Skip to main content

NBA Analysis: Why Do We Continue to Overrate Closers and Closing?

Since Michael Jordan‘s last shot as a Chicago Bull, every pundit and blogger has obsessed over the need for a crunchtime scorer in the mold of His Airness. Kobe Bryant, Tim Duncan, Dwyane Wade, Paul Pierce and Dirk Nowitzki have all been lauded as exemplars of “Closers,” marking the difference between Pretenders and Champions. The thinking, so they say, is that great clutch offensive players are necessary to win a title.

But we know they are not. We know teams aren’t winning championships based on their play late in close games. We know good clutch offenses come from teams without superstars. And we know people have a massive Late-Game Bias that fuels most of this. But then why do so many championship teams seem to have these superstar “go-to” players on offense? Why do so many great teams switch to what’s commonly called Hero Ball at the end of close games?

There is simple explanation based on a logical fallacy at work here.

The Hero Fallacy: Winning teams usually play Hero Ball at the end of a game because they have a player good enough to be used as a Hero. They are not winning because they play Hero Ball at the end.

Teams are not successful on offense because of what they do in the final few minutes of their close games. They are good because they are outscoring the opponent during the rest of the game, and the teams that do so often have (at least) one elite offensive player who can also be used as a go-to scorer. This is a subtle, yet simple issue of causation.

Crunchtime doesn’t always start with a tied score. The other points in the game still contribute to a victory. Good teams can still be winning games not because of their crunchtime Hero Ball strategy  but they can be winning hordes of games even in spite of their crunchtime strategy.

Winning teams have been employing this Hero Ball strategy for years. The likely reason they do this is because they are paying stars $15 million, and because it’s a risk-averse strategy of simplifying the offense. Coaches won’t be questioned about “why the star didn’t take the big shots” and players don’t have to worry about overcomplicating the offense when they run an isolation for their star. What’s important here is that there isn’t much of a noticeable difference between “winning because of a closer” and “winning without a closer.”

Would You Know the Score if There Weren’t a Scoreboard?

Ask yourself that question. If you are so bold to think that the answer is “yes,” ask yourself if you know who has had the best clutch offenses in the games you’ve seen recently. Or in this year’s postseason so far. It’s a little more difficult now, isn’t it? Only it’s the same question, and it’s essentially an impossible one for a human being.

Clutch then becomes an easy explanation for why teams win. It’s easy to remember. It’s easy to write stories about. And what we remember is Hero Ball on winning teams…which leads to the Hero Fallacy. When Bill Simmons constantly perpetuates clutch myths, it is because he can mentally juggle those final 10 possessions. Those possessions are also most memorable because of our Recency Bias; We remember the last stuff we saw. We remember it so well that it grossly affects our perception of the entire event.

We are not objective computers designed to make sense of large data sets. To borrow from Simmons, we just aren’t designed that way. We just aren’t.

Why does it Seem Like so Many Successful Teams are Winning in the Clutch?

On a frequent basis, go-to scorers make shots in the clutch. Only they are doing so at a lesser frequency than they normally do throughout the game. The human brain cannot detect this difference over the course of hundreds of possessions. All teams are scoring less efficiently in these situations, but the naked eye and mind can’t decipher the differences between a 110 Offensive Rating in the clutch and a 90 Offensive Rating in the clutch, despite this being the difference between the best team in the league and the worst team in the league over the course of the year.

In reality, teams are no more winning “because” of Hero Ball than the Lakers won titles “because” of Shaquille O’Neal‘s free throw shooting. The correlation is incredibly strong between go-to scorers and winning teams just like the correlation from 2000-2002 was strong between “NBA champion” and “terrible high-volume foul shooter.” But the Lakers weren’t winning because of Shaq’s missed free throws and most teams aren’t winning because of their Hero Ball strategy at the end of games. This is the Hero Fallacy:

  1. No one has noticed hero ball is producing lesser results (see above).
  2. Because Hero Ball is often seen on a winning team, people falsely believe Hero Ball is good. Or even worse, Hero Ball is necessary to win.

Based on the distribution of players throughout the league, good teams almost always have good players. Good teams are on TV in the playoffs. These teams win close games, not necessarily because their offense is good down the stretch, but because they score more points throughout the entire game (the other teams offense might not be very good at the end of those games as well). Hero Ball gets to the biggest stage (NBA Finals) and ultimately a Hero Ball team wins a championship.

Over time, people have grown to falsely believe that Hero Ball is why teams are winning a championship.*

*Meanwhile, the Non-Hero teams have been slighted: The 2001 and 2002 Blazers were excellent without a traditional “closer,” which probably means the 2000 Blazers were as well. That team nearly won a title.  The 2004  Pistons did win a title.  The 2005 Pistons and 2010 Celtics came within a whisker (or Robert Horry and Ron Artest 3-pointers) away from a championship.

Just How Bad is Hero Ball…and Why do Teams Keep Using It?

It’s not nearly as bad as some make it out to be. In 2010, about 14% of the points in the league were generated by fast-break opportunities, according to If we assume a ridiculously high efficiency on those fastbreak possessions of say, 1.7 points per possession (per the curious results on this page) that means the average halfcourt offensive rating was 101.1 (with technical fouls removed a well). The league-wide average for offense in the final three minutes of five-point games in 2010 was 90.4 points per 100.

That’s a sizeable drop, but it’s also hardly noticeable. Such a decrease will amount to an extra 3-pointer every four games in those clutch situations. This isn’t going to cost a lot of teams much in the standings, and it goes largely unnoticed without our trusty scoreboard.

Furthermore, the difference in 2010 between “Hero Ball” teams and everyone else was almost negligible, although much of that depends on how we define Hero Ball teams.

If we look at the difference between teams led by everyone else and Chris Paul, Carmelo Anthony, LeBron James, Steve Nash, Dirk Nowitzki, Dwyane Wade, Deron Williams and Kevin Durant, the Heroes saw their offense decrease 4.7 points less on average than everyone else, although all of that difference is generated by three players who arguably played the same way for the entire game, James, Nash and Paul. The other six teams performed at league average...

Get the rest of this article over at Back Picks.


Popular Video