3 ms·

I think generally modelling life in 2d grid with simplifications such as talent = 0.0 - 1.0 and events doubling or halving capital with a deterministic number o

I think generally modelling life in 2d grid with simplifications such as talent = 0.0 - 1.0 and events doubling or halving capital with a deterministic number of "good/bad" events isn't going to give you anything real as output. I don't like this sort of science, it stinks of foregone conclusion. Running the model itself is irrelevant as the outcome is pretty obvious by the maths put into it. How far is it really from just writing console.out("my hunches are correct!"); Maybe in a future world of modelling where we feed real-world data into the model then maybe we can get somewhere?

BlackFly 9y ago
They introduced talent as a bias in capitalizing on doubling events (the less talent, the less doubling events in proportion). A priori you would consider that this should lead to the talented being overrepresented in the higher percentiles of capital, yet they discover this bias does not dominate over the underlying uniform randomness of merely encountering a doubling event. So no, I wouldn't say the conclusion is foregone. It shows an example where something that can reasonably be called talent (the ability to capitalize) is not as important as luck (the chance encountering of an event where you can capitalize).
- Quarrelsome 9y ago
  > So no, I wouldn't say the conclusion is foregone. With the distribution of people and events over space and the distribution of talent the end result was a given because its more likely for an arbitrary person to trigger multiple good events than for very specific "talented" people to trigger multiple good events even if their factor means they don't need as many good events to be "even". Same way that if you had more events then people then the outcome would completely change because then "talent" would become a bigger factor than "finding an event". Its just about what they decided to make scarce (events) in the model.

BlackFly 9y ago

They introduced talent as a bias in capitalizing on doubling events (the less talent, the less doubling events in proportion). A priori you would consider that this should lead to the talented being overrepresented in the higher percentiles of capital, yet they discover this bias does not dominate over the underlying uniform randomness of merely encountering a doubling event. So no, I wouldn't say the conclusion is foregone. It shows an example where something that can reasonably be called talent (the ability to capitalize) is not as important as luck (the chance encountering of an event where you can capitalize).

Quarrelsome 9y ago
> So no, I wouldn't say the conclusion is foregone. With the distribution of people and events over space and the distribution of talent the end result was a given because its more likely for an arbitrary person to trigger multiple good events than for very specific "talented" people to trigger multiple good events even if their factor means they don't need as many good events to be "even". Same way that if you had more events then people then the outcome would completely change because then "talent" would become a bigger factor than "finding an event". Its just about what they decided to make scarce (events) in the model.