Playing for the singleton queen

A few days ago I made some predictions about the “Ashes” Fifth Test at the Oval, a cricket match between England and Australia. My predictions were: England would win the toss, bat first, put up a decent score on the first day and part of the second day; that Australia would come in on the second day and crumble; that England would proceed to set up the game for a solid win; that Australia would fight back heroically but ultimately to no avail, they would lose. And England would win the Ashes back.

Well, England won the toss, batted first, put up a decent score. Australia came on the second day, crumbled. And England are set fair to win.

Since I’d shared my predictions on Twitter, there were comments about my prescience. And I promised to respond via a blog post. Which is what I’m doing here.

First off, there’s no need to call James Randi in, there’s no sixth sense here, nothing paranormal. I am not prescient. Nor do I claim to be.

Here’s all I did:

  • I learnt that I was going to be at the Oval on the second day.
  • Since I’m a cricket fan, I had a reasonable idea of the state of play between the teams, the strengths and weaknesses.
  • Using those two bits of information, I constructed a scenario that would make for an exciting second day’s play, but which ultimately would lead to the result I wanted, an England win.
  • Then I proceeded to share that scenario in public.

In summary, I took the facts as I knew them, built them into a hypothesis that was sympathetic to my frame of thinking, then published that hypothesis. It was a process that I learnt from observing my father playing contract bridge, something he described as “playing for the singleton queen”.

Let’s say you’re declarer at bridge, the lead has been played, dummy’s come down and you’re assessing your options. You realise that there is just one way to make contract is to have the spades break 4-1 with the singleton on your left, and for that singleton to beĀ  the queen.

You then have two choices. You can work out the probability of a 4-1 break with the singleton on your left: the spades could break 0-5, 1-4, 2-3, 3-2, 4-1 and 5-0. You can then work out the probability of the singleton being the queen. And you can convince yourself that the chances are too low and start playing to curb your losses.This is natural, since we all have loss aversion.

Or you can play for the singleton queen to be on your left. And find the only way to make contract. Also a valid choice, but one far less likely, especially if you are loss averse.

Back to the cricket. I wanted England to win, I still want England to win, but I wasn’t going to get upset if England lost. Similarly, I didn’t mind being proven wrong with my prediction. It was nothing more than a way of constructing the known information into a sympathetic hypothesis. Loss aversion didn’t come into it, since the loss would have no material effect on me.

Whenever the win or loss has no material effect on me, I try and play for the singleton queen to be on my left. Why ever not?