## Saturday, July 15, 2023

### Uncertainty and Randomness

We are fans of both, some links below.

From The MIT Press Reader, July 11:

The Psychological Depths of Rock-Paper-Scissors
An excerpt from veteran game designer Greg Costikyan's book "Uncertainty in Games."

Unless you have lived in a Skinner box from an early age, you know that the outcome of tic-tac-toe is utterly certain. At first glance, rock-paper-scissors appears almost as bad. A four-year-old might think there’s some strategy to it, but isn’t it basically random?

Indeed, people often turn to rock-paper-scissors as a way of making random, arbitrary decisions — choosing who’ll buy the first round of drinks, say. Yet there is no quantum-uncertainty collapse, no tumble of a die, no random number generator here; both players make a choice. Surely this is wholly nonrandom?

All right, nonrandom it is, but perhaps it’s arbitrary? There’s no predictable or even statistically calculable way of figuring out what an opponent will do next, so that one choice is as good as another, and outcomes will be distributed randomly over time — one-third in victory for one player, one-third to the opponent, one-third in a tie. Yes?

Players quickly learn that this is a guessing game and that your goal is to build a mental model of your opponent, to try to predict his actions. Yet a naïve player, once having realized this, will often conclude that the game is still arbitrary; you get into a sort of infinite loop. If he thinks such-and-so, then I should do this-and-that; but, on the other hand, if he can predict that I will reason thusly, he will instead do the-other-thing, so my response should be something else; but if we go for a third loop — assuming he can reason through the two loops I just did — then . . . and so on, ad infinitum. So it is back to being a purely arbitrary game. No?

No....

....MUCH MORE

On randomness:
Randomness: "A Drunkard’s Walk in Manhattan"
"...Believe the seemingly impossible — that you can win at a number guessing game with absolutely no information".
The Black Death: Ecological Invasion Resembles a Drunken Walk More so Than Waves
The Case For Using Random Benchmarks In Portfolio Analysis
"Detox democracy through representation by random selection"
That last piece of research was awarded Harvard's own Ig Nobel prize in 2010.
Ya see, ya got your complex systems and ya got your chaotic systems and then ya got your complex-chaotic systems like weather or the economy or the stock market and when you endeavor at those levels of sophistication you realize:...

There may be issues.

And on uncertainty:

New York Fed on Forecasting: "What Does Disagreement Tell Us about Uncertainty?"

"The Greedy Doctor Problem"

The Bezos '70 percent rule' for decision-making"
There is a history of Decision Making Under Uncertainty* as a distinct discipline going back to Pascal and his wager and Bernoulli and his Expected Utility vs Expected Value. It is now taught as the heart of Decision Theory and as a cousin of Game Theory.

Related:
Ogilvy's Rory Sutherland Has Some Thoughts
"Poker and the Psychology of Uncertainty"

That second post has many many links, including:
Paul Tudor Jones On 'Imperfect Information'

"How to Choose With Less Than Perfect Information"
Less-than-perfect-information is a term from decision theory.
"Poker, Speeding Tickets, and Expected Value: Making Decisions in an Uncertain World"

"The future has become a financial fetish, sliced into probable outcomes to be calculated and bet upon."

Richard Feynman: "THE RELATION OF SCIENCE AND RELIGION"

And, as always, many more. Use the 'search blog' box upper left if interested.