Losing Is Not a Winning Strategy in the NBA

Mark Cuban, Dallas Mavericks owner and "tanking" proponent.  (Photo: sdk)

Mark Cuban, Dallas Mavericks owner and “tanking” proponent. (Photo: sdk)

The NBA season is beginning this week and fans of each team are, of course, optimistic. At this point, everyone can hope a title is possible come next summer. 

Although everyone could theoretically have dreams of a title in 2014, it is clear that every NBA fan isn’t actually hoping their team is successful in 2014. Some NBA fans are actually dreaming of an event that happens just after the conclusion of the NBA Finals.  For fans of a few teams, the focus is already on the 2014 draft.  For example, some fans of the Philadelphia 76ers seem convinced that not only are the Sixers not trying to win this year, but that this is actually the best course of action for this franchise.

Proponents of “tanking” dream of such number one picks as Shaquille O’Neal or LeBron James. Each of these players were selected number one and went on to win multiple NBA titles.  Of course, other number one picks – like Yao Ming, Michael Olowokandi, Allen Iverson, Joe Smith, Glenn Robinson, Chris Webber, Larry Johnson, etc. –  played their entire careers and never won an NBA title.

Despite this list, fans of the NBA’s losers still have dreams that success in the NBA lottery will lead to title contention and an NBA championship in the future.  But is this likely to happen? About a year ago, I offered the concept of the “lottery treadmill” in an effort to understand how title contenders are built.  Now I want to address what happens to teams that are not contenders.  

Or more specifically: if a team wins 25 or fewer games – a result needed to maximize success in the lottery – what happens in future NBA seasons?

Before answering this question, let’s make an observation.  Since 1985, only two teams (the Miami Heat in 2006 and the Houston Rockets in 1995) have managed to win an NBA title without winning at least 66 percent (54 wins in an 82-game season) of their games.  And since 1984-85, about 20 percent of teams have won 54 or more games.  So it seems likely that a team needs to be in this group to really be considered a contender. 

But it appears that teams that win 25 or fewer games have a hard time joining this elite.  Of the teams that won 25 or fewer games since 1984-85,

  • 2.3 percent won 54 or more games the next year
  • 3.9 percent won 54 or more games two years later
  • 5.7 percent won 54 or more games three years later
  • 10.1 percent won 54 or more games four years later
  • 10.6 percent won 54 or more games five years later

In sum, nearly 90 percent of teams that win 25 or fewer games are not contenders five years later.  This suggests that “tanking” is a strategy that is very unlikely to lead to NBA success.

Of course, some may say that this is still better than just being “mediocre” in the NBA.  Back in 2011, three different NBA executives made a similar observation:

Towards the close of yesterday’s basketball analytics panel, Mark Cuban (owner of the Dallas Mavericks) and Kevin Pritchard (general manager of the Portland Trail Blazers) showed their cards in terms of fast-tracking a franchise rebuilding project. 

Cuban confessed that once Dirk Nowitzki retires he expects the Mavericks to lose, and, if he gets his way, they’ll lose badly. Kevin Pritchard seemed to agree and introduced a new term into our lexicons: “the mediocrity treadmill.” 

There is no championship future for a middling team that is stuck in the embattled space between those who struggle to make the playoffs and those that struggle and miss. Cuban has no desire for the Mavericks to be such a team. Charlotte Bobcats owner Michael Jordan recently defended trading Gerald Wallace to the Portland Trailblazers by saying, “We don’t want to be the seventh or eighth seed.” The Bobcats have been, at best, mediocre, and so perhaps we can interpret his statement as one owner casting his philosophical lot with Cuban and Pritchard. 

So are teams better off avoiding the “mediocrity treadmill”?  Let’s define a mediocre team as one that wins between 34 and 49 wins.  Of the teams in this group,

  • 9.1 percent won 54 or more games the next year
  • 13.9 percent won 54 or more games two years later
  • 14.8 percent won 54 or more games three years later
  • 16.5 percent won 54 or more games four years later
  • 19.8 percent won 54 or more game five years later

In sum, a team that is mediocre is much more likely to contend in the near future than a loser.  And that means if your team is actually trying to build a loser (i.e. avoid the mediocrity treadmill), they are reducing their chances to contend.

