A popular gripe among Rockets fans, this year in particular, is the presence of "hero ball" in late game situations or when the offense is sputtering. Specifically this criticism is laid at the feet of James Harden, the franchise shooting guard Houston acquired last year from Oklahoma City in a highway robbery of a trade. Definitions of hero ball vary from person to person that uses it, but Henry Abbott first publicized the term on the True Hoop network (Clicky). Abbott defined hero ball as isolation plays near the end of games resulting in low percentage shots. The ball movement stymies and often the shot is contested. James Harden has drawn the ire of Rockets fans under the guise of hero ball but it's worth taking a critical look at what Abbott decried about the tactic and an element that he ignored, the psychology of the sport.
Harden on the Numbers.
Efficiency is the primary factor in assessing whether or not a shot was a "hero ball" shot or not. Generally contested shots in isolation are not efficient. If you choose to go by the Sun Tzu quotes littered throughout the Hero Ball article, a well executed and open shot is the best one available. The numbers on SynergySports.com bear that out more often than not. Spot up shots and transition plays tend to bear more points per possession than all others. Since we're attempting to look at late game situations we're more often than not going to have to look at isolation and pick and roll, where one player can dominate the ball and "play hero."
Now, in order to assess whether or not a style of play is hero ball one must assess whether or not a player's efficiency is drastically reduced by the chosen play. In order to fleece that out we need to look at how James converts on his plays. 50.5% (24.3% in isolation, 26.2% as the P&R ball handler) of his plays are isolation or pick and roll plays (As the ball handler). These two sets returns a 33.2% turnover rate. Harden records 69 touches per game (According to NBA Player Tracking) and 16.4 field goal attempts per game. James manages to record .92 points per possession in isolation and .87 as the pick and roll ball handler in raw numbers. Let's see how these stats bear out:
The Rockets currently average 105.6 points per game on 96 possessions a game, good for 1.1 points per possession. Keep this in mind as a baseline.
16.4 Field goal attempts per game with 24.3% in isolation yields 4 isolation attempts a game and 26.2% of plays as the pick and roll ball handler yields 4 plays there as well.
James turns the ball over at a 14% rate in isolation, good for .564 turnovers in those 4 isolation attempts. Harden converts at .92 points per possession in isolation, on 3.4 possessions, yielding 3.12 points in 4 isolation possessions. Harden is rewarded with 1.5 free throws per attempt (Average of 6 free throws on iso's counting and-1 possessions, converts at 85% which leads to an additional 1.2 points per play) which inflates his conversion rate to 4.3 points per isolation set. For comparison's sake, the Rockets 1.1 points per possession would yield 4.4 points.
Out of 4 plays in pick and roll ball handler situations .7 end in turnovers (19.1%). If 3.3 plays end in points at his .87 points per possession rate Harden produces 2.9 points per game off of pick and roll ball-handling plays. Harden returns .786 free throws attempts per pick and roll play (As the ball handler), converted at 85% free throw percentage yields an additional .7 points per and yields 3.6 points per pick and roll ball handler.
What we can develop from these stats is a ratio and a likelihood of success or failure. Based on the figures above Harden converts 1.075 points per isolation play (Adjusted for turnovers and free throws). When you adjust Harden's pick and roll efficiency for free throws and turnovers he produces .9 points per possession. Harden's true shooting percentage currently sits at .598, which means whatever he's doing it tends to work. Primarily, Harden's ability to draw free throws in isolation and on the pick and roll helps to keep his production nearly on par with the Rockets total efficiency production.
The Rockets average 2 possessions per minute (96 possessions per game, 48 minutes in a game). Houston has a 14% chance of turning the ball over by going to a Harden isolation play. They have a 19% chance of Harden turning the ball over in a pick and roll situation. As a team, Houston has a 17% chance of turning the ball over (Average turnover rate). Harden's average turnover rate is 15.9%, even adjusting the team turnover rate by reducing Harden's total rate it will not bring the team average down to 14%.
