You are correct, but for the sake of the simulation, I think this works. However, instead of getting "wealth" you get here "a share of all wealth". Basically, you simulate the proportion of the total wealth.
In the real world, the total wealth increases but the proportion might still stay the same. You could basically get the same in the simulation if you were to add some linear offset that's based on the timestep.
If the wealth created in a transaction is proportional to the size of the transaction, that sounds like it would increase the Gini coefficient. Most new wealth is created by transactions involving those with the most wealth, and redistributed to them. Probably simple to check in the simulation. Whether this is any good as a model for the real world is then a different question...
I agree. Both the simulation and the analysis are quite useless for two reasons.
1) It doesn't perform any analysis of the individual population members between time-steps. A members of the population jumps randomly within the income distribution after a transaction and the plots don't account for this. I would like to see the average wealth of each of the population members over time. I'm guessing it will be 100.
2) The transaction function is entirely unrealistic causing and making addressing 1) pointless since each population member is essentially a new born between time steps.
That said, it an interesting exercise in how one can generate a beta-like distribution.
Overall, this is too far removed from reality to be of use in debating how to address income inequality.
Some trades are zero. You exchange money for some good or consumable. For example you buy food at a supermarket. If you were to bake a cake with that food and sell it then it would be non-zero sum but not if you eat it.
If there is no difference in value between the money and the food, then why did you exchange the money for the food? Why not just keep your money and skip the trip to the store?
Same with the store: why bother selling the food for money? If the value is the same, why doesn’t the store just keep the food and save themselves the trouble?
The alchemical transformation between needs is in fact what money is for. And money is lost in the transformation due to friction. The computer age is impressive but easily overlooks advances in fields, like Foucault's Order of Things and his presentation of Economic theory. Such awareness would prevent such drastic oversimplifications imo
> From another angle, why would the customer spend their time (which has value to them) in a transaction where they don't get any?
Well I for one always took issue with the rational consumer model. It is possible that people trade things they value for things they don't, because people do not always act rationally.
That’s a larger discussion but much of the economics of it gets cancelled out with the notion of “revealed preferences”. Economics doesn’t really account for whether an individual consumer is rational for buying junk food, scratch tickets, and liquor; it just sort of observes that you behave as though you want those things and models from there. The internal struggle of will between the “you” who binges on Oreos and store brand vodka and the “you” who wakes up on New Years morning with a hangover and a sincere desire to turn your life around isn’t necessarily within the scope of economics (unless the New Years Day “you” promptly purchases a gym membership that you never show up for!)
Have you seen the variety of bread available, and the price range thereof? If all of it is purchased on rational valuation grounds, I think we must end up with a fairly loose definition of "value"
I'm sorry, but you still seem to be missing the point ? This isn't about bread varieties either, I would assume that most bread is bought within an order of magnitude of prices.
This is incorrect. You agree to trade because you value the food (and eating the food) more than the money. The store agrees to trade because they value the money more than the food. Each side is more satisfied after the trade than before. Hence it's non-zero sum.
If they don't buy it from the store, they have to make it themselves. Everybody are satisfied in the transaction because they played well on this trivial zero sum game instead of falling into a negative sum swirl.
By the way, economic concepts are statistics. Arguments are only meaningful on large number of samples or over long period of time.
That’s an illusion if you can’t use it directly to make more money. In the abstract sure it’s positive value but for this simulation or any simulation it doesn’t matter.
The transaction has impacts beyond the immediate direct interaction, and those must also be considered. Examples:
* The company now has more money for research and development, which will result in the creation of new value.
* Demand has been demonstrably increased (from the moment before the transaction), providing a small but real signal that production should increase. Increased production will result in new value being created.
And so on. This is all on average, of course. Some companies destroy value through incompetence, ignorance, malicious action, or other mechanisms.
No, the value of 3 dollars varies from person to person.
For Bill Gates, 3 dollars is near worthless.
For a poor, 3rd world country citizen living on less than a dollar a day, 3 dollars is very valuable.
Secondly, from your point of view it might be that 3 is 3. But if we look at the entire economy, things get extremely muddy. Maybe those ten dollars that the transaction is based on come from a loan based on a quantifier. So the bank lend out 10 dollars based on a 1 dollar holding. Now the money supply has increased, which all things equal will lower the value of each dollar.
But all things are not equal. The economy is growing and if it is growing faster than the money supply is growing then the value of the dollar is still growing.