Now some might argue that this next draft is different.  This next draft is supposed to have such players as Andrew Wiggins, Jabari Parker, and Julius Randle.  Each of these players are supposed to be stars.  Of course, such players have only “starred” so far in high school.   So it is possible that these players will not excel in college or the NBA.

But let’s imagine these players are like LeBron.  It is important to remember that LeBron never won a title with the teams that acquired his services on draft night.  In fact, in the lottery era (since 1985) only the San Antonio Spurs (with David Robinson and Tim Duncan) have drafted a player number one and won a title with that player.  Every other number one pick failed to bring a title to the team that “won” the lottery.

So we see that the NBA draft lottery is quite similar to the state lotteries so many people play.  Winners of these lotteries often don’t get the life of their dreams.  And the odds of winning are so poor that state lotteries are often described as a tax on the mathematically illiterate.  Despite all this, people still buy lottery tickets in the hopes of realizing their dreams.  And likewise, some people in the NBA have dreams that winning the NBA lottery is the path to a future NBA title.

But it is important to remember as fans that these are often just dreams.  Losing is not a winning strategy in the NBA.  And if you see your team lose frequently this next season, you shouldn’t think that unhappiness experienced today is likely to lead to much happiness in the future.


David Levine

I think you are missing a few distinct selection effects. In addition, you'd need to investigate whether raw win total is a good predictor of likelihood of winning a title. You may find that title contenders have something in common that is different than just 54+ wins.

Michael

Yup, not all wins are equal when it comes to differentiating between a playoff also-ran and a true title contender.

Martin van Elp

There is a false comparison in there. You cannot just compare a team that is intentionally losing with all losers as you do here "And that means if your team is actually trying to build a loser (i.e. avoid the mediocrity treadmill), they are reducing their chances to contend." You can only do that if all losers are assumed to lose intentionally. A better (but still imperfect) check would be to compare intentional losers with mediocre teams in later years.

In more detail: losing normally signals a bad team and probably overall weakness (accomodation, staff and so on). So you expect them to perform worse in later seasons too than a mediocre team, because a mediocre performance signals having a better team, accomodation, staff and so on. However, if one starts to lose intentionally, that losing no longer tells us implicitly that the team is bad, so the signal 'losing' is not what we expect from a mediocre team to signal. I have doubts too about intentional losing, but you cannot compare a mediocre team (intentional loser) + number one with all losers + number one.

Read more...

billy

There is no real comparison between the two. If you ask a professional athlete to intentionally lose, they would be poised to say no. The only way to intentionally lose is to make your team extremely bad by trading your good players away for draft picks. The teams that get the players will benefit immediately, while the team that gets a draft pick will bet on the next few seasons. Either way, the team is bad simply because they have bad players. If you had one superstar to a core of terrible players, you still won't win. You might contend for a while, but you won't win. Look at Lebron, he didn't start winning until he went to Miami.

Michael

A key factor that doesn't seem to be accounted for in this piece is the manner in which teams purge bad contracts.

In the NBA, where free agent acquisitions and trades are as often about cap space as they are about talent, teams that are tanking or trying to get to the championship have to align not only the talent of the team, but their cap situation so as to afford the players, either on their cheaper rookie deals or their supermax free agent deals.

A much more complicated and detailed regression analysis would be needed to make a more accurate assertion regarding the wisdom of "tanking"

Dave

"But let’s imagine these players are like LeBron. It is important to remember that LeBron never won a title with the teams that acquired his services on draft night. In fact, in the lottery era (since 1985) only the San Antonio Spurs (with David Robinson and Tim Duncan) have drafted a player number one and won a title with that player. Every other number one pick failed to bring a title to the team that “won” the lottery."

