On Tanking

April 9th, 2013 by Nate Smith

Does a team always have an obligation to play its best players on any given night?  At what point does a team decide that playing it’s younger players in order to build experience for the future trumps its need to win the game that evening? What if a team is made up of barely qualified NBA players who probably have no long term future with the team for which they’re playing?  The Cavaliers were built that way at the end of the season last year.  Is not fielding a team that has any chance of being competitive a violation of sporting ethics?

The NBA has become a place where losing is rewarded.  As a team loses its odds of getting a lower draft pick become higher.  To prevent teams from losing on purpose to better their draft position, the NBA instituted a lottery.  Starting in 1985, the first three picks of the draft were determined randomly, first by drawing envelopes out of a hopper, and then starting in 1990, according to a number of ping pong balls.  After 1993, when Orlando had a 41-41 record and still won the lottery, the rules were changed to favor bad teams even more.

The problem with the draft lottery is that it provides incentives to fail.  It can be argued that these incentives are are antithetical to the concept of competition, fair play, and trying one’s hardest.  The term “tanking” has been coined to describe the process where “competitor deliberately loses without gambling being involved.”  Why is tanking so different from point shaving, which is one step better than fixing a game?  I imagine that the giant of American baseball, commissioner Kenesaw Mountain Landis would have had something to say about tanking, or even the appearance of tanking.

Regardless of the verdict of juries, no player that throws a ball game; no player that undertakes or promises to throw a ball game; no player that sits in a conference with a bunch of crooked players and gamblers where the ways and means of throwing ball games are planned and discussed and does not promptly tell his club about it, will ever play professional baseball. Of course, I don’t know that any of these men will apply for reinstatement, but if they do, the above are at least a few of the rules that will be enforced. Just keep in mind that, regardless of the verdict of juries, baseball is competent to protect itself against crooks, both inside and outside the game.

I suppose that since there is no grift involved against a perspective gambler, tanking is a step above match fixing and point shaving.  Furthermore, it can be impossible to tell if a team is merely losing because of circumstance, because of effort, because of substitutions, or because of holding players out of games when they could be playing.  David Stern was reportedly livid earlier this year when Gregg Popovich sat Tim Duncan, Tony Parker, and Manu Ginobili against Miami, because “the Spurs did a disservice to the league and our fans.”  But little was made of a game in March, when the Phoenix Suns played the Utah Jazz, and sat out Goran Dragic for “rest.”  Of course Phoenix lost that game 103-88. Why was that game important?  Because Utah, of course, is batting the Lakers for the final playoff spot, and Phoenix receives the Lakers’ first round pick this year, if the Lakers do not make the playoffs.  So Phoenix did its part to help Utah keep the Lakers out of the playoffs by sitting arguably its best player, Goran Dragic.  Of course, this can’t be proved.  It’s a wink wink / nudge nudge situation.  And that’s the whole problem with tanking.  I doubt that Landis would think too highly of Phoenix’s actions.

A further problem with tanking is that it erodes competitive balance.  In theory, every team in the East and West conferences has an equal schedule.  But teams that have more “bad” teams at the end of their schedule have an advantage against teams that have those same teams at the beginning of the season, if those teams are trying to lose.  Is it even possible to tell if teams are tanking?  I have become more sensitive to it in the last few years.  We basketball fans have all become more aware of the issue because of current articles like yesterday’s Plain Dealer which illustrated the draft implications of the Cavs recent two day win streak or yesterday’s SB Nation Draft lottery watch subtitled: Magic threatening to outsuck Bobcats.  Grantland’s Brett Koremenos explored this idea in Grantland last month with his article titled Solving the Real Problem with the NBA’s Tanking Epidemic. Bill Simmons and Malcom Gladwell discussed the topic in a 2009 series of letters, with Gladwell summing up the tanking problem as well as anyone.

You simply cannot have a system that rewards anyone, ever, for losing. Economists worry about this all the time, when they talk about “moral hazard.” Moral hazard is the idea that if you insure someone against risk, you will make risky behavior more likely. So if you always bail out the banks when they take absurd risks and do stupid things, they are going to keep on taking absurd risks and doing stupid things … If you give me a lottery pick for being an atrocious GM, where’s my incentive not to be an atrocious?

Furthermore, when front offices become incentivized to fail, they have a few different ways of doing it.  They could tell the coach to play young players, or to put them in bad positions.  The coach could make baffling decision that were in direct violation of common sense if one wanted to win the game.  I wrote semi-sarcastically of Scott and Tyler Zeller in loss to Boston a couple weeks ago, “[Zeller] finished 5-6 for 11 points and 9 boards in 24 minutes.  In a masterful move Scott left him on the bench for much of the fourth quarter, knowing that his play might turn the game in the Cavs’ favor late.”  The problem compounds itself when the players realize what is going on and stop giving maximum effort.  If Cavs observers are reading the tea leaves, this moment might have come in an awful loss to Brooklyn a few nights later.

