Believe it or not, game-winning and game-tying shots aren't the only things that matter when talking about the NBA players who fare best in clutch situations.
Getting to that point matters as well.
For the purposes of this article, "clutch situations" will be defined as the last two minutes of games separated by five points or fewer. That's enough time left for teams to put games out of reach or draw closer during a tight contest, but it's also close enough to triple zeros that the pressure is truly on.
Unfortunately, our minds often lie to us when we're figuring out which players perform at the highest level in the clutch. We're swayed by volume rather than effectiveness, which tends to create warped and inaccurate perceptions.
You'll see a pair of great examples once you make it all the way to the final slide of this article.
To counteract the misleading attempts of our brains, it's time to turn to the numbers. That's why I'm using what I call "clutch rating," a new metric that combines efficiency with involvement.
To calculate it, three steps are required:
- Make sure the player in question qualifies for the rankings. To avoid the effects of small sample size, I'm setting cutoffs as follows: top 100 in clutch minutes played (which is 33 minutes at this point in the season) and at least 15 field-goal attempts.
- Calculate points produced per shot, which you can do by multiplying assists by two, adding points, subtracting turnovers and then dividing that result by field-goal attempts. Unfortunately, there is an inherent flaw—assists leading to three-pointers and passes leading to free-throw attempts don't get extra credit.
- Multiply points produced per shot by PIE, which is an estimate of a player's involvement in the relevant game events. You can read a full description here.
Remember, these rankings are objectively determined. Players aren't receiving bonuses for game-winning shots or anything of that sort, and only the numbers matter.
Note: All stats, unless otherwise indicated, come from NBA.com's statistical databases.