The value of those 3 dollar profit is
3/[size of money supply] x [value of economy]
And valuating the economy is extremely difficult. To the point where you could call it subjective.
The price you paid says you value the thing you got more than the money you gave in exchange. At the same time, the person who received the money values it more than the thing he sold. That is what makes economic transactions positive sum games, according to some theories.
I would call market prices more “consensus” than “objective”, since they’re based on the aggregate subjective valuations of those involved in the market.
In the details, even though the simulation shows great inequality, it also shows that everyone gets a chance to be part of the 1% for a bit.
Which is better: everyone is always mediocre, or everyone gets to be super rich for a bit and poor for awhile?
I also created a sum positive version of his sim awhile back, and one interesting outcome is that monopolies became essential for stabilizing the economy in the face of catastrophic recessions. And if we had the zero sum economy where everyone is mediocre, then no one survives catastrophic recessions.
So, capitalism and monopolies seem to be the best overall in terms of survival benefit, and wealth redistribution schemes get wiped out by natural selection. I guess that's what happened to the USSR during the Cold War.
Not in the tech sector. Everyone is competing for monopoly position for each area and niche. Once an area is monopolized, the incumbent gets complacent and inefficient but doesn't lose their winning position.
Also people don't make money by creating value, they make money by picking the winners and following the money.
It can still be a positive sum trade to purchase something from a corrupt and inefficient monopoly. Indeed generally the only reason corrupt and inefficient monopolies can stay in business is because having access to the good/service still has value to its customers, even if they'd be happier having an alternative.
Sure, VCs make a few negative sum trades when they invest in companies that aren't going to generate the exit they were hoping for, but they're aware of that and trying to balance them out with ridiculously positive sum investments anyway...
The 'picking winners' part is most certainly providing value! If you can, more accurately than at random, look at a bunch of ideas and say which ones are going to be successful and which aren't, so that you (or others) can fund effort toward the likely-to-be-successful ideas and not the unsuccessful ones, that is very valuable and deserves to be rewarded.
Indeed, this is the fundamental contribution of capitalism to practical economic performance. It’s simply the most efficient and successful approach to allocating capital to productive enterprises. It’s still vulnerable to abuse of course, as any approach to allocating resources can be, hence the need for regulation.
Regulation is also vulnerable to abuse. The naive “solution” to that is more regulation. This is great for successful incumbents since it raises barriers to entry for potential competitors.
There will always be competition for resources. It’s my observation that every proposal to eliminate that coincidentally allocates more resources to the proposer or a group he identifies with.
Laissez-faire is no exception. In fact I’ve never heard of one.
If you have enough capital and social connections you can pick randomly and still get a return on your investment even if you provided negative value for society. Some monopolistic corporation will end up buying the startup you invested in for a lot of money, shut it down so that the corporation's shareholders will take on the cost but the corporate monopoly means that they won't be affected much anyway. The rest of society ends up paying for that exit through lost opportunity cost because that statup used social manipulation and advertising money to steal customer, supplier and investor attention away from more deserving projects.
Investors win, founders win, everyone else loses. To say that it's a zero sum game is actually a very generous way to put it.
I have heard this argument numerous times, and I'm sick of hearing it over and over again.
Yes, real world economy is (almost) practically a zero-sum game.
Look at GDP growth rate. A rate of 4-6% of a developed country, with population growth of 1-3% is considered a success (after accounting for inflation). Don't tell me tens of millions of individuals can get out of low-income situation, and billionaires can still keep getting rich, with a measley overall growth of 4-6%. The situation is even worse when every few years, it dips to 2%, 0%, or even goes negative.
With a total best-case growth of 4-6%, in a population growing 1-3%, you have to make many folks worse off in order to get ahead.
As others have said, a stagnant (0% growth) economy still has many positive sum trades within it. If all trading stopped happening, you wouldn't see the GDP remain at the same level; it would crash to basically 0 as almost no one would even have access to food.
>With a total growth of 4-6%, you have to make many folks worse off in order to get ahead.
I mean, kinda, in that if you make something really good you make the person that made something slightly worse worse off, as they now don't get the money you earn instead? But this is not what you mean.
> Don't tell me tens of millions of individuals can get out of low-income situation, and billionaires can still keep getting rich, with a measley overall growth of 4-6%.
The way billionaires are rich is by controlling organisations that provide a lot of value and capture some of that value for themselves. One of them getting 10% 'richer' on paper doesn't mean that organisation grabbed some more resources from others; it's not real, current wealth, it's estimated value of total future income from that organisation.
