NaClhv

(This is a work in progress, and it may take a while. Come back when it's done.)

How do we make good decisions? This is arguably the most important question that anyone can answer. And while a perfect decision-making process is probably impossible to specify, maybe we can outline a good-enough version. It would go something like this: we explore our options, gather and process the information, then choose and act out the decision. But that's probably still too abstract to be useful.

So let's narrow our focus, to just the "information" portion of that process. Now, information comes in two flavors: it is either continuous, like the weight of a cat, or categorical, like whether it's alive or dead. In particular, such either/or categorical information of just two states is also said to be a binary. It has a special place in decision making, in that all decisions in the end turn out to be binary decisions. You either do, or do not. There is no try.

At this point, we can be quite pedantic about this classification system. Are there other ways to represent information? Can one type of information be used to represent the other? But again, I want to avoid being overly abstract here. So let us narrow down still further: the goal of this post is to focus on a specific kind of error in decision making - that of taking some continuous information and pigeonholing it as a binary. This oversimplification then gets used to mock or belittle a good rule of thumb for decision making, and thereby encourages bad outcomes.

Let's get into some examples. If you want to lose weight, it's worth trying to replace your regular soda with diet soda. But then, "oh, yeah, like THAT's really gonna help" says the mocker. "Go ahead and order a diet soda with your extra-large combo meal. You're surely going to look like a swimsuit model after THAT meal". Well, is the mocker right? And if he's not, what's the nature of his mistake?

Obviously, metabolism and weight loss are immensely complicated topics, and the effects of of diet soda consumption is inconclusive - as a simple google search will show. No doubt it depends on exactly how you're consuming the soda, what the counterfactual is, and much more. I'm not trying to settle that issue here, nor is that the argument of the above mocker. Rather, the mocker effectively reduces all this complexity to a simple binary bit: it will or will not help. And since most diets fail, he claims that yours will too.

This is, of course, a lazy way to think. After all, who wants to think through all the metabolic pathways and calorie counting and behavioral psychology? Isn't it easier to just reduce all this information to a binary bit? Of course it is -and that's why lazy thinkers do it. And their decision making power suffers as a result.

How about another example? Have you heard that, in case of a nuclear attack, you should try to get underneath your desk? This is actually a good rule of thumb. And yet, the mocker says "oh sure, like your desk has any chance of withstanding a nuke!" In doing so, they collapse a great deal of information into a binary bit. For instance, they ignore the distance between your location and ground zero, the presence of walls and other shelter that can mitigate the blast, and the dangers of flying debris or direct radiation blast. And indeed, the level of damage in such an explosion falls off quite rapidly, and far more area would suffer "light damage" compared to the devastation at ground zero.

Here, a comparison with earthquakes is useful. Many earthquakes are more powerful than many nukes, and your desk will not help you with either if their full energy is focused on you. But much of their damage is done in the periphery, and for earthquakes, ducking under your desk is well-understood to be sound advice.

In both cases, note how the information changes throughout the decision process. The complex, multi-variate /ignorance

Make it VERY short? Yes.

The correctness of a decision is binary, but the wrongness is continuous. There is only one (or a very small fixed set) of "best" decisions, whereas the wrongness can be irrelevant to catastrophic.

The decision itself is binary, but the alternatives are often continuous.

Information is often best collected on a continuous manner, but this is, in a pretty deep sense, an admission of ignorance.

But everything is expensive, and often times shortcuts are necessary. You have to respect that.

Mixing this inappropriately generally leads to bad results.

When binary is right

When binary is wrong

Always towards clearer picture.

In the end, do or do not. Right or wrong. There is only one right answer.

But there are a continuum of decisions to make.

The decision-making process requires full calculations. But not all calculations can be full.

Measure what you can, and make measurable what you can't.

In addition, make finer shades whenever possible.

Don't do Lazy binaries, or out of judgement

Yes, you should hide under your desk for a nuke

Yes, ordering a diet coke helps after ordering a big fast food meal

General "falling off the wagon"

"did you know the nazis were bad? That's why you should be communist!"

Yes, setting arbitrary cutoff for qualifications are bad (e.g. for dating). Even non-arbitrary cutoffs can be bad.

You're dead or alive. You're pregnant or not.

Even these binaries break down into continua, but even those continua break down into binaries.

"statistical significance" is often oversold. So is "trust the science".

Yes, laws are not absolute

Laws mediated through many layers

Laws to account for random chance, etc.

Games like chess or Go, and a good or multipurpose move.

"if 100 guilty man go free.." etc.

"let justice be done though the heavens fall"

Lots of errors in politics: "you'll waste your vote", "you have to vote: more voter participation", "if A then B then C..."

Gender binaries.

"If you break one law, you're a lawbreaker"

"if you're not for us, then you're against us"

"if you're not against us, then you may be for us"

Go directly into "why laws are not absolute"?

ACTUAL OUTLINE:

The case of a nuke - complex info has to be reduced to a simple rule.

The case of a diet - complex biochemistry has to be reduced to a simple rule.

In a democracy, it's not possible to reduce to a simple rule. It's complex issues, and everyone votes. If it were possible to reduce to a simple rule, we would not need to vote. That's why independence is important, and why everyone tries to subvert it.

Other concerns - more general comments about decision-making

When deciding on a real-world rule, all this has to be reduced down to a simple rule.

All real-world rules are rules of thumb. True of actual rules, true of laws, true of even Divine commands.

All rules meant for the real world have exceptions. That's what rules ARE. They're these things with exceptions.