Too Long, Didn't Read
What does this say? Well, in simple terms, it states that in isolation, Houston has a better chance of safeguarding the possession and converting if James just does it himself. Further, given Harden's conversion rate of 1.075 points per possession in isolation, he is only .035 points per possession off from the Rockets average points per possession. Harden's isolation turnover rate is lower than that of the Rockets as a team, even adjusting down for his turnover rate. In the pick and roll initiated by Harden, the Rockets will only slightly underperform their team average (.2 points per possession differential) and will see a 2% increase in turnover possibility. Offset this by the fact that Dwight Howard tends to be the pick and roll man for Harden and Howard converts at 1.16 points per possession and it's an even sexier proposition.
Basketball is a game of calculated risk. In a situation where isolation is nearly as efficient as execution throughout a game, it's inherently not a faulty proposition to go with the play that eliminates a wide range of negatives for one play that stands to net you the same result. That's purely from a numbers standpoint, though. Numbers can be argued and debated as alternatives are offered. Abbott did state hero ball won't go anywhere, but, not enough attention was adequately paid as to WHY it won't go anywhere.
The Psychology of the Sport.
In the founding hero ball article (And subsequent later articles) a great deal of stress is put on the idea of moving the ball and garnering open shots. In an ideal world, this is always the case. However, basketball is a game that grinds on a player, wears on his mental and physical state, and causes usually reliable muscle memory to fire a little more erratically when the games get closer. Multiple examples of NBA legends critiquing modern players are given. A myriad of examples of NBA playoff teams moving the ball for a win is given. What is missed, however, is that, of all of those anecdotal examples and situations, there was either a transcendent talent present to allow players to get open, or that transcendent talent placed his mark on the game and converted the shots.
Reuben Fischer-Baum of deadspin.com offered a simple rebuttal to the points per touch theory advocated by Abbott's anti-Hero Ball charge. Specifically in his points per possession chart and touches per possession chart he has shown that, at least early in the 2013-2014 NBA season, that ball movement does not necessarily correlate to more efficient basketball. The Blazers, Heat, and Rockets produce the highest points per possession totals and all fall at or below the NBA's average mark for touches per possession. Intriguingly enough, but for a pair of outliers in San Antonio and Dallas, the more touches per possession the less efficient the points per possession stat becomes.
This exposes a pretty gaping hole in the hero ball argument. The more the ball is distributed, the less likely it is that execution will remain flawless, fluid, and productive. That's not to say the ideal touch per possession is one or two but to place some emphasis on something the anti-hero ball argument ignores. The more complex a plan gets, the less likely it is to succeed. Intricacy requires mastery and although the NBA is the mastery level of basketball, there are clearly differences in talent level and grace under fire. The reason certain players are paid franchise money is simply because they have reduced the flaws in their games. James Harden is an elite scorer in the NBA in the same conversation as Kevin Durant for his scoring prowess. Dwight Howard is the best center in the NBA and a workhorse you can count on (Up until critical free throws enter the conversation). Much is made to the average fan about killer instinct and the ability to deliver in the clutch. When a player has ice in his veins and can execute to finish off a team, he's a thing of legend. Yes, this all sounds like creating a cult of heroism but basketball is as much a game of heart as it is a game of numbers.
What's it all mean?
The hero ball trope looks at the numbers and assumes all things are equal at all stages of a game. How often has an athlete "choked" under pressure? A stupid turnover here, an errant step out of bounds there, a JR Smith shot assuming it was necessary every so often. All of these issues are psychological and not quantifiable in numbers. These mistakes happen. If you can limit those mistakes when possessions become more critical, why wouldn't you? The pressing issue to take away from this discussion is simple. James Harden converts on isolation plays almost more efficiently than the Rockets offense operates as a whole. The Rockets are lauded for having a wildly efficient offense and James matches that IN ISOLATION. For all the impatience and anger directed at him he's a one-man army on the court when the Rockets call his number. Rather than directing ire at one of the Rockets first franchise players in years, perhaps a closer evaluation of the numbers and the psychology of the sport need take place. Only once that context is set can we actually discuss the hero ball cliché. Is James Harden playing for his ego? Who knows? What is known is that, flying solo, James turns the ball over less than the team and scores about as well as the team. He's also fouled at a high rate in isolation. At the end of the game, what do you think is more likely, Harden's ability to hit a shot with some separation, or four other Rockets flawlessly swinging the ball around the perimeter to free up a shooter? If you've watched enough Rockets games (Especially in the fourth quarter) you know one of these is more likely to result in points than the other.