Dara O’Kearney: Poker, Game Theory, and a Squid in a Box

  • At the recent IPO, Thomas Murphy challenged David Lappin and I to a game of Squid in a Box
  • The aim is to interrogate your opponent to determine if they have a squid in their box or not
  • We used game theory to make it more interesting, with my tactic involving an RNG
A mystery box
I took on my VSO News peer David Lappin in a battle of game theory. [Image: Shutterstock.com]

Cash or nothing

I have written in the past about how game theory can be applied not just to poker but to any game in which decision-making plays a part.

1 mark out of 10 “guessed” right

One such example was a game Neil Channing told me about in Vegas years ago that involved a player taking a cash chip, putting it behind his back, and wagering that the mark couldn’t guess the correct hand. Once the mark had accepted the wager, he would bring both his hands to the front for them to guess left or right. The twist was in the presentation: he would thrust one arm so far forward that the closed hand was literally right under the nose of the mark, as if daring him to pick it. The other hand was kept well back. Only 1 mark out of 10 “guessed” right, except it turned out he didn’t actually guess. Instead he used game theory. I won’t repeat how, because you can read about it here.

Squid in a box

At the recent International Poker Open, WPTGlobal social media gofer Thomas Murphy told my fellow VegasSlotsOnline News man David Lappin and me that he wanted to record a short of us trying to outwit each other at “Squid in a box.” For those unfamiliar with the game, it’s basically a variation of “Carrot in a box” where there are two players and two sealed boxes, one of which has a carrot inside. Player 1 unseals one box which has a carrot inside it (or a squid in this case), and is then interrogated by the other player who has to guess whether the box contained a squid or not.

Lappin suggested we use game theory to make it more interesting

As we were setting up, Lappin suggested we use game theory to make it more interesting (and on brand), so on the fly, I went to random.org on my phone and generated a random number that was either 1 or 2 to decide whether I would answer all the questions truthfully or untruthfully. Randomising this decision cut off the possibility that I was likely to choose truth or lie more than half the time, which mattered because someone who knows me as well as Lappin does might have a very good read on any likely such bias.

What I forgot

Lappin, who obviously knew I’d rolled randomly, started asking me questions and realised I’d already let the RNG decide whether to lie or not quickly. But then he realised something else: that I knew whether I was lying or not (which allowed him to try to get a read on that). I then realised on the fly that my GTO solution was inadequate: what I should have done was only pretended to but not actually look inside the box so that I didn’t know whether it contained a squid or not, and used the random number generated to decide whether to claim the box contained the squid or not.

The game concluded with Lappin claiming he was guessing without any read on me. Was he lying? We will never know for sure, but have a look at the short and judge for yourself.

Leave a Reply

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