HR and the Unconscious Mind: Do we automatically judge a book by its cover (in poker and promotion)?

I watched an episode of The World Series of Poker recently.  Anyone who’s watched the show knows that the viewers and announcers get a secret glance at the cards each player is holding (and anyone who’s ever played poker knows this makes the game a whole lot easier).  But this time, for one hand a player’s cards remained hidden.  After some serious betting, with several players staying in, this fellow acted just a little nervous and then made a big (but not too big) raise.  Several players stayed in the pot with him, though he continued to bet strongly as the shared cards were laid on the table.

Having played quite a bit of this game myself, everything I saw led me to one conclusion: “This man’s got a huge hand.”  However, as the hand continued one announcer said something like “You know I don’t think he’s got it.”  The other replied “Yeah…I don’t know what it is about this guy, but I just don’t trust him.”  Apparently the players at the table felt the same, because they called all his bets.  Naturally, he ended up turning over the best possible hand (two aces) and raked in a monster pot.

What happened??  All rational signs—the stakes, his bets, his mannerisms, his responses to other players—said he had a great hand.  Yet nobody believed him.  The explanation, I think, is that he looked just a little bit like this:

A very untrustworthy face.

A very untrustworthy face.

Whether or not we admit it, this looks like somebody we probably wouldn’t trust, whether sitting at a poker table or interviewing a potential employee.  If asked whether we judge a person by his face, most of us will say “absolutely not.”  But before being too sure, watch this transformation:

If you’re like most people, you probably saw the face get more and more trustworthy for the first 5 seconds, and then morph in the opposite direction, so that by the 12th second you’d cross a street to avoid him.

Research from Alex Todorov’s lab at Princeton shows that people easily and consistently rate faces along this “trustworthiness” dimension, and make similar inferences in other domains like dominance and competence.  (To watch computer generated faces morph along various dimensions, check out their website.)  In fact, these judgments are so natural and effortless that even children—so young they can barely speak—make them nearly as consistently as adults (Cogsdill, Todorov, Spelke & Banaji, in press).

You’re probably wondering whether such judgments are accurate.  There are some counterarguments, but most of research seems to say “no, not really.”  In spite of this, they influence our judgments in even very serious domains. In one frightening example, Todorov and his colleagues have shown that perceptions of facial competence play a substantial role in how we choose our nation’s leaders. During the last few congressional elections, researchers in this lab had people independently rate the competence of each candidate’s face.  Importantly, the raters had no knowledge of the candidates or their actual competence: They simply rated how competent each person looked.  These ratings predicted nearly 70% of election results!

What does this sort of thing mean for HR professionals?  Imagine you’re deciding who to hire or promote.  In this day and age, most of us know not to judge a person by things like race or gender.  And that awareness helps us make decisions both more fairly and more pragmatically.

But are we as aware of our own minds when processing the trustworthiness or competence of a face?

When we have a gut feeling that someone cannot be trusted, we can be right…or we can be wrong.  We may be unconsciously processing some good information.  But our (occasionally pesky) automatic mind can just as easily pick up on irrelevant information.  This leads us to choose senators carelessly and make bad poker calls.  Might we just as easily pass over a great person in our organization, because we notice the bone structure of her face rather than the content of her character?


  1. Avatar ZenMadman   •  

    This suggests that it’s more efficient/accurate to interview people by phone rather than in person. It also suggests that playing poker online instead of in person could lead to more accurate bluff catching. It also explains why people rarely call my bets and frequently ask me for directions. 🙂 <— (because that's what I look like.)

    • Steve Lehr Steve Lehr   •     Author

      It’s true…though of course we also have biases based on things like accent (and there’s also legitimate information one might pick up in person). The truth is meta-analyses suggest that interviews in general work much more poorly than we’d hope…

    • Steve Lehr Steve Lehr   •     Author

      Btw, as I was writing this I was thinking it would be a lot of fun to set up an informal experiment on a poker site where I’d make a series of bets that are as close to identical as possible, using different pictures (like the one for this blog) as avatars. I’ve often thought people don’t pay enough attention to the influence of their avatars on other people’s online play.

Leave a Reply

Your email address will not be published. Required fields are marked *