Double-Dipping in Dunning—Kruger

Origi­nally posted at sandy­maguire.me.

I have to get you to drop mod­esty and say to your­self, “Yes, I would like to do first-class work.” Our so­ciety frowns on peo­ple who set out to do re­ally good work. You’re not sup­posed to; luck is sup­posed to de­scend on you and you do great things by chance. Well, that’s a kind of dumb thing to say. I say, why shouldn’t you set out to do some­thing sig­nifi­cant. You don’t have to tell other peo­ple, but shouldn’t you say to your­self, “Yes, I would like to do some­thing sig­nifi­cant.”

Richard Hamming

I want to talk about im­pos­tor syn­drome to­day. If you’re un­fa­mil­iar with the term, it’s this phe­nomenon where peo­ple dis­count their achieve­ments and feel like frauds—like at any mo­ment, oth­ers will re­al­ize that they don’t be­long. Peo­ple with im­pos­tor syn­drome are con­vinced that they’ve some­how pul­led the wool over oth­ers’ eyes, and have conned their way into their jobs or so­cial cir­cles. And it’s very, very com­mon.

On the other end of the spec­trum, there’s an­other phe­nomenon known as Dun­ning—Kruger. This one is widely re­ported in the me­dia as “the av­er­age per­son be­lieves they are bet­ter than 70% of drivers.” They can’t all be right—by defi­ni­tion, 50% of peo­ple must be worse than the me­dian driver. This too is very com­mon; it seems to hold true in most pos­i­tive-feel­ing, ev­ery-day do­mains.

While this is in­deed sup­ported by Dun­ning—Kruger’s data, the pa­per it­self fo­cuses on the fact that “the worst perform­ers in a task are of­ten so bad that they have no idea.” For ex­am­ple, the least funny peo­ple are so trag­i­cally un­funny that they wouldn’t know hu­mor if it bit them. As a re­sult, un­funny peo­ple are en­tirely blind to their lack of hu­mor:

Par­ti­ci­pants scor­ing in the bot­tom quar­tile on our hu­mor test not only over­es­ti­mated their per­centile rank­ing, but they over­es­ti­mated it by 46 per­centile points.

A less well-known find­ing of Dun­ning—Kruger is that the best perform­ers will sys­tem­at­i­cally un­der­es­ti­mate how good they are, by about 15 per­centile points. The pro­posed rea­son is that they found the task to be easy, as­sume oth­ers must have also found it easy, and go from there. In other words, top perform­ers are so good that they don’t no­tice many challenges.

It’s un­for­tu­nate that Dun­ning—Kruger has been pop­u­larized as “most peo­ple think they are bet­ter than they are.” Not only do high-perform­ers already un­der­es­ti­mate them­selves, but those who know about Dun­ning—Kruger in its pop­u­larized form are likely to dou­ble-dip, and fur­ther ad­just down to com­pen­sate for this. For ex­am­ple, if you are in fact in the top 90% for some skill, due to Dun­ning—Kruger you will likely es­ti­mate your­self to be at 75%. But, if you know that peo­ple rou­tinely over­es­ti­mate their abil­ities by 20%, you might drop your es­ti­mate down to 55% in com­pen­sa­tion—sig­nifi­cantly lower than your true skill level.

If this is true, it would sug­gest some of the world’s best peo­ple will es­ti­mate them­selves to be worse than the self-eval­u­a­tions of the wor­lds’ worst. The take­away is this: if you’re the kind of per­son who wor­ries about statis­tics and Dun­ning—Kruger in the first place, you’re already way above av­er­age and clearly have the nec­es­sary meta-cog­ni­tion to not fall vic­tim to such things. From now on, un­less you have ev­i­dence that you’re par­tic­u­larly bad at some­thing, I want you to as­sume that you’re 15 per­centile points higher than you would oth­er­wise es­ti­mate.

The math­e­mat­i­cian Richard Ham­ming is said to have of­ten ruffled feathers by ask­ing his col­leagues “what’s the most im­por­tant prob­lem in your field, and why aren’t you work­ing on it?” Most peo­ple, I’d sus­pect, would say that the most im­por­tant prob­lems are too hard to be tack­led by the likes of them. That it would take minds greater minds than theirs to make progress on such things. They give up be­fore hav­ing even tried.

In­stead, the smartest peo­ple I know join Google or go work at B2B star­tups and are si­mul­ta­neously bored in their day jobs, feel like they’re frauds while they’re there, and don’t have enough en­ergy to work on per­sonal pro­jects af­ter hours. But at least they’re mak­ing wicked-big pay­checks for them­selves. And for their less-qual­ified lead­er­ship.

The best minds of my gen­er­a­tion are think­ing about how to make peo­ple click ads.

Jeff Hammerbacher

Here’s an in­ter­est­ing thought ex­per­i­ment. If you ran­domly swapped places with some­one for a week—you do­ing their job and they do­ing yours—how would it turn out? If you’re a com­pe­tent pro­gram­mer with pass­able so­cial skills, I sus­pect you would be a lot more suc­cess­ful in that week than your re­place­ment. Most things just aren’t that hard, and a good per­centage of the peo­ple do­ing those jobs are phon­ing it in any­ways.

The bar on com­pe­tency is trag­i­cally low. And yet the world re­volves.

If you agree with this line of rea­son­ing, it means the world is just ooz­ing with po­ten­tial, ready for the tak­ing. Most of the world’s most com­pe­tent peo­ple are un­aware of just how good they are. Most things re­ally aren’t as hard as get­ting through that graph-the­ory class you took, and don’t take nearly as much effort. The world is be­ing run by peo­ple who are too in­com­pe­tent to know it; peo­ple who are only in power be­cause they’re the ones who showed up, and be­cause show­ing up is most of the bat­tle.

Which leads us to the in­escapable con­clu­sion that this world we live in is par­tic­u­larly amenable to change. That if you’re will­ing to trust in your own in­stinct and tackle hard-seem­ing prob­lems, you’re go­ing to ex­pe­rience liter­ally un­be­liev­able amounts of suc­cess. Every­one else is defer­ring to bet­ter minds. So be the change you want to see in the world, be­cause we’ve been wait­ing for you.