But is this fair?  There is an unwritten code in sports: try your hardest.  If Byron Scott and the Cavs organization is not trying their hardest, why should the players?  Is it even a fair observation?  Are we as a society so jaded that we see conspiracies even in our trivial pastimes?  Are we simply confusing fatigue, injury, and normal human behavior (i.e. incompetence) for a conspiracy to lose?  In examining those factors, I decided to do a quick experiment.  My hypothesis was, if tanking has gotten as bad as it seems, then we should be able to see the results of it.  I charted a couple things over the course of the last 26 seasons.  First, team winning percentage.  My hypothesis was that if we’re seeing record tying winning streaks by teams like the Heat, is this partially because of tanking?  If so, then the winning percentage of all the non-playoff teams ought to be going down over the last few seasons.  The results surprised me.

Remember that the Charlotte Bobcats were added to the league in 2005.  Winning percentage of non playoff teams went up that year and the two years after, and then dropped the next three years.  The Raptors and the Grizzlies were added in 1995, causing the number to drop in the 1996 and 1997 seasons.  This is actually counter-intuitive because as the pool of non playoff teams grows, one would expect an “averaging effect” to push the winning percentage of the losing teams up.  The Hornets and the Heat were added in 1988 and the Timberwolves and Magic in 1989.  The winning percentage increase after these years indicates this effect.  But what is clear, is that the last three seasons have seen slightly more competitive non playoff teams than the previous three seasons.  This season is certainly no outlier when it comes to non-playoff team winning percentage.

But winning percentage is certainly not the best barometer of how good a team is.  There’s a lot of noise in it.  Many NBA statisticians have long preferred point differential as a barometer of team quality.  Basketball-reference has a normalized stat called Simple Rating System SRS which takes into account point differential and strength of schedule to come up with a number that is slightly better than point differential as a barometer.  This normalizes the number a bit giving us the ability to compare teams in the two different conferences.

As can be seen, SRS took a big jump in 1998 and 1999 , and then dropped quickly.  The 2008 season was particularly bad for SRS, two years after the the Bobcats joined the league.  But once again, the current season is a slight uptick, and the last few seasons don’t seem like statistical outliers at all.  But if there is tanking going on in the NBA, it was at its worst in 1988, 1994, and 2008, and  then has been around the same level since, on average.  Or there were just a lot of really bad teams those years.

There are certainly a lot of limits to this analysis.  This averages the best and worst teams that didn’t make the playoffs into one group.  An analysis that breaks the non playoff teams into tiers and analyzes those tiers over time would be a more precise way to measure if tanking is going on.  Also, looking at winning percentages post all-star break would be an interesting method as well.  We could even start looking at post all star break injuries and correlating them to average games missed, and seeing if “tanking” teams are holding their players out too long.  I hope to be able to break this down a little more in the future.  And I know there are mathematical implications that come from having a finite number of available wins and losses in a season that I am not nearly bright enough to have contemplated yet.  But what we have shown is that this season is no worse than the last few, and if tanking is going on, it can’t be detected this easily, or it is much more ingrained in NBA culture than we’d like to suspect.

I would like to see the league take steps to eliminate tanking.  I’d like them to redistribute the lottery percentages a little more evenly.  The league overreacted to the Magic in 1994.  I also proposed a system in the past that would disallow a team getting the first pick to get it the next year.  A team that picked in the top three two years in a row would not be able to get there a third.  Similarly, if a team picked in the top give for three years in a row, I’d like to see the best pick they could get the following year to be a number six.  A team that has been in the lottery four years in a row ought not be able to get a pick higher than ten, and a team that has been in the lottery five years in a row ought to have to sit at the end of the lottery for a year.  Of course this is just a framework, and these numbers can be tweaked, but you get the idea.  Don’t over-reward teams for losing.

This season and the Cavs don’t seem any worse than the last few when it comes to tanking.  I don’t like to think that people pick and choose when it is most advantageous to play hard, or when losing might be OK.  And as much as any Cavs fan, the last few seasons have worn on me.  It is a difficult situation to be in when the choices for explaining ten game losing streaks are incompetence, laziness, injuries that may or may not be real, or losing on purpose.  As painful as the Cavs might seem in moments like the Brooklyn loss, those moments are uplifted by jubilant victories over teams like the Clippers, Oklahoma City, Chicago, and Boston.  I hope very soon that I won’t even have to ponder the question of tanking at all.   Such is the hope of fandom, especially in Cleveland.  There’s always next year  — well, unless the Cavs are in a playoff race against a team with a lot of tankers on its schedule.