Rat Beats Human

Rat Beats Human

By Mark Chussil

Once upon a time, there was an experiment.

In this experiment rats could press either Bar X or Bar Y. If the rat pressed the correct bar, it would receive a pellet of food; if it pressed the incorrect bar, nothing. (Why do experimental rats get “pellets” of food? Why not morsels, nibbles, tidbits, noshes, or treats? Poor rats.) The bars were calibrated in a random sequence such that 60% of the time Bar X would dispense the delight and 40% of the time Bar Y would serve the snack.

A similar experiment, also with a 60/40 random sequence, was run with humans. It is unknown whether the humans received pellets of food or some other form of reinforcement.

The rats quickly figured out that the best strategy was to press Bar X every time. Mathematical­ly, that’s the best possible strategy. They got a 60% success rate, the best possible outcome.

The humans tried to outsmart the system, to find patterns in the random sequence, to predict the next move. They switched cleverly between Bar X and Bar Y, and they succeeded less than 60% of the time. Rat beats human.

No one — well, no human — says a rat is smarter than a human; yet, competitively, rat beats human. Why? How?

I believe I am not a rat so it is difficult for me to understand their decision-making process. I am also not all other humans. So, based on conjecture:

Pride and ego. Rat: None discernible. Human: Price and ego never rest.

Practicality. Rat: Accepts simple solution. Human: Tries to figure it out even if there’s nothing to figure.

Driven by evidence. Rat: Dispassionately gathers and analyzes intelligence. Human: Seeks patterns, rejects the very idea of randomness.

Interpretation of luck. Rat: I was lucky. Human: I was smart.

Now, there’s a quirk here. We humans are willing to compare ourselves to other humans. We are not willing to compare ourselves to rats. (Or even to computers, but that’s another story.) I compare my pellet-record to yours and you compare yours to mine. You got 57.3% and I got 57.1%; I want to know what you did that I didn’t do. (Note that I may think you got lucky, but you won’t.) I’ll do my best to adopt your best practices. You’ll go into consulting and sell them to me — a proprietary framework with the buy-me-now title “pellet-acquisition optimization in today’s unpredictable global economy” — along with an annual subscription to The Journal of Pellet Optimization and Prediction.

Humans compete not only over pellets but also over planets. A long time ago astronomers strove mightily to devise an accurate model of the geocentric heavens; that is, to write equations that would correctly predict the movement of the planets in a cosmos with the earth at its center. Their models grew more and more complex, and even so failed to fit the data. Then along came a smart human named Nicolaus Copernicus who published a heliocentric model with planets, including the earth, circling the sun. Copernicus was no rat but he possessed the enviable rat-skill of preferring data over ego. (Do we have a similar situation today with John Bogle, founder of the Vanguard Group and creator of its index funds, versus stock-pickers? I’m just asking.)

Unless it’s a really, really special individual, a rat knows nothing about random-number generators, probabilities, and expected values. The rat is doing something different. It learns from its experience. It is satisfied with its simple solution. And it will learn something new if its solution stops working.

Rats evidently have a competitive advantage over humans in making predictions under certain conditions. Anyone who steps back and says “wait a minute, what if…” may be channeling his or her inner rat. Copernicus did. Alexander did, as he sliced the Gordian knot. I’ll wager that people who come up with the innovative, disruptive strategies we prize so much — “it’s so obvious, why didn’t I think of it?” — are doing it too.

The rat isn’t doing anything a human can’t. The rat is doing something a human won’t. To com­pete effectively against humans, sometimes it pays to think like a rat.

“Clear? Huh! Why a four-year-old child could understand this report. Find me a four-year-old child. I can’t make head or tail out of it.” — Groucho Marx as Rufus T. Firefly in Duck Soup

Share This Comment

Comments
No Comments