>> It could be that the results were built into the model’s design.
>Every true mathematical proposition is a tautology. I don't find this point particularly convincing. As long as the model's predictions match reality in interesting ways, it can lend possible insights on why reality might have certain characteristics.
I think the problem is that nothing about the model lines up with reality. For example:
- Each agent begins with an identical amount of capital. -- nothing like the real world.
- Every 6 months you have the possibility of doubling or halving your capital. -- nothing like the real world.
- Chance of doubling capital is proportional to talent. -- nothing like the real world.
As an argument for such a simple model they provide:
> The previous agents’ rules are intentionally simple and can be considered widely shareable, since they are based on the common sense evidence that success, in everyone life, has the property to both grow or decrease very rapidly.
Sure, success/failure _can_ increase or decrease rapidly, but is it the rule? Arguing the basis of your entire model on "common sense" seems weak. It would be better if they had provided at least some real world data where this pattern emerges. In my experience people seem to move a little bit up or down from their baseline, but I have never personally seen a swing from rags to riches or the other way around.
If they had begun by drawing parallels to the ways in which we see the real world and letting the model simulate it into the future it would be more compelling. As it stands, it seems like a set of arbitrary rules designed to reach their desired conclusion, albeit weakly through "it looks pretty similar".
> - Each agent begins with an identical amount of capital. -- nothing like the real world.
Not relevant. Part of the main point was that even in the case of equal opportunity/equal starting points, bad luck readily overcomes talent, good luck rewards even the less talented, and so talent isn't any kind of guarantee, contra many people's claims.
> - Every 6 months you have the possibility of doubling or halving your capital. -- nothing like the real world.
I don't think this is relevant either. You can pick any progression you want and it won't affect the final results. The point is to simulate regular opportunities of dramatically increasing one's wealth. So again, even when faced with equal opportunity, dramatic wealth inequality just due to luck seems inevitable.
> - Chance of doubling capital is proportional to talent. -- nothing like the real world.
The claim that wealth is proportional to "talent" is widely believed. This paper puts it to the test and refutes it. That's one of the main points.
> In my experience people seem to move a little bit up or down from their baseline, but I have never personally seen a swing from rags to riches or the other way around.
I addressed this in the post you replied to: small swings up and down are probably wage-driven. This doesn't explain wealth disparity, because dramatic swings in wealth will largely not be wage-driven. These events thus aren't of interest.
> If they had begun by drawing parallels to the ways in which we see the real world and letting the model simulate it into the future it would be more compelling.
Not as easy as you think. We have no idea what factors are involved. That's why we study models, and see how closely the results match with reality. That then gives us insight into how some parts of reality might work.
In their model the only benefit of talent is it makes you more likely to double your wealth given luck. This means that they assume luck is needed for advancement while talent is not. Hence they assume that luck is more important than talent, and it's unsurprising that they see the result.
Likewise the power distribution is built in. Power distributions are created by multiplicative laws and the only ways wealth changes in this model are multiplication and division. This means they also assume the power distribution they claim to discover.
It's definitely very hard to build models that match reality, but this one doesn't really seem to try and it's outcomes are so predictable they don't offer meaningful insight. Luck surely plays a huge role in success, but something like longitudinal studies where we measure the abilities of children and then follow their outcomes are likely far more insightful.
> This means that they assume luck is needed for advancement while talent is not.
Which is clearly true. You can be as talented as you like, but talent without a corresponding opportunity to use that talent for monetary gain obviously yields no wealth. There's nothing objectionable here.
It's also widely believed that greater talent makes one more successful at exploiting opportunities, and the paper accepts this premise in order to test it.
Now you can quibble that perhaps the random variable governing opportunities is not independent of talent, in that, those with more talent are simply presented with more opportunities. Perhaps that's true.
But the paper actually shows that our world is also consistent with the hypothesis that opportunities are completely random.
> Power distributions are created by multiplicative laws and the only ways wealth changes in this model are multiplication and division. This means they also assume the power distribution they claim to discover.
I think you're looking at this wrong. The power distribution is not the discovery, the discovery is that what we see in real life is consistent with such a simple formulation based on a small number of factors where wealth disparity is largely dominated by luck.
Perhaps that's not actually the case in reality, but we now know that it's absolutely not impossible, which some previously believed.
> Luck surely plays a huge role in success, but something like longitudinal studies where we measure the abilities of children and then follow their outcomes are likely far more insightful.
These aren't mutually exclusive. Modelling and empirical studies both yield insights.
You listed 3 ways in which the model is "unrealistic". However, it seems to me that neither of the ways that are supposed to be more realistic will actually help the talented against the lucky.
Maybe you disagree and can expand why do you think making things more realistic in some aspect will make it behave better for the talented (especially on the high end of income distribution).
>Every true mathematical proposition is a tautology. I don't find this point particularly convincing. As long as the model's predictions match reality in interesting ways, it can lend possible insights on why reality might have certain characteristics.
I think the problem is that nothing about the model lines up with reality. For example:
- Each agent begins with an identical amount of capital. -- nothing like the real world.
- Every 6 months you have the possibility of doubling or halving your capital. -- nothing like the real world.
- Chance of doubling capital is proportional to talent. -- nothing like the real world.
As an argument for such a simple model they provide:
> The previous agents’ rules are intentionally simple and can be considered widely shareable, since they are based on the common sense evidence that success, in everyone life, has the property to both grow or decrease very rapidly.
Sure, success/failure _can_ increase or decrease rapidly, but is it the rule? Arguing the basis of your entire model on "common sense" seems weak. It would be better if they had provided at least some real world data where this pattern emerges. In my experience people seem to move a little bit up or down from their baseline, but I have never personally seen a swing from rags to riches or the other way around.
If they had begun by drawing parallels to the ways in which we see the real world and letting the model simulate it into the future it would be more compelling. As it stands, it seems like a set of arbitrary rules designed to reach their desired conclusion, albeit weakly through "it looks pretty similar".