So yes you can totally have those organisations become more valuable by more than 4-6% a year without other people losing out. You're confusing the total amount of value/utility produced in a year (which GDP is more-or-less trying to estimate) with projected total amount of value produced in all future years.
I don't know if you're misreading on purpose or missing a point.
Suppose I each year I produce two eggs and you produce two slices of bacon. We both value these at $1 each. The total value produced this year was $4.
Next year, we produce the same, but we agree to a trade - one egg for one slice of bacon. Egg with bacon is clearly superior to just two eggs or just two slices of bacon for breakfast; so we value the combo of egg+bacon at $3, giving total value of $6.
Next year, we do the same. Total value was just $6, for 0% growth. However, that doesn't mean there was a 'winning side' to the trade. It just means the previous estimate for produced value already included the benefits of the trade, so there's been no change.
When your example is included in Norvig's simulation, it will result in two parties coming together, and instead of an interaction of (-1, +1), the interaction (+1, +1) will happen.
However, when such interactions are randomly mixed with zero-sum trades, and the simulaion is run over N (population) and t (time), it will result in net wealth W_after > W_before.
In order to mimic real-world scenario, the prob distribution has to be chosen such that the W_after is between 0.98 to 1.06 of W_before (in other words, between -2% to 6% growth, more or less). Anything far outside this bound would be highly unrealistic for a functioning economy.
However, when you impose this condition, you would realize that the (+1, +1) events have negligible impact, and that (-1, +1) type of events dominate. In other words, Norvig's conclusion pretty much does not change.
And that is my point. Mixing in a fraction of individual positive-sum interactions doesn't change the results, if the total wealth growth is capped based on real-world macroeconomic data.
>In order to mimic real-world scenario, the prob distribution has to be chosen such that the W_after is between 0.98 to 1.06 of W_before (in other words, between -2% to 6% growth, more or less). Anything far outside this bound would be highly unrealistic for a functioning economy.
No, you're missing the point.
0.98 to 1.06, to borrow your numbers, is the ratio of GDP one year to GDP the previous year. It is not the ratio of GDP accounting for positive-sum trades to GDP under only zero-some trades, and falsely equating those is disingenuous at best. The latter ratio might well be an order of magnitude larger.
Growth over a period of time isn't even strictly necessary for all trades to be positive sum anyway. A village economy where the basket maker trades the same number of baskets for the same amount of food every year may have a zero growth rate [assuming the baskets' lifespan is limited], but those trades are positive sum because the basket maker and the farmers using broken baskets would both be worse off if those trades didn't continue to take place.
Are you telling me that capital powerhouses aiming for 8-20% year over year, and consistently having it their way, in an economy of best case 4-6% with instances of 2%, 0%, and negative mixed in, and a population consistently growing by 1-3%, still leaves the rest of the population with growth?
You’re assuming the 4-6% growth is a given, and were all just fighting over our bit of it, but this is _completely_ backwards.
Where did that 4-6% growth come from? Did it rain down from heaven and were all scrabbling for our share of it? No, it came from positive sum trades where value was added. An enterprise that develops a new technology, that makes an economic activity dramatically more efficient might sell that and become very successful. It might achieve 20% growth in its industrial sector and also boost the growth of its customers due to their better technology or more efficient services.
It’s _this_ sort of activity that much of that 4-6% overall growth is coming from. It’s not draining away growth from others, it’s making the growth in the first place.
Yes. except I suspect once you take negative externalities such as carbon emission into account the true rate is currently more like 1 percent if not actually negative.
I only assumed that to be charitable to OC's argument. If we don't assume that, then OC's criticism of Norvig's simulation assumptions is even more invalid (only slightly so though; it's already invalid enough).
No capital powerhouse achieves 20% year over year. Some capital company maybe gets that each year, but they take losses in different years, so it's just swapping places. Berkshire Hathaway averages what, 9% a year?
An important take-away here is that larger personal economic buffers (the percentage of their wealth each person does not need to transact regularly) resulted in a lower final Gini coefficient. This matches our real-world experience that growing savings is the path to financial stability, and real-world data that forced saving functions (home ownership, social security) improve average financial stability.
I would be interested to see death included in the model, as that is an important and common way that pools of wealth get redistributed over the long term in an economy.
