Its no more ivory tower wishful thinking for the other applications than for predicting behavior; they are all idealized in that one never has certain knowledge in any of everything relevant in any of the dimensions.
> they are all idealized in that one never has certain knowledge in any of everything relevant in any of the dimensions.
Which is why it's particularly useless for geopolitics. It's literally the furthest you can get from "idealized" anything - far more so than plain sociology, economics, or psychology. It's the global sum of all of the most unpredictable traits of humanity. It's one of the messiest human pursuits and always has been.
It's a nice "cows are spherical" approximation to write papers about but game theory really ain't that great at predicting basic human behavior either. Except like with evolutionary biology you can always come up with a "just-so" game theoretic story that sounds right but has no real predictive ability.
similar to math, just because you dont know all the variable to solve an equation, doesnt mean it is useless. It is a tool to study the interactions and possible solutions, inferring some and ruling out others.
Just because cows aren't spherical doesn't mean it is useful to act as though the world is devoid of physics, random and inscrutable.
We're not talking about an extra variable here or there, we're talking about the equivalent to string theory which is famously intractable. It's an n-body problem where n = ?????? and you don't know any of the boundary conditions.
It's not random and inscrutable, but it sure isn't conducive to naive academic analyses like game theory. Just because you can model something simple, doesn't mean you can predict something as complicated as reality. The science of biology, sociology, psychology, and economics have proven that over and over again - geopolitics is all of those combined and more.
maybe there is a disconnect on what we call prediction. There is more than enough information to predict if China would be in favor or against US building a military base in Taiwan. There isn't enough information to predict what Xi Jinping will have for breakfast.
There's definitely a disconnect because if that's the best example of a game theoretic prediction, I stand by my statement that game theory is useless in geopolitics.
What would China's first concrete move be after it's final final warning [1]? Which faction in the party would have most influence over Xi Jinping in such an event and how much influence do they have over the military? How much would China be willing to compromise it's status as a trading partner with the West if push comes to shove? Those would be useful predictions that can be acted on.
China's final warning meme is actually very illustrative. The context is PRC was both warning and doing - actively attempting to shoot down TW / US based U2s despite having limited capabilities. The chefs kiss that this is an USSR meme is that PRC end up shooting down more U2s using modified soviet hardware than USSR herself. Even more so when consider PRC issued final warning to USSR that ended up in border skirmishes. PRC's actual final warning is "don't say we didn't warn you" has been predictative of PRC kinetic action with near certainty. USSR/India border skirmishes. Korean war against UN. PRC also has directly supported North Vietnam against the French, and threatened UK when they hinted at granting HK independence under Thatcher. That's every NPT nuclear state over territorial/security issues less important than TW. It doesn't always lead to immediate action, but has consistently been prelude to it.
