In the NBA, there has been a lot of controversy about the correct basketball play. In particular, many have criticized Lebron James for his proclivity toward passing the ball in the most important moments of games, while praising Kobe Bryant. As the popular talking point goes, a true superstar should take his team onto his shoulders and will his team to victory when the pressure is greatest.
I’ve always felt this was an argument more steeped in machismo than reality. My goal in this article is to present my definition of the “correct play” and to determine what evidence we have about whether a superstar in general, and Lebron James in particular, should try to take over a game on his own vs. looking to his teammates for assistance.
First off, what is the correct basketball play? My definition of the correct basketball play is the play that leads to the greatest chance of a team winning. Most of the time this will be the play that produces the greatest expected number of points, although in specific end-of-half and end-of-game scenarios, this may not always be true.
Now to be clear, I’m saying that the actual result of the play does not justify the play in and of itself. Rather, it’s the expectation from the play that makes it correct or not. The result of a single play is far too small a sample size to determine whether it was correct or not, but the average result of many such plays, while still imperfect, is a much better gauge of what is right and what is not.
So which plays are the best plays? What is the Correct Basketball Play?
Synergy broke down various types of plays by points per possession produced and found that plays involving off-the-ball cuts (1.18 ppp) and transition plays (1.12 ppp) are the two most efficient plays. The least efficient is an isolation play, which is worth 0.78 points per possession. In iso plays in which the player passes the ball, that number jumps to 0.93 ppp. The least efficient way to score, it would seem, is 1-on-1.
Another way to break down the value of a play is to look at assisted versus unassisted shots. The folks at 82games.com calculated that an assisted shot had an 8% better chance of going in. Given that free throws are more likely on an unassisted shot, however, we find that an assisted shot is only worth about 0.13 more points than an unassisted shot. This is also evidence that trying to score on one’s own is, more often than not, not going to be the right decision.
Of course, there are numerous game-specific factors that alter these numbers for any given team or player, not the least of which is which are the players actually involved. Furthermore, you can only run one play so many times over the course of a game before the defense is going to adjust. Yet on the whole, I doubt many will disagree with the idea that getting the whole team involved is the best way to run an effective offense.
Some would argue, however, that the endgame, be it the final few minutes or the final possession, is a completely different animal, and the right basketball play can become the wrong basketball play. There is no doubt that added pressure has a tendency to bring out the best and the worst in athletes, and there’s also no doubt that those at the peak of their craft are most likely to excel under such circumstances. Yet is this phenomenon really so powerful that it fundamentally changes how basketball should be played? I have my doubts.
We know that as the game progresses, fatigue sets in, fewer shots are assisted, more outside shots fly, and shooting percentages decline. The closer we get to the end of the game, the worse offensive efficiency gets, with it reaching its nadir in the few seconds before clock strikes zero.
However, one team consistently bucked that trend since 2000: Lebron James’ Cleveland Cavaliers. El Gee at backpicks.com has looked in depth at clutch performance, and among his findings were that no team improved their shooting in the clutch from 2006-2010 other than the Cleveland Cavaliers, whose eFG improved by an impressive 2.6%. In 2009, they improved by a shocking 13.5% in the clutch, which is the best of any team since 2000. During Lebron’s last 3 seasons in Cleveland (2008-2010) the Cavs were the only team with a crunch time +/- of over 100, a mark they reached all three seasons. And lest you question their ability to put it together in the postseason, their eFG% was an otherworldly 66.7% during the postseason, and their simple rating system was the best of any team within the past 12 years in 2009. During this time period, James averaged 44 points, 10 rebounds and 7 assists per 36 minutes with a TS% of 63%. Once he left, the team regressed to the mean.
When he went to Miami, Lebron joined a team that had been outscored in the clutch 3 years running and got them to +45 and +27 the past 2 years. He scores more, scores more efficiently, and assists at a higher rate during the clutch. This from the guy Skip Bayless referred to as not even Robin, but Alfred, during crunch time. With all due respect to Bayless, Lebron makes a strong case as the best crunch time player in the league, both in terms of individual and team results.
Lebron’s main competition for that title has been Chris Paul, whose Hornets outscored every other team on a per possession basis in the clutch during his time in New Orleans. According to Paul, “Why shoot with three people on me if one guy is open? If I’m open, I’ll shoot it, and if I’m not, I’ll pass it.”
All of this is evidence that demonstrates that keeping everybody involved is still the right move, even in the last few minutes of the 4th quarter. But what about the final possession, or the final shot? Shouldn’t this be where the wheat is separated from the chaff?
Jordan Sams put together a compilation of clutch shots since 2000 and potential game-winners since 2006. While Lebron is tied with Tim Duncan for highest field goal percentage in the clutch (last 5 minutes, game within 5 points) at .460, he is only 18 for 61 for a .295 field goal percentage in such situations in the final 24 seconds and a chance to win or tie. Among the twelve players with the most potential game winning/tying shots since 2006, the average clutch FG% was .417, but in the final 24 seconds, it was only .314. Even the best players are going to miss over 2/3 of their shots in the final 24 seconds.
At the Sloan Sports Analytics Conference, they looked at all players trailing by 3 points or less with 30 seconds left, and they found that on the road, such players had a FG% of .359, while at home, their FG% was .314. Admittedly, the timeframe is not identical to the 24 seconds used in Sams’ dataset, but it’s pretty close. And what the data suggests is that a potential game-winner or game-tying shot from a superstar is, at a minimum, not more likely to go in than that of one of their teammates. If there’s time to make that extra pass, there sure doesn’t seem to be any downside.
Now frequently, with the game on the line, having your superstar shoot will be the right option. Often the superstar will be able to get a better look than any of his teammates or will be able to draw a foul and get to the line. At other times, the defense won’t give you any better options. Sometimes, there just won’t be enough time to do anything else. But the desperation jumper over multiple defenders with no time left on the clock is the worst play in basketball.
Given the option to take the guy willing to defer to his teammates, or the guy who will do it all himself, all things being equal? I’ll take the former. It was good enough to help Michael Jordan and Larry Bird win titles, and it may one day do so for Lebron as well. But regardless of whether it does or not, looking to your teammates when they’re open is still the correct basketball play.
- Bill Simmons: Larry Bird Would Love to Play with Kobe Bryant (2/8/12)
- The Wilt Chamberlain Debate – Basketball’s Goliath Conundrum
- The Top Ten Reasons Why I’m Glad That Basketball Is Back
- David Friedman: The Strengths and Limitations of Advanced Basketball Statistics (3/10/12)
- Michael Jordan: Game Winning Shots