>I would be interested to see death included in the model, as that is an important and common way that pools of wealth get redistributed over the long term in an economy.
Only in the narrowest sense of 'redistributed' though, right? I.e. upon death wealth typically transfers to kin, modulo whatever estate taxes apply for a given country.
Yeah, I mean just the basic inheritance tree—if there are 2 kids then each gets half, and if each of them has kids then it gets further subdivided, plus some heirs will spend down or lose money and have less to pass down...
Slightly off-topic: it was mentioned here earlier that the reason Norvig switched to Python (from Lisp) was because of his job (Google). But in fact he uses Python for most of his personal stuff as well.
As a long time python user who recently got into s-exp languages (scheme to be specific), I realized the differences between python and scheme are more than just syntactic, in the following sense:
Programming in scheme forced me to think and code with a functional mindset. Python never forced me like that, even though you could stick to a functional subset of python (essentially forcing yourself).
I've also realized that a strictly functional mindset doesn't match well with CS curriculum. Case in point, there is a textbook called purely functional data structures, and I'm personally fully convinced that there is a whole class of functional algorithms that I have not explored, and these look nothing like the usual algorithms and data structures taught in a standard CS curriculum.
I also believe that SICP doesn't even scratch the surface of this different kind of thinking. SICP is small booklet of (dynamically typed) functional programming 101.
So if your standard CS graduate doesn't have a bleep of an idea about how to think with a functional mindset, and all of a sudden, has to program in a language which practically forces you to do so, they're going to have a bad time.
P.S.: You could develop an imperative framework on top of scheme (or lisp) and move on with your life, never having to think functionally. But that's a whole different story.
> I've also realized that a strictly functional mindset doesn't match well with CS curriculum. Case in point, there is a textbook called purely functional data structures, and I'm personally fully convinced that there is a whole class of functional algorithms that I have not explored, and these look nothing like the usual algorithms and data structures taught in a standard CS curriculum.
Schools are trying to strike a balance between theory and practice. A strictly functional mindset is, for one, a pretty limited way of viewing computing, but also disconnected from how CPUs actually work.
That depends pretty much on the university's teaching quality.
I was lucky to study in a very good one, so not only did I got exposed to all programming paradigms during those 5 years, I also got access to a rich library that exposed me to the real history of systems programming across several platforms all the way back to the late 50's, due to their rich book and conference proceedings collection.
If you only ever code with a functional mindset, you will probably write really inefficient programs that needlessly re-compute data. To program well on actual CPUs, on some level you need to be able to think about the assembly that your code gets compiled to
And I believe the reason of having the pseudocode Algol-like instead of writing ASTs (i.e. Lisp) is the same - making more accessible to readers. I wonder if he mentions this somewhere.
MMmh ... there's no wealth creation function anywhere. Is Peter assuming that economy is a zero-sum game?
Or is he assuming that the simulation is unaffected by working with normalized numbers?
My gut says both assumptions need to at least be justified. I'd be happier to see an actual wealth creation mechanism added.
Also, the final metric (Gini coefficient, aka the jealousy index), how much does it actually to speak to individual's happiness? A rather controversial choice: happiness is probably much correlated to absolute rather than relative wealth, something Gini entirely fails to capture.
happiness is probably much correlated to absolute rather than relative wealth
Studies show that it's much more complex than that. Basically both are always a factor, but if you're below a certain level absolute wealth is the most important factor, but once a above a certain level relative wealth dominates[1]. The level where the switch over happens differs quite a bit between countries and places, although most studies looking at salary seem to put it in the $75k-$120k/year range for the US. All that being said most studies also find that income/wealth (both absolute and relative) have relatively small impact on overall happiness when compared to other factors.
> Is Peter assuming that economy is a zero-sum game?
Of course production is also part of the economy but that is not what he's modelling. He's modelling the distribution of the wealth created.
Of course you could model positive-sum trades but that just affects how much wealth is distributed after in each transaction, not how.
> A rather controversial choice: happiness is probably much correlated to absolute rather than relative wealth,
Absolutely not. Else our ancestors would all have died of depression in the stone age. In addition wealth has diminishing utility so inequality is intrinsically not efficient.
