“Now you’re just arguing semantics.”

It’s an accusation commonly employed in debate as a way of suggesting the other party is attempting to dismiss a valid point on a linguistic technicality. Quite often, however, it’s the accuser who is actually arguing in bad faith. People implicitly redefining terms midway through a debate is a common yet dirty tactic, and pointing out when it’s happening is totally valid.

That said, debates can easily be derailed by pointless semantic quibbling, but this generally requires both parties to be complicit. What a word or phrase means according to the dictionary is of minimal importance, but it’s crucial that both parties understand and use it the same way, and that this meaning remains consistent throughout.

Quite often, in such cases, digging into the semantics of a term is exactly what the debate needs. The way we use words is reflective of our psychology, and examining what someone understands by an expression can reveal a lot about how they’re thinking about the subject as a whole, and why they’ve come to the conclusions they have.

What is a “Muslim ban”?

I’m bringing this up because of an exchange I had on Twitter with popular poker ambassador Matt Glantz. The subject was US politics in general and Donald Trump’s recent executive order on international travel specifically. I realize that I’m paid to write about poker, not politics, and I promise to bring this around to the subject of poker in a bit. Just bear with me for now.

Matt’s stated stance is that he opposes the executive order, but feels that those opposed to it are engaging in disingenuous rhetoric by calling it a “Muslim ban” and that doing so hurts resistance to the order more than it helps. His rationale is that not all Muslims are banned, only citizens from a short list of countries considered hotbeds of extremist Islamic violence, and that these citizens account for only 20% of Muslims globally by his arithmetic.

I like Matt a lot, and have generally found him to be reasonable in the past and capable of changing his views when provided with a convincing argument. In this case, however, I think Matt is emotionally invested in minimizing the gravity of the situation, for the simple reason that he was at one time a Trump supporter (though to his credit he changed his mind shortly before the election). More specifically, he was a Trump supporter of the camp that believed a serious disruption of the US government was needed, that Trump would accomplish that, but that his most shocking statements during the campaign were just pandering for Tea Party votes and that he had no intention of keeping any of his outrageous promises.

Of course, high on the list of those promises was a ban on Muslims entering the United States. It’s easy to understand, then, why Matt might find it emotionally difficult to call the order in question a “Muslim ban.” And yet, his point can’t be dismissed as mere semantic quibbling: There is a real question to be asked here, and it’s one that’s going to be asked a lot in the coming years.

The sorites paradox

One of the last tweets I exchanged with Matt on the subject was the following:

This is something known as the “sorites paradox,” a philosophical conundrum dating back to ancient Greece. If you’re not familiar with it, it may seem unrelated to the subject at hand, and perhaps that’s why Matt didn’t respond. It’s also possible that he did catch my meaning, and simply didn’t have a good answer.

The sorites paradox is something that arises when one asks strict yes or no questions while using terms which seem clear enough on the surface, yet lack specificity when examined more closely. The sand pile (or “heap”) example is a common formulation of it. The following three statements all seem intuitive enough, but can’t all be simultaneously true:

  • We all understand what a pile of sand is, and can correctly identify one when we see it.
  • A single grain of sand is not a pile of sand.
  • Adding just one grain of sand to something will not make it a pile of sand if it wasn’t one already.

In other words, if we understand “a pile of sand” in a rigid sense, then there must be some minimum number of grains of sand you need to collect in one place to form a pile. And yet, it seems absurd to say that removing one grain of sand from a sufficiently small pile would instantly make it not a pile.

Mixed strategies and opening ranges

I promised that this was all going to have something to do with poker, and it does. Poker is a game of imperfect information, and like most such games, its equilibria will generally involve mixed strategies. If you’re not familiar with those terms, what it means is that there isn’t one “best thing to do” in most situations, but rather that perfect play consists of doing one thing a certain percentage of the time, and something else a different percentage of the time, perhaps a third thing the rest of the time, etc. That is, holding the exact same cards, on the same board, in the same position, facing the same action, if you’re doing the same thing each time, you’re probably not playing correctly.

And yet, the human mind has a difficult time thinking of things probabilistically, so almost no one plays this way. To the extent that we do make different moves in effectively identical situations, it tends to come down to our emotional state, or perhaps a read, rather than deliberate randomization.

Go to any training site and you’ll find opening range charts. The exact ranges depend on whose you’re using, but the chart may tell you, for instance, that all your suited Kings down to K8 are a raise from the cutoff, but K7 suited or Q8 suited is a fold. For practical purposes, this is good enough, but hypothetical perfect play probably looks quite different. Perhaps KQ suited is always a raise and K2 suited is always a fold, but perhaps KT should occasionally be folded and K5 should be raised some small percentage of the time.

In other words, because of the limitations of human cognition, we ignore this probability gradient and classify hands as either “a raising hand” or “a folding hand.” In theory, maybe K8 should be opened 52% of the time and K7 48%, but because they are not substantially different hands, it’s good enough in practice to always open K8 and always fold K7. Hypothetically, a hyper-perceptive opponent could catch on to this approximation and use it to exploit us on 8- and 7-high flops, but unless you’re playing super high-rollers, very few if any players are that sharp.

If you’re paying attention, it should be clear at this point what the connection is between the Muslim ban, the sand pile and our opening range. The sorting of things into discrete bins is in most cases a convenience of human psychology, and doesn’t reflect reality. One grain of sand is not at all a pile, and a metric ton of sand is definitely a pile, but somewhere in between there exist collections of sand which possess some fractional degree of “pile nature.”

