Sana Khadar

Those weasels!

I should say, they were lying of their own volition.

Participants were not instructed to lie.

It's just that the conditions of the game incentivize people to lie sometimes.

Kind of sounds like a lot of real-world scenarios, but anyways.

From that game, Arthur and his colleagues could start to construct a lie detector.

Now, before we get carried away, I should remind us that this was a specific experimental environment.

In the context of a game, it's not the real world.

But even with those caveats, you might be thinking, wow, 70% is pretty good.

Certainly it's better than chance.

But as you might expect, there is a catch.

You just had to state your preference.

And you could be as selfish as you wanted, but you weren't lying.

In case it's not clear, that is not good.

Oh, and by the way, this is also the broad problem with polygraphs, the lie detectors you see in movies and TV.

They're not used in Australia, but some law enforcement in other countries do still use them.

Anyways, with their lying data and the selfishness data, Arthur and his colleagues did something interesting.

Over simplifying things, if lying and selfishness are two different frequencies, they kind of subtracted the selfishness signal from the lying signal.

This is an enormous job for the future because you probably need to do a similar sort of study to what Arthur and his colleagues did with every mental state that might potentially overlap with lying.

Nervousness, weighing up risks and reward, arousal, fear, the list goes on and on and on.