What makes the increasing-Gini result confusing is ignoring variation. This is done by thinking of the full distribution as if it were its average. In this case, the pot is distributed using a pull from a uniform random distribution: The assets of a pair are combined into a pot, and a number, s, from 0 to 1 is drawn. One person gets s of the pot. The other gets 1 - s. On average, s = .5. When s = .5, the Gini coefficient of the pair after the transaction is zero. Their pair of assets are equal. As ABS(s) approaches 1.0, their inequality becomes larger. For example, in Russian Roulette, s is either 0 or 1. With p = the portion of the pot that came from one person, then, when ABS(s - 0.5) > ABS(p - 0.5), the transaction increases inequality. When s is a constant 0.5, every transaction will either reduce inequality or maintain the perfect equality. The system increases inequality in the population until it matches the inequality in the pool from which s is pulled. There's nothing surprising there unless you make a simplifying assumption that s = 0.5, the average of its distribution. Using this simulation to model economic systems comes down to choosing a distribution for s: whatever you choose, that's what you get. UNIFORM(0.49,0.51) will result in lots of equality. UNIFORM(-1,2) will produce some people in debt to others.
The standing ovation problem is another nice and easily understood problem, that allows a view into dynamics and the impact of structure on the results. One can easily model that as an agent-based simulation and introduce all kinds of stuff on top.
Real life wealth distributions have a very interesting property not shown here. The "random give/take" always ends up with an exponential distribution, which models real life for about the bottom 99% and doesn't have any really high values or long tails. However the top 1% follows a power law in real life, which does spread wealth across several orders of magnitude and has long tails, and the transition is very sharp. Presumably this is because the top 1% made their wealth in a way that is scalable, such as though starting businesses, investing, etc. If you search papers in econophysics you can see this effect.
This is interesting and a fun link between statistical mechanics and observed income and wealth inequality.
But I think the statistics chosen to compare to the real world are a little cherry picked? The wealth distribution matches, but in the simulation, I think the identities of the wealthy change over time. The rich get poor, then rich again, then poor again. Over an infinite time, everyone is in every wealth percentile.
But an important political economic problem in the real world is that the rich stay rich and the poor stay poor. So in that regard the simulation does a poor job of explaining income and wealth inequality.
> But an important political economic problem in the real world is that the rich stay rich and the poor stay poor.
I’m not sure if this 100% true. The highly visible super rich like Bezos and Gates and, say, the top 50 billionaires may stay rich, but the merely rich like millionaires and low billionaires are more likely to have turnover in their ranks. I recall seeing some data on this but forgot where. Will post if I can find it.
I might be wrong, but IMHO the amount of billionaires don't really constitute a big enough population to say something definitive about that claim. Clearly the list of rich people isn't static. A billionaire might be shoved off the list of billionaires and further down, for instance. Though certainly, the chance of a billionaire becomming poor is most likely lower than a millionaire becomming poor, though I'm sure this can be tested. But without a better definition of "rich", then it really doesn't say much about social mobility or economic equality, if that is what you're going after.
If you assume that the claim "the rich becomes richer" is true—which seems to be the case in this simulation—then the thing to find out is probably, "why?" and "what can be done to make a more fair distribution?" Given that this system is even unfair that is. As I see it, this is clearly up for discussion (i.e. whether what is basically Capitalism is unfair), though the simulation does not take it up explicitly, and instead merely hinting to that with such a system, the rich would necessarily become richer, and the poor poorer.
I mean, who said economic equality is even a goal? Do not those who make life better for others, also deserve a good life themselves? And if you don't make life better for others, you obviously still deserve to live, but do you deserve as much comfort and self-determination (in terms of economic wealth) as those who manage to offer more?
Huge economic inequality breeds resentment and a bad society for everyone to live in. Not to mention the risk of a bloody revolution if the poorest people end up with not enough food.
Huge inequality of consumption breeds resentment, yes. Fortunately, in the US, inequality of consumption is much, much, much smaller than inequality of income or wealth.
One of the most economically unequal societies in the West is the USA. Why does Socialism still fare so badly over there, then? Clearly rich Americans should be taxed more, and Americans in general should give up their freedom in order to enjoy more economic equality. ;)
> I mean, who said economic equality is even a goal? Do not those who make life better for others, also deserve a good life themselves? And if you don't make life better for others, you obviously still deserve to live, but do you deserve as much comfort and self-determination (in terms of economic wealth) as those who manage to offer more?
Not sure why you even went there since I said nothing about it. But imho any discussion of economic inequality that doesn’t take into account the different between inequality due to wealth creation and inequality due to wealth extraction is disingenuous and likely a smokescreen.