A Zen koan

Although people of all cultures use this kind of shortcut in day-to-day life, its presence in philosophy and spirituality is largely a Western phenomenon. Many Eastern schools of thought more readily embrace the idea that a thing can be its own opposite, that the whole can simultaneously contain parts of seemingly contradictory concepts. That is, after all, what the yin-yang symbolizes.

The yin-yang is primarily associated with Taoism, but Zen Buddhism is likewise a philosophy which embraces paradox, rather than attempting to attack it with rigid logic. Zen koans are, essentially, very short stories intended to draw one’s attention to paradox and thereby help one to escape the trap of rigid black-and-white thinking. The sound of one hand clapping is the koan most familiar to us in the West, but there are a great many others.

The following is a koan called Shuzan’s Short Staff.

Shuzan held out his short staff and said: “If you call this a short staff, you oppose its reality. If you do not call it a short staff, you ignore the fact. Now what do you wish to call this?”

That phrasing is unnatural to the Western ear, so the implication may not be clear.

Shuzan is holding out a stick, which he presumably uses for walking, and he is asking his students whether or not they would call it a walking stick. Neither answer is correct. If you were to ask someone who has not seen Shuzan’s staff to imagine “a walking stick,” their mental image of a typical walking stick would likely differ in several ways from the actual object he holds. And yet, to say it is not a walking stick is patently false, as it is in fact a stick that he uses to help him walk.

So, with that in mind, is a tablespoon of sand a pile? Is K7 suited a raising hand from the cutoff? And what of Donald Trump’s executive order on immigration? Is it or is it not a Muslim Ban?

Mu, as the Zen Buddhists say. Unask the question, because the question itself is wrong.

How to read a poker thread

Strategy threads on poker forums can be very useful. Specifically, the debate itself is often a great exercise in thinking deeply about a hand, putting the opponent on a range, calculating equities and so forth. However, the ultimate conclusion people tend to draw from such threads is often wrong, because the question they have in mind – what is the right play? – is itself wrong.

Since the game theoretically optimal answer to the question is likely to be not an action but a probability distribution of actions, the right question to be asking is “what actions are viable here and how often should I take each?” Most posters will be answering the first, incorrect question by recommending one specific action, but that doesn’t mean we can’t use their answers to produce an answer to the correct question.

The key is in treating the forums community as a single person, and pretending that it would make its decision by choosing a random answer given in the thread. So, if 40% of the answers say we should bet pot on the river, 40% say half-pot and 20% say check-it back, most people are going to infer that betting is the right option and that maybe three-quarters pot as a compromise is the right sizing. Instead, I would recommend interpreting this information as a suggestion to bet pot 40% of the time, half-pot 40% of the time, and check back the remaining 20%. Since the human brain can’t randomize effectively, we’re going to select between these actions based on our reads, perhaps betting full pot against the loosest 40% of our opponents, checking back against the tightest 20% and half-potting it against the rest.

There is one caveat here, which is that this approach to reading a thread works best if all the people involved are similar in skill. People whose justification is hugely flawed or are known fish can be discounted, while the most thoughtful responses can be given additional weight. If Fedor Holz were to grace the podcast with his presence, for instance, and he said to fold in a spot while Andrew and I both said call, the answer would probably not be to call 67% and fold 33%. However, it probably wouldn’t be to fold 100% of the time either; you would have to make a judgment call about how many Andrews a Fedor is worth, and how many Alexes to an Andrew.

Fuzzy logic and fuzzy fascism

A similar principle is at work when it comes to political debate. It’s not usually semantic arguments that derail an otherwise productive discussion, but the assumption that someone in the group has the correct perspective on the situation and that the others are wrong. Going back to the argument which inspired this article, a discussion of Trump’s executive order is never going to reach a satisfactory outcome if we fixate on whether or not it’s correct to call it “a Muslim ban.” Rather, some fruitful discussion could be had about the degree to which it resembles a Muslim ban, after which the conversation could move on to its implications.

Another term for what we’re talking about here is “fuzzy logic,” which is a computing concept. Computers employ binary logic at the fundamental level, encoding things as 0s and 1s. Traditionally, logical statements are interpreted by a computer as a single bit, 0 for false, 1 for true. This is straightforward, but runs into the sorts of limitations we’ve encountered in this article. Instead, sometimes true/false statements are encoded as decimal values between 0 and 1, representing a degree of truth or level of certainty.

Getting used to talking about things in this way is going to be important in the coming years, because the nature of evil is that entities – whether we’re talking about people, corporations or governments – get away with it by maintaining plausible deniability. Whatever your mental model of an evil dictator may be – Adolf Hitler, for most people – Donald Trump is never going to fit it exactly. He’s not literally Hitler, but he is also not entirely unlike Hitler, because no person is. Everyone has Hitler-nature, some just have more of it than others.

No one knows what events will unfold in the next few years, but it’s probably correct to assume that they’ll have greater-than-normal repercussions. It’s important, then, that we remain clear-eyed and not get sidetracked by questions about whether a given label like “fascist,” “terrorist,” “racist,” etc. applies or does not apply to a given person, organization or government. Rather, the question must always be to what extent their behavior and the impact of that behavior resemble the thing in question… and, more importantly, what we’re going to do about it.

Alex Weldon (@benefactumgames) is a freelance writer, game designer and semipro poker player from Montreal, Quebec, Canada.