[Previous] Toxic Attitudes about Greatness | Home

Chess Time Loop

There was a viral discussion about being trapped in a time loop that you can only escape by winning a game of chess against Gary Kasparov (one of the all-time best chess players). I think the discussion is bad and it going viral shows bad things about people.

The scenario is ambiguous. Removing ambiguity from the problem/scenario/goal is one of the main things Critical Fallibilism (CF) says to do at the start of the discussion, before debating, quarreling, or reaching conclusions. So I want to talk about how CF would approach this and how it differs from typical discourse.

First, as a minor point, the vast majority of people participating in this discussion don't know enough about chess to have an informed opinion. It's viral among regular people, not just chess players. Yet people who are ignorant about chess seem to like arguing over this scenario about chess. CF says to consider what prerequisites you need in order to realistically and effectively accomplish your goals (like figuring out the correct conclusion to a scenario), not to ignore the concept of prerequisites.

One of the main reasons people disagree about the chess time loop scenario is they interpret it in different ways. Each person bases their conclusions on their interpretation. Then they disagree with people about their conclusions and argue over them. So, as I see it, they are debating a downstream consequence of an earlier disagreement, and they aren't focusing their debate on the earlier disagreement, the origin of their disagreement. They aren't looking for the root cause of their disagreement and trying to focus discussion there. This doesn't make sense logically or organizationally. You don't have to always look at root causes, but they're often useful to consider.

For an example of different interpretations, some people think that Kasparov would play a variety of different moves in the same position. Other people think if you reach a specific position, Kasparov would always play the identical move. Whether he's extremely predictable, or not, makes a large difference to the difficulty of beating him. But from what I've seen (and admittedly I didn't research this much), people are more inclined to debate their final conclusions than their interpretations/premises like whether or not Kasparov would always play the same move in the same position.

I've also seen people say basically that determinism implies he'd always play the same moves in the same position. This is false and careless. Determinism (with some standard assumptions like no multiverse) implies he'd always take the same actions in the identical situation. But the current position on the chess board is not the full situation. Even under determinism, he could play different moves, in the same position, depending on your demeanor and how long you spend on your moves. If you played particularly quickly or slowly, or seemed especially nervous or calm, any of those (and many more subtle things) could cause Kasparov to go down a different train of thought in his head and therefore end up playing a different move, even under determinism. If you want to repeat moves based on a deterministic premise, you need to be a great actor who can put on the identical show every time. Or maybe a similar show would work. You'd have to find out from experience how much leeway you have to behave differently on different moves, as well as how much your behavior can result in changes on later moves not just on the next move.

Also, if you have to behave the same to get the same moves from Kasparov, then it may ruin the strategy of switching colors and playing his own moves against him. Some people assumed you can choose your color at the start of the game, so you can play black for the first game and see what Kasparov does, then resign after he moves, play white in your second game, play his move, see his reply, remember it, resign, play black again and find out his second white move, etc. But this means I need to memorize one series of moves for black, and one series of moves for white, and the assumption is I can play on either side and Kasparov will always play the same moves. If I'm white and I use the same demeanor and move speed every time, maybe that works. But how can I duplicate the same game with black? I'd have to, by trial and error, figure out some different demeanor, suitable to a different set of moves, that'll cause Kasparov to play the white side of the same game I got him to play with black? Just using my white demeanor, while playing black, wouldn't make sense because the moves are different and have different conceptual meaning. And if I don't know their conceptual meaning, my demeanor will be off, and Kasparov will notice something is weird, and he won't play the same moves he normally would against a strong player. The idea with switching colors is basically to get him to play against himself, but if you don't know how to act like a strong chess player it won't work right. If you play Kasparov's moves that you memorized, but you don't act like a strong player, he may change his moves.

Also, as someone who actually knows about chess, I can tell you the likely result of getting Kasparov to play against himself: a draw. Draws are the most common result among top players. You have to win, not just draw, so you have to outplay him somehow, not just match his skill level. Chess has a large margin of error to get draws; it takes a significant advantage to win; small advantages often fizzle out to draws instead of being increased to decisive advantages. Usually losing requires making a bunch of mistakes that add up; one severe mistake works too but grandmasters are great at avoiding huge mistakes.

Anyway, there are different interpretations of the scenario, and people bicker about the right conclusion to reach, and that is viral for some reason among people who know little about chess, logic, philosophy, debate or the physics of time travel.