I think the assumption is that the tanking team will build a better team around the newly-drafted superstar than the Cavs did. Players like LeBron, Dwight, Yao, Durant (#2 pick), Rose and Irving greatly changed the fortune of their respective franchise. Their teams may have not won the championship, but they were much much closer to winning it than before. It's a superstars' league, and for a lot of teams, the draft is the only way of obtaining them, especially in the era of superstars joining forces via trades/free agency.

Read more...

Bayan

It would be interesting to see this analysis with revenue as the measure, not wins or championships. I wonder if tanking to get a first pick results in higher revenues over a 5 year period. Believe it or not, most NBA owners are more concerned about that than winning.

Nil

The big issue for NBA teams is unless you have one of the top 5-10 players in the league you have almost no hope of winning the title. With 30 NBA teams that consistently leaves 2/3rds of them without a realistic shot every year.

A team can get those superstars either through the draft or through free agency. However top superstar free agents very rarely choose to go to teams that are not already title contenders or that aren't in top markets (LA, NY). So if you are the GM of Milwaukee, Utah, or Charlotte you are unlikely to ever attract a top 5-10 player in free agency unless you already have one, thus putting all your hopes in the draft is the only rational decision.

The draft maybe a crapshoot, but #1 picks do turn into elite players at a far higher rate, and picks 2-5 also have far better outcomes than the 15-16th pick that a mediocre team will land. One flaw in a modern draft analysis would be that in todays 18 year old players are banned from entering and there is far better scouting of foreign players. Several of the current & last generation of superstar players came into the NBA as lesser known and higher risk high schoolers or foreign players who could slip down the draft boards. Had teams had to wait an extra year for them to become more known quantities in college or international competitions many of them would have required much higher lottery picks to select.

Read more...

Devin Lavelle

First of all, by limiting your statement to "in the lottery era", you're essentially making a distinction without a difference and severely limiting your sample size. Every team that won a championship in the 1990s was lead by a player they drafted with one of the top 3 picks in the early 80s.

Which brings me to my second point, being a bad team can also get you the second or third pick ... where very good players are often found.

And my third point, there aren't very many championships to go around, especially in the NBA, which tends to see a lot of runs. Many of those #1 picks that didn't win a ring did contend for them. (Iverson, Yao, C-Webb, LJ)

Every team that has won a championship in decades has built around a great player (or players) they acquired through the draft. They have usually been drafted in the top handful of picks, with a couple of notable exceptions (Kobe and Dirk, who had significant question marks and were gambles that clearly paid off, that's the nature of the draft).

Most of these teams (Bulls, 1990 Pistons, Rockets, Spurs, Mavs, Heat) were quite bade before drafting the great player(s) that helped bring them to the promised land. A few (Lakers, 2004 Pistons) managed to build the team without sinking below mediocrity.

Bottom line. You don't win in the NBA without great players. Acquiring great players through free agency requires you to have already built a good team, while retaining cap space (which is rather challenging). The only other way to do it -- and the way most championship winners have gone -- is by drafting them (almost invariably early in the draft) and then adding aggressively to the team.

Read more...

Matt

A team that wins 49 games is being considered mediocre? That's a .600 winning percentage or good enough to probably host a playoff series. Mediocre teams usually win 35-44 games.

bob

Cuban won his title, now he just wants the Mavericks to use as little of his money as possible, and he's using the top draft pick as an excuse to prevent Mavericks fans from dumping the team.

Jamie

I agree that the selection bands are a little arbitrary.

- 25 or less: really bad
- 26-32: bad but not the worst
- 33-44: mediocre in the Pritchard vein

Anything 45+ is generally acceptable overall, and not usually worth breaking down unless everyone got old/injured/left.

49 wins can get you home court in the East many years.

But I'd expect that a lot of that 54+ future upside of mediocrity came from the teams in the 45-49 range in year zero.

Dan Murphy

Expect a lot better than this article from Freakonomics. There are so many holes and flaws in the article that it would take more words than the article itself to go through them all. With that in mind, I'll resort to all the other comments below that do it for me. Most are logical, and most are correct... damn, that was a bad article with a poorly constructed argument.

