Shaquille O’Neal is the most dominant big man in NBA history. He is widely recognized as one of the absolute greats, one of the few truly special players to ever grace a basketball court. Where his reputation falters quite a bit, however—and for good reason—is at the free throw line.
The 7’1’’, 325-pound behemoth, affectionately dubbed ‘Shaq’, is routinely, and often comically, chided for his historically poor free throw shooting. It’s become such a popular diss amongst naysayers, former on-court rivals, and colleagues alike, in fact, that the big man has fashioned a go-to rebuttal in recent years. Nowadays, upon being called out for his failures at the charity stripe, most frequently on the post-game talk show “Inside the NBA” he co-hosts with former on-court rival-turned-funnyman Charles Barkley, Shaq will quickly shoot back: “Sure…but I made ‘em when they count.” Of course, this is an easy, off-hand assertion for the big man to make, and being that fact check is not a particularly prominent feature on the show, Shaq’s claim largely goes unchecked. In the interest of statistical reasoning (and sheer curiosity), I decided to go back and look at whether Shaq, in fact, “made ‘em when they count”—in the clutch—and whether his performance in the clutch varied enough from his success rate in non-clutch situations to suggest that his ability was, in fact, somehow enhanced (or whether it faltered) in the clutch. To do so, I began by determining the two sub-categories O’Neal’s free throws would fall into. As such, I decided that “clutch” shots would be represented by free-throw attempts that came in the fourth quarters of one-possession games, while non-clutch shots would simply cover all other free-throw attempts. I continued along this course by collecting data for all of Shaquille O’Neal’s free-throw attempts since the 2000-01 NBA campaign (disclaimer: O’Neal’s career actually dates back to 1992, but given the fact that the play-by-play data (via Basketball-Reference.com) only dates back to the year 2000, that, by default, became the baseline for how far back this investigation went). Upon doing so, I set O’Neal’s shooting percentage on non-clutch free-throw attempts (of which he had 6,449) since the ’00-01 season as his overall ability, and measured his free-throw percentage in the clutch since the same ’00-01 season against this pre-determined ability. By setting the project up as such, I hoped to answer the question of whether Shaq’s free-throw shooting performance in clutch situations did, in fact, deviate from his ability in otherwise “regular” (non-clutch) situations.
From the year 2000 through the end of his career in 2011, O’Neal took 6,449 free throws in non-clutch situations—of these, he made 3,356 (52.039%). By comparison, the big man took 607 free throws in clutch situations over this same time frame, making 306 —a success rate of 50.412%. The following two-way table and its accompanying graph highlight Shaq’s make vs. miss rate in the clutch, as opposed to non-clutch situations:
As evidenced by the above data, O’Neal did not make a higher proportion of his free throws in clutch situations. However, despite O’Neal actually making a lower percentage of his shots in clutch situations than in non-clutch situations, the difference in success rate was a mere 1.637%. As such, the data seems to suggest that his ability in the clutch was neither diminished nor heightened; rather, his lesser percentage seemed to remain in a reasonable ballpark for a free-throw shooter of his capacity. However, in order to truly investigate whether his success rate legitimately deviated from his general ability, as I deemed it, I ran a dot-plot simulation. The simulation was set up according to the belief that his success rate (52.039%) on a whopping 6449 non-clutch free throw attempts more or less reflected his overall ability. As such, I ran 100 separate simulations of his 607 clutch free-throw attempts, based on the idea that his ability to make them sits at 52.039%, and received 100 different possibilities for how many of these 607 “clutch” attempts a free-throw shooter of his calibre would plausibly make. The two dot-plots below display the results; the first photo shows the number of makes (out of 607 attempts) each of the one hundred simulated performances yielded, while the latter shows the proportion of makes:
According to the results of the simulation, Shaq’s clutch free throw makes, of 607 attempts, fell in a range of 290 to 343 baskets—47.8 through 56.5%. Given that O’Neal’s number of makes, in reality, was 306 (or about 50.4%), firmly within the set of plausible results, it is reasonable to say that O’Neal’s success rate on clutch free-throw attempts was indicative of neither a heightened nor lessened ability in the clutch. Rather, O’Neal’s 50.412% free-throw shooting in the clutch was a reasonable outcome for a player of his general ability, and his ability to “make ‘em when they count” was no greater than it otherwise would be.
A note on any potential errors encountered during this experiment: given the fact that we used what was, in truth, an expansive performance value (Shaq’s free-throw percentage, even for his entire career, represents his *performance* at the line, after all) it is certainly possible that O’Neal’s true “ability” at the free-throw line does not lie exactly at 52.039%. However, given the large enough size of the sample (the data spanned a whole eleven years), it is extremely unlikely that O’Neal’s true ability, however slightly deviant from the one used in this experiment, was variable enough to suggest a statistically significant difference in his clutch ability. Furthermore, while his “clutch” attempts were different than his non-clutch attempts in the sense that they came at a particularly defined point in a game, they occurred during the same game nonetheless. As such, the performance and ability values were both still bound by the same general set of circumstances (being that they came over the course of the same games, seasons, and were shot by the same exact player) that dictated all of the attempts, so any potential concern that the sample size to represent Shaq’s ability does not do so accurately enough should really be a non-starter altogether. All of his attempts were conducted according to the same general set of conditions, after all.
All in all, given the proximity in success rate between clutch and non-clutch free throw attempts, and the very large and expansive sample size used, it is unlikely that any error or limitation particular to the information used was at all significant enough to affect the conclusion that Shaquille O’Neal’s success rate on clutch free-throw attempts was reasonable for a player of his ability, and not suggestive of a deviation from his general ability to shoot them.