Anchors, and Anchoring

Question 1: what’s your estimate of the rest-of-season wRC+ numbers for Marcell Ozuna, Matt Carpenter, and Dexter Fowler? Don’t look up their projections. Don’t spend more than ten seconds thinking about it.

Now, choose A or B, and answer the corresponding version of Question 2. Again, don’t take more than ten seconds, and don’t use a calculator.

Question 2A: what is 1 x 2 x 3 x 4 x 5 x 6 x 7 x 8?

Question 2B: what is 8 x 7 x 6 x 5 x 4 x 3 x 2 x 1?

This isn’t a real experiment, obviously. This is the set-up for I don’t know, 1,200 or so hopefully-thoughtful words about the St. Louis Cardinals. But if this were a properly controlled experiment, I can confidently make a couple statements:

1. We would, collectively, be lower than we should be on Question 1. (And on both versions of Question 2, but that’s not really important. FYI, it’s 40,320, which the average subject misses by an order of magnitude.)
2. The people who answered 2A would have a lower guess than the people who answered 2B. And those results on Question 2 explain the results on Question 1.

Question 2 isn’t something I made up. It’s an experiment that the cognitive psychologists Daniel Kahneman and Amos Tversky did over 40 years ago, and which helped form the basis for one of their early articles on common cognitive biases. This particular bias is called “anchoring,” and the multiplication question is a simple and famous example of what anchoring is.

Nobody but a mathematical savant (or a person with the factorial of 8 memorized) can do the math in Question 2 in just a few seconds. Our brains are great at shortcuts, though. It’s part of what makes us so smart. But it’s also part of what makes us, in some cases, so dumb. In the case of Question 2, the natural human shortcut is to start doing the math, then to guess when time runs out based on what you’ve done so far. The result is both intuitive and, in a certain sense, maddening: the people doing Question 2A (starting with small numbers) will almost always guess a lower number than the people doing Question 2B (starting with big numbers). The former group has a smaller number in their heads when the time for reasoning ends and the time for guessing begins, and vice versa for the latter group, and the strength of those cognitive anchors is too much for our brains to overcome.

Put simply: the first stuff we see matters a whole lot to our shortcut-happy and suggestible brains, and we are bad at adjusting away from it. The first offer in a sales negotiation skews the results. Asking a person “is the population of France more or less than 150 million” will result in much higher estimates than if you anchor it at 20 million (real answer: 67 million). And so on.

Once this phenomenon was identified, and studied further, a few additional things became clear. For one, it doesn’t matter much how well-educated the subjects are. Even more annoying, it doesn’t even help all that much to be informed of the phenomenon; experimental subjects told about the nature of the experiment still aren’t able to avoid the effects of anchoring. This is just a thing our brains do to make sense of the world. Placing outsize importance on initial data kept our distant ancestors from eating poisonous fruit, and today it makes us pay too much for used cars. Not the worst trade-off, really.

Right, okay: the St. Louis Cardinals. This is a baseball blog.

Coming into Friday night, Matt Carpenter’s wRC+ for 2018 was 72. Dexter Fowler’s was 60. Marcell Ozuna’s was 66. Those are terrible numbers. Those are numbers which, if the very best defensive shortstop in the world put them up, he’d still only be borderline MLB player. That’s three guys who are in the lineup nearly every night, and each one of them has, so far, been fairly describable as an anchor around the Cardinals offense’s neck.

But each of those guys is a good hitter. Do you think one or two or all of them should be benched? You are not alone (not by a long shot), but you are also wrong. Your monkey brain is misleading you. So is mine. That is what our monkey brains do.

How many times have you read (or thought) this of late: “yes, [slumping hitter] will probably figure it out, but right now he’s terrible and they need to use somebody else”? There’s a ton of that kind of talk in May. That’s because early-season slumps get in our heads more than late-season ones do. Right now (for example) calls to bench Dexter Fowler are rampant. Fowler actually isn’t slumping in a way that’s outside of his career norms, but he’s done it in April and May instead of August. So we’ve all anchored on his terrible early-season results and it’s very hard to recognize that he’s been a good hitter for a long time and yes, he’s going to hit this year. Same with Ozuna. Same with Carpenter.

I say this not to toot my own horn, but to illustrate a point: a little over a month ago, Kolten Wong had an absolutely awful 2018 line, and I wrote this, stating that there was likely nothing to worry about. Since then, Wong has a 97 wRC+ — right in line with his career production. It didn’t feel like he was himself (and because Wong’s 2018 wRC+ is just 72, it still kinda doesn’t), but after the first couple weeks of the season Wong really has been himself. Our early assessments were, as they tend to be, just wrong.

Similarly, Matt Carpenter, Dexter Fowler, and Marcell Ozuna each project as above-average hitters, but they don’t feel like they are right now. That’s a normal feeling. Even being an educated fan, there’s no way to avoid it. That feeling is also very likely incorrect. It’s not incorrect a month from now, after they’ve proven themselves again — it’s incorrect right now.

I can’t promise you any of these guys’ lines will turn around starting today, the way Wong’s did when I wrote about him. Pointing out the likelihood of regression toward career norms isn’t a magical incantation that makes it start immediately. All I can do is offer you the comforting (or not) fact that your internal impression of the future production of the slumping heart-of-the-order Cardinals hitters is almost definitely too low. Mine is, too. It’s because they were bad early, and we’re dumb this way.

But the thing is, these guys are good hitters, even if it’s hard to believe right now. They’ll hit. And you’ll forget how wrong you were.