Julien Rodger

Great article Dave, hard to argue with those numbers

I will say this. The fact that 'tanking' teams don't go onto contention, doesn't necessarily mean that tanking is a poor strategy. To me it's more likely that the teams who lose that much did so because they were poorly run. And if you're poorly run, for the most part it doesn't matter if you tank or not, you're going to fail regardless. The 'under 25' and 'over 25' Ws stats are skewed by the fact that the stupid teams like Washington, Charlotte, Sacramento, Minnesota, etc. are the ones tanking and a lot of the smart teams like Spurs, Utah, Houston, Indy etc virtually never tank. I bet if the Spurs and Rockets committed to tanking and picking in the bottom 5 for a few years, their success rate may be alright. Likewise it's possible that for a team with terrible management, tanking is actually a better idea than 'trying to win', in that the latter ends up even more terribly. As a Raptors fan where we really got burned under Bryan Colangelo was when he tried to make a playoff team by trading for Jermaine O'Neal, overpaying Turkoglu, trading for Gay, etc. When his entire job was to lose and draft Valanciunas and Ross the team at least added to their assets and got themselves in a better position. In other words, tanking is 'easier' than trying to build like the Morey Rockets have been, so if you have a terrible GM you may as well tank. But the real correct strategy is just to have good management, no matter your plan.

Read more...

Matt

Typical freakonomics click baiting through pathetically ignoring obvious problems in casual logic. Assuming that in the past almost no teams have purposefully tanked, the comparison of future win rates between horrible teams and mediocre to solid teams gives zero information about the casual question at hand. If your team talent, coaching, and management is good enough to be a five hundred team but you spend one season purposefully losing games to get the next Jordan, we have no idea from historical data how effective that strategy could be.

truzo

Are you guys recycling posts? Wasn't there an entry on statistical analysis of the lottery and draft picks compared to winning towards the end of the last NBA season? Picking right back up...

Alex in Chicago

Sigh.

This guy keeps posting this, but doesn't understand that the people who advocate in favor of the strategy don't care about winning 54 games. In fact, they only care about winning 4 games, at the end of the season, in a NBA finals.

Recent history suggests 3 ways of doing that:

#1. Drafting an elite player #1, and surrounding him with talent that is severely underrated by other teams in the draft (Spurs)
#2. Drafting an Elite player and trading for/signing in free agency an elite lottery pick player to pair with that player (Heat/Lakers).
#3. Draft well, make smart trades, compile several elite veterans, and then mortgage the future for a short window to win a title (Celtics/Pistons).

The 3rd model is not only the hardest, it is also the least profitable because you have no long-term elite talent, you just get a few good years out of the Pierce/Garnett/Sheed type players.

Your analysis basically discounts the fact that most teams are bad because they have bad management, thus they don't even draft well, or make good trades, and that to make a true contender in the NBA you need to have more than a 54 game win season, you need a lottery pick who has ascended.

Read more...

Voice of Reason

The author of this article is asking the wrong questions, and supporting his argument with data that proves correlation, not causation. Of coarse teams that don't do well won't do well in the future. But the question we are asking is if you take a given team, and are not given information on their actual quality, would they be better off in five years if they chose to lose most of their games on purpose as if they played their hearts out.

Another interesting argument that was failed to be brought up is accountability and who is rewarded and punished for the performances. Head Coaches and GM's are hired and fired based on their records. They have no incentive to tank a team so that somebody else can be brought in to use their #1 overall pick. Similarly, players have no incentive to let their stats fall, have their replacements drafted, get cut and not have a job.

Only the fans and owner could possibly benefit from intentional tanking (if you agree that strong play-off success in the long-run outweighs consistent mediocrity). Unfortunately, they have a limited impact on who wins and loses once the season is underway (it probably wouldn't go over well if the owner tried to talk the team into losing).

Read more...

Big Dog

Glenn Robinson won an NBA title in 05

Patrick Conway

Glenn Robinson did fall in to a ring in 2005 as a total bench warmer.

http://www.basketball-reference.com/players/r/robingl01.html