The Ludic Fallacy consists in the error that probability can be calculated on unclosed systems, whereas outliers are of greater influence on consequences that change state than are regularities that maintain state.
In other words, there are very few conditions under which dice are a model for probability, and the ratio of influence (change) is a log of the tail. Dice are closed systems. There are no outliers. Whereas in all other categories (real world) we are almost always measuring variations in a norm, not possible outliers – which although rare, are far more influential than the regularities we measure. In other words, we get what we measure but what we measure is largely unimportant, because it’s obvious and not influential. What we don’t measure is that which is not obvious and rare, but influential.
When we predict the future we depend upon regularities. but if regularities exist then there is no profit to be made. it is from outliers that profits are made.
This is a via negativa strategy, just as is falsification.
Or stated otherwise, the unimaginable and improbable is more influential than the imaginable and probable.
This is – reductio version – the whole point of Taleb’s work.
And Taleb is, even if he doesn’t succeed, the counter to Keynesian Probabilism, the same way I am counter to Marxist pseudoscience.