Outcome Bias

BiasedWhile I was concluding a difficult cross-country trip last evening, my Chargers were wrapping up a surprising win against the Indianapolis Colts. To clinch the win, Nick Novak hit a 50-yard field goal, his fourth of the night, with 2 minutes left to give the Chargers a 19-9 lead. ESPN’s highlights are shown below.

There should be at least three clear takeaways from these highlights and from the game itself:

  1. The Bolts’ “powders” are the best uniforms in sports;
  2. Rookie receiver Keenan Allen was a steal in the 3rd round of the draft (full disclosure: he was a teammate of my younger son at Cal); and
  3. Kicking that last field goal was a mistake.

My focus here is #3.

We are all prone to outcome bias, whereby a decision is evaluated with far too much focus on the ultimate outcome instead of on the quality of the decision at the time it was made, given what was known at the time. That Novak nailed the kick and the Chargers won masks the reality that Coach Mike McCoy never should have ordered it. Even so (and not surprisingly), the pundit class loved the decision. Jon Gruden of ESPN labeled it “a gutsy call.” Local sports radio is abuzz with those who approve of the choice because of how it worked out. However, McCoy should have run the ball instead. Teams attempt field goals rather than “going for it” far too often. Bill Barnwell’s wonderful, weekly “Thank You For Not Coaching” series at Grantland is essentially an ongoing tutorial on when and why to go-for-it on fourth down. The Chargers should have gone for it.

The situation last night was pretty straightforward. The clock showed 2:00 left in the 4th quarter and the Chargers faced fourth and very short at the Colts’ 33 after a pass play to Antonio Gates was stopped just shy of the necessary first down yardage. As Barnwell has pointed out, punting wasn’t an option. At most, the Chargers could expect to gain only about 20-25 yards of field position by punting while giving the Colts a chance and plenty of time to win outright in regulation.

On the other hand, success at either kicking the field goal or making the first down would effectively end the contest.  So which was the right call? Field position wasn’t a big deal. The Colts would have gotten good field position with a 4th down stop (most likely at their own 33), but they would have gotten slightly better field position (their own 40) with the extra yards provided by a missed field goal attempt. Either way, since success at whatever approach is selected wins the game, the only question that really matters (also provided by Barnwell) is whether it is easier to convert on fourth-and-less than-a-foot than it is to kick a 50-yard field goal. Whatever choice offered the best chance of success was the right choice. And that’s a question to which the available data can speak pretty dispositively.

So far this year, teams made good on 55.2 percent of fourth and one situations overall.  But over the previous five seasons and a much larger sample size including nearly a thousand attempts on fourth and one alone, the average success rate on third or fourth and one is 68.8 percent (64.2 percent on fourth and one). Since the Bolts had run the ball extremely well the entire game and since less than a foot was needed in this instance, the actual likelihood of success was a fair amount better than that — probably something like 75 percent.

On the other hand, NFL kickers have converted field goal attempts of between 50 and 52 yards over that same time frame 64.7 percent of the time. Novak had been 6-for-12 on kicks of 50 yards or more for his career (his stats have been nothing special in the aggregate), but he was playing in calm conditions at home and kicking at the shortest “50 and over” distance. Moreover, his performance has been much better since coming to San Diego (13 of 15 overall this season, with his only two misses coming on blocks; and 18 of 20 overall last year, with both misses coming from outside 50 yards).  However, such hot streaks likely tend to revert toward the mean (even though the subject is hotly debated). But it’s understandable that McCoy might have real confidence in Novak, even if it is likely based upon recency bias. It’s possible if unlikely that Novak has improved markedly in San Diego, but it was still a mistake to kick in that situation.  Even if we assume the higher 65 percent confidence rate consistent with the overall success rate of NFL kickers from similar distances, that success rate is decidedly lower than the success rate (well above 68.8 percent and probably about 75 percent) for running the ball there. If the goal is having the best chance of success, kicking was a mistake.

The move worked out for McCoy, of course, but the process was flawed. “It’s an opportunity to go up 10 or, what are you going to do? You go for it, you don’t get it, you give them a situation to tie the game,” McCoy said. “I have all the confidence in the world, the way he’s been kicking the ball.” That approach is understandable and even a reasonably close call, but it was still a mistake.

When making decisions amidst uncertainty, as in football or investing, the process is what matters — not the outcome, despite our bias in that direction. The available data should control the process, which is much more important than the outcome, driven in large measure by chance. Sadly, like NFL head coaches, we don’t usually work that way.

5 thoughts on “Outcome Bias

  1. Great post Bob.

    As stated above, the 4th & 1 conversion percentage is 64.2% while the field goal conversion percentage is 64.7%. I would argue that using broad averages is dangerous to begin with, but if we do agree to use averages, it seems the statistical play (using the applicable averages) is to kick the FG. Any deviation from those averages seems like guesswork.

    Or, perhaps Mathews inability to stay in bounds earlier in the game tilted McCoy’s confidence away from the run game!

    • It seems to me that the right approach is to start with the base rates and move from them only with very good reason. I doubt that Novak’s personal numbers are reason enough to move from the base rate with great confidence, but 4th and a foot for an offense that had run well all night ought to be factored in. That’s why I think running was the right call. But it was a close case. Thanks for reading and commenting.

  2. Pingback: Worth Reading | Above the Market

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s