This one will get weird, but I think it’s cool, so read it all the way through, and if you can’t read have someone else read it to you.

Many of you know that I play poker, and if you don’t you probably missed my article from a few weeks ago entitled Poker at the end of the World. I have a regular, friendly poker game and even before that I was always something of a student of human behavior. Kind of like a profiler, but not for the FBI. I could offer a service where I play poker with you and talk with you a few times and then I write a profile for you… not something like “You are a liar” (everyone lies a bit in poker) but more of a discussion of social and mental “tells” that maybe you yourself don’t recognize. This is part of being a writer since I am always studying people to create realistic characters for my work. But I don’t have to play poker to do it. A couple of conversations over a beer or a cup of coffee will suffice.

But that’s not the topic today. Today I’m going to talk about something I call “Quantifying Unreasonableness” or “Quantifying Unreason.” For those of you who are bad at words, another more awkward way to put it would be: “A sort of math for measuring, studying, and predicting human unreason.” Expecting, studying, and being somewhat able to predict unreason is a highly valuable skill in poker and life.

Now, there may already be a term for this in mathematics, and if there is let me know. But I just make up terms to describe very real things and I’m not a mathmagician.

In poker, VARIANCE is the word we use to describe the swings of “luck” or “chance” that often seem to defy statistical norms. Variance just means a very long and sometimes statistically weird run: Weeks of not getting any good cards, (we often call this ‘running bad’,) or just really weird ways of losing that seem to be almost mystical in their unlikelihood. Likewise, there can be runs of “good luck” where nothing goes bad or wrong. When we are “running good”, we win hands we shouldn’t, and everything appears to be magical for us. In quantifying unreason, people will often play hands that they shouldn’t because they feel that they are operating under a spell of some sort (and sometimes it works.) See the logical fallacy called The Gambler’s Fallacy.

These variances (periods of good or bad runs) are not unscientific or counter-statistical. They are quite reasonable and to be expected. If you flip a coin a million times, statistically your average for heads or tails is going to be very close to 50-50 overall, but there could be runs within that million flips where you get heads 12 times in a row or more. It is not likely, but it could happen. Likewise, humans can and do act unreasonably and illogically, but over time and with sensitive enough awareness and attention and even record keeping, we could quantify this unreasonableness. On a personal (one-to-one) basis, some of us can very loosely predict human behavior based on the skill (or talent) of quantifying unreason. Read the 50-page epilogue to War and Peace where Tolstoy discusses the unlikelihood of any individual Frenchman deciding reasonably to march to almost certain death in Russia under the command of Napoleon, but the fact that almost 500,000 Frenchmen did that very thing.

Quantifying Unreason is a way to describe a measurable (but not always predictable) level of unreason in human behavior and systems. We can use it in things like playing poker, self-help, psychology, and even probably in programming AI (which is neither A nor I, but I digress.) It would also help bridge the uncanny valley, but that’s for another time.

Stick with me…

The point is that people make decisions to “stay,” “call,” or “go all in,” every day, even when it seems unreasonable to do so. Using the term “unreasonable” makes it seem like we’re judging the actual behavior, but we’re not. We’re merely identifying the behavior in relation to what is expected or what would be predicted by a machine (or math.) Why do people stay in difficult relationships? Why do people stick in tough marriages? Why do people stay at jobs that perhaps they do not enjoy? I’m not saying there are no legitimate and good reasons for these things, I’m saying a computer might judge these things (with no human or emotional context) to be unreasonable.

This unreasonableness factor intersects every segment and part of society. And the ability to quantify unreasonableness and use pattern recognition to identify it, can be quite valuable. Replicating it or understanding how to apply it to systems will also, in the long run, make those systems more human.

Someone might get insulted when we say “we’re quantifying the unreasonableness of your behavior,” but that’s just because we haven’t taken enough time to hide or cloak the language in less offensive terms. Unreasonableness (as a computer might measure it) may not even be unreasonable in human terms. It’s just a behavior or action that would unemotionally be considered unlikely based on the available information.

Example: A father throwing himself in front of a car to save his child or a soldier dying for his brothers and his country might be considered “unreasonable” if a measuring system solely considered self-preservation to be reasonable. However, considering human nature and tendencies, sacrificing oneself for others can be very reasonable.

Does a specific behavior that seems to be unreasonable (say… raising a bet in a situation where you might reasonably assume you have a weaker hand or that you shouldn’t win) indicate that the will is operating contrary to preference? (I’m not asking you, I already know.) And can purposely acting contrary to reasonable expectations boost the opportunity for success in certain situations? And can the likelihood of someone acting unreasonable be assumed or predicted (within certain tolerances) enough to give a person (or a machine) a tactical advantage?

Yes.

Quantifying Unreasonableness will be more than a chapter in the book I might write – Game Theory, Logos, and Cocktail Recipes.

***

Michael Bunker is a local columnist for BrownwoodNews.com whose columns appear on Wednesdays and Sundays on the website.