. But we can’t judge the effect the rules have on competetive balance without any idea of what things would be like in the absence of those rules. If the draft, salary cap, and max contracts were all abolished, it could conceivably be the case that the NBA’s Noll-Scully would soar even higher. In other words, the NBA’s rules may be doing a fine job of *minimizing* the Noll-Scully to the greatest (reasonable) degree possible, given all the other features of the game (dearth of tall players, multiple scoring events, etc.)
dberri
April 18, 2007
Brian,
There is some information left out of the post. In the book we looked at competitive balance in 15 different sports leagues. What we found is that within a sport, balance was the same. Between sports it was different. So all the soccer leagues were pretty similar. The ABA and NBA were similar. The NFL and AFL were similar, etc..
Some have said that is because it the structure of the sport that causes competitive balance. But we also find that within a sport balance changes over time. That suggests it is not about the nature of the sport.
In a 2003 Economic Inquiry publication, Marty and I did present evidence that it is changes in population that drove changes in balance in baseball.
So to summarize… we have three proposed causal factors for competitive balance. League rules, nature of sport, underlying population. If it is league rules, why do different leagues in the same sport have the same level of competitive balance (despite different rules)? If it is nature of the sport, how can competitive balance change so dramatically in baseball? The population story tells us both why balance would be the same in the same sport, and also why it would change over time.
Moronski
April 18, 2007
The draft did help the lower ranked teams more when the players being drafted played three to four years of college or became too good for college. The players had more experience and probably more importantly it was easier to figure out which players would be productive. However, because these players were being paid so much money it was catastrophic to draft someone high in the draft and have them flop. The contracts for players like Glenn Robinson, Larry Johnson, Donyell Marshall were getting out of hand and these guys weren’t worth the money they were being paid. It wasn’t that long ago that the first couple of draft picks were among the highest paid players in the NBA so the big contracts were corrupting. With the rookie scale, the players are going pro earlier so that they can cash in on their first big contract as soon as possible (post rookie scale) and have to work for it. This also means that the teams drafting them are really getting their development years instead of the player’s peak years so the draft isn’t necessarily helping to turn around franchises much. It also creates more risk since these guys are so young and unproven there is more risk which also helps some of the better teams who have higher draft picks and are better at developing players. Bottom line, I agree that the draft basically exists to restrict salaries and to help ownership from blowing money on unproven rookies like in the past and not necessarily to help promote competitive balance.
BadgerBucco
April 19, 2007
One important factor that plays off the scarcity is the free agent process. The team that has a player can always outbid any other team for that player’s services (provided they have enough cap space). The effect of this is that the scarce great players will rarely if ever change teams as free agents. Their teams will always make sure they have enough cap space to give the great player a max contract. This eliminates one of the three methods of talent acquisition when it comes to the true game-changing players.
Jason
April 19, 2007
Have the rules regarding free agents and the holding team’s abilities to outbid others actually cut down on player movement? Empirically, the number of “stars” who are with their original team seems to be few. I don’t think there’s much evidence that this league policy has actually resulted in teams being able to monopolize scarce talent.
It also doesn’t seem that the draft was much better at helping bad teams in the past either. When a seasoned college player was the rule at the top of the draft board in the 70s and 80s, the Clippers still managed to pick high and miss the playoffs for 15 consecutive years. Again, I’m not sure the conventional wisdom really matches the empirical data.
geebo
April 19, 2007
so what would the standard deviation of winning percentage be if there was no draft, or the order was random? how can you say the structure doesn’t work without looking at what would happen without it? maybe the standard deviation would go up to 4, in which case it is doing something.
Dynamire
August 6, 2007
I don’t like the conspiracy theory that these rookies are being exploited. Frankly, when rookies have bargaining power, we have year after years of Larry Johnsons where they hold teams to ransom. You might end up paying Greg Oden about the same as you might pay the combination of Jordan and Pippen in their prime.