I suspect many people like debating something where they secretly feel there's no right answer. They act like there is a right answer. They'll be pushy about their answer. But I suspect deep down they feel safe because it's silly and they can't really be wrong because the scenario is impossible and ambiguous. They want to bicker over stuff where there is no objective truth for smarter people to discover and win with.

If people were being reasonable, many of them would be less interested or approach it in other ways such as how CF would: make a list of interpretations of the scenario, make a list of conclusions of the scenario, then match them up. For each interpretation, you could agree on what conclusion makes sense. You could still disagree about which interpretation to believe, but you could discuss some stuff objectively and reach a significant amount of agreement about issues like which versions of the scenario imply which outcomes. You could agree about whether many scenario plus conclusion pairs are correct or incorrect. This is like the CF method of pairing multiple ideas with multiple goals and evaluating the pairs instead of trying to evaluate the ideas independent of the goals or with ambiguity about which goal(s) you care about.

But people aren't trying to have structured, productive discussions. That's not what they want from the scenario. That isn't why it's viral. They aren't approaching it with the attitude of a scientist or rational philosopher. They don't know how to and don't seem very interested.

As to the right interpretation of the chess time loop, there isn't one. I think that's the objectively correct, logical conclusion, but I grant that it's not a very satisfying or fun conclusion. The scenario wording is ambiguous and the scenario also violates the laws of physics. Sometimes you can figure out a good interpretation of something ambiguous by thinking about what makes sense, and bringing in additional constraints on what could be meant like physics or logic. But this scenario explicitly does not follow the laws of physics. Time loops aren't a real thing. There is no science that can help fill in the details that the problem statement left out. People tend to assume things like the amount it violates the laws of physics or logic should be minimized, but that doesn't actually make sense and isn't meaningful guidance. It's well known that, logically, you can use a single contradiction to reach any conclusion whatsoever. One contradiction, which violates the laws of logic, implies anything at all, of your choice, which you can prove using the standard laws of logic plus the one contradiction. A single contradiction completely breaks all of logic. Similarly, there's no good physics-based way to reason about how time loops would work since they're simply made up and not part of real physics. Trying to extrapolate from real physics to them isn't going to work well because real physics doesn't allow them. If it did work well, people interpreting the scenario in different ways would be less of a problem.

What people are actually doing is closer to extrapolating how time loops work based on some movies and TV shows they've seen. So the answers they reach depend on their TV viewing history and memory, not on physics or logic. They wouldn't want to admit this, but TV, movies, video games and novels are probably where their intuitions about time loops come from. Sometimes they like to pretend to be reasoning like scientists and logicians, but they're using a ridiculous scenario that's actually based on fictional stories they watched actors portray on sets that would look extremely fake when viewed at most angles and distances besides the actual camera angle and distance used. And the actors did many takes and the scenes were edited together. And they based it on scripts that were edited: people made stuff up, then crossed things out and made up different stuff.

But, again, it's viral. Something about this kind of unproductive debate, that could never be very productive, really appeals to millions of people. And it could be made more productive with CF methods – discuss the scenario, clarify it, write down multiple interpretations, and then evaluate the right conclusion separately for each interpretation. But I don't think most people want to do that. I don't think that would appeal to them even if they knew that method existed. They know the scientific method exists and they aren't trying to use it. They know there are experts at debate and they aren't trying to research what those people do and emulate it. They're just quarreling on social media because, apparently, they wanted an opportunity to do that and jumped at the chance even though they aren't chess players.

I think the thing they're doing is basically what they want to do and like. It's not all that different than bullying people who don't wear fashionable clothing. People are used to, in many parts of life, taking sides on arbitrary issues. Sports fans will strongly take sides in favor of their team even though it doesn't really matter who wins the sports game and they don't even try to discuss which team objectively merits the win more or how that could be determined.

This article has over-generalized some. People vary. I'm trying to comment on some themes and issues that apply to a lot of people but certainly not everyone. People are individuals but there are also some common patterns that come up with many people that can be worth understanding. Even when a pattern applies to a person, they're still also an individual who generally doesn't fit the pattern perfectly, just pretty well (or you could build a lot of margin of error into the pattern, in which case people could fit perfectly because the pattern itself is less specific and gives a lot of room for variety and individuality).


Elliot Temple on June 6, 2025

Messages

Want to discuss this? Join my forum.

(Due to multi-year, sustained harassment from David Deutsch and his fans, commenting here requires an account. Accounts are not publicly available. Discussion info.)