One of the most memorable moments of the 2018 NBA Playoffs was Damian Lillard’s buzzer beater to send the Thunder home in the first round.
It was certainly an impressive shot, but was it a smart one?
After the game, Paul George said that it was “a bad, bad shot. I don’t care what anybody says. That’s a bad shot. But hey, he made it. That story won’t be told that it was a bad shot. We live with that.” George was raked over the coals for this comment, even being scrutinized by media companies from Oklahoma City. The Oklahoman wrote that, “No Lillard shot over the span of the last 11 days was bad.”
But what would the media have said if Lillard missed the shot? They would have blamed him. He made no attempt to get anywhere near the basket and took an off balance shot from 37 feet away over a much taller defender. That’s a bad shot.
What we are observing with this discrepancy is a limitation of our way of thinking. We are inept at considering alternative outcomes from what actually happened. This gives us a hindsight bias, where we believe that the result of a decision was the only possible outcome of that decision. We assume that Lillard stepping back from the logo was a smart decision because the shot went in, but we fail to recognize the overwhelming evidence that it was in fact a poor shot choice. In short, we “assess the quality of a decision not by whether the process was sound but by whether its outcome was good or bad.” – Thinking, Fast and Slow
Let’s take a look back at the 2018 NFL season.
Down 14 late in the fourth quarter to the Falcons, the Giants scored a touchdown to cut the lead to 8. Rather than kick the extra point, as most coaches would, Pat Shurmur elected to go for two. The Giants failed the attempt and ended up losing the game.
Naturally, Shurmur was questioned about this decision after the game because the outcome of his decision was negative. But was going for two the right call? Let’s take a look at the math for a team down 8 points after scoring a touchdown late in the game.
*Note: We assume that the trailing team goes on to score another touchdown (because they have to in order to have a shot at winning anyway. We also assume a 50% chance of converting the two point conversion, and a 100% chance of making the extra point, whose real, league-wide probabilities are 47% and 97% respectively).
As you can see, going for two gives you a 62.5% chance of winning (50% for a regulation win and 12.5% for an overtime win). But kicking the extra point each time gives us just a 50% chance of winning (guaranteed to go to overtime, and then a 50% chance to win in overtime). Therefore, it is objectively a correct decision to go for two if you score a touchdown near the end of a game to go down eight.
Yet when Pat Shurmur made this decision during the 2018 season, he was ridiculed in the postgame press conference. Why? Because the Giants didn’t convert. Now let’s say that the Giants kicked two extra points and lost in overtime. Would Shurmur have been criticized? Of course not! Because this is what most coaches would have done.
Decision makers who will be criticized for negative outcomes, even if it’s the result of a correct decision, are less likely to go against the consensus because they know they will be criticized if the outcome is negative, but they will always be let off the hook if they follow the crowd.
In summary, hindsight bias makes decision makers far less likely to make unpopular decisions, even if the logic behind them is sound. This is partially why we see coaches continue to make decisions the “old-fashioned” way: it increases their job security, even if the decisions are wrong.
It takes a lot of bravery to put yourself in the crosshairs by making a correct decision. Coaches that do so are the unheralded heroes of the NFL; they risk their jobs and their reputations by doing what is right.