Inequality due to differences in ability to create wealth is absolutely fine. That’s capitalism at its best, and is the “make life better for others” you’re talking about. Silicon Valley startup ecosystem is the penultimate example, along with various manufacturing activities and energy production.
Raw materials + innovation + capital + labor + time = useful new physical things worth more than the sum of their inputs = wealth creation. I’m all for people getting filthy rich this way, it ultimately benefits all of society and is how we advance the human race (for the most part, minus our failure at dealing with some externalities like pollution).
But inequality due to wealth extraction is a problem, and becoming an increasingly bad one, in the US at least. Wealth extraction is about market failure. One or few entities corner a market, drive out competitors (or collude) and drive up prices.
Examples are the Telecoms and Internet service in the US blocking last mile alternatives, or an increasingly consolidated and powerful banking sector that can privatize profits and socialize losses, or hospital billing administrators enriching themselves by creating obfuscated pricing schedules based not on what some medical service costs to provide, but on what they can get the insurance companies to pay. Things like that.
It’s extremely important that this conversation about inequality constantly makes this distinction, lauds and exonerates wealth creators while closely scrutinizing and course correcting wealth extraction. Otherwise you get revolts like the Bolshevik Revolution, French Revolution, Occupy Wall Street, etc full of people who are just so angry, fed up, and oblivious to the difference, that they will throw out the baby with the bath water if given the chance. Then we all become worse off.
That's amazing. I've always wanted to do something similar but I had a much more complicated simulation in mind. Apparently, you can get some interesting insights even with such a rough approximation.
I would be very interested to see, how the "connectivity" and information would affect transactions. For example, I am more inclined to buy from Apple because Apple is a well-known company and my friends bought from there. The more you are known, the more transactions you can expect to make.
This doesn’t model B2C-type asymmetrical situations, but pure P2P-type symmetrical situations.
An easier first step for the interested beginner would be modelling visibility as a kind of dynamic network. That leads to interesting considerations about imperfect information and agents’ profit-maximisation under such conditions.
Back in 2017, I was wondering about this in the context of the 80-20 rule, and believe I have a simple answer (posted to hacker news way back, https://news.ycombinator.com/item?id=16618385). Basically you need two things: 1) Some slight advantage 2) The network effect, that is, the probability of competing depends on the current winnings.
If you have these two things, you get 80-20 like distributions, you get the explanation for why winners keep winning. If you are interested, you can find my simulation and analysis at
Kind of shocking how well this works. The intuition is, why has coke won, well they had some initial advantage, and so they won a bit. Now that they have won a bit, they can finance themselves into more competition. For example, they can place themselves into more stores, into more restaurants etc. Now they get a chance to compete more.
Running the simulation yields interesting results, for example, in the two columns below, the left is Household income in 1970 broken into quintiles. The right column is simulation results.
Interesting how well the top 3 or 4 quintiles match between the simulation and the real world data.
If you run the simulation with different rules, the real world quintiles do not match the simulation quintiles nearly as well. You can tweak the simulation to see this as well.
The simulation can be tweaked to handle cases such as inheritance, so an actor with different ability inherits the wealth of a past actor.
I modified the simulation as follows:
I allowed it to evolve for a single generation, enough time for the top 20% of the population to have 80% of the wealth. I now choose a random sample from the top 20% of the population to follow, lets call them T20.
I now repeatedly
1) pass the wealth of all actors to actors with new, random abilities
2) let the new actors compete for a generation (the same number of competitions we used above)
Result: After 3 generations of steps 1 and 2 above, 80% of T20 has lost almost all their wealth, 10% has lost 75% of their wealth, 10% has done really well, growing it by a factor of 8, due to capable ancestors for three generations.
I doubt wealth is exchanged 1 on 1 so clearly, there is lot of wealth accumulated in assets that change price for example. Or the commons, a limited wealth up for grabs at the expense of everyone else.
Hypothesis does not rely on wealth exchanged 1 on 1. The idea of modelling is to capture essential features.
This model looks at the case where everything is fair and everybody is equal. In the sense it's the best possible world scenario and it demonstrates that inequality builds up even if no player is better than other.
Another fascinating aspect of simulations like these is how the shape of the final distribution depends on the number of actors - with links to statistical thermodynamics.
Sometimes we needs to start with a simpler model, demonstrate concepts, and work our way up. Of course, it is important to acknowledge the model is simple.
Even if they do know other economic simulations which are better, it isn't a given that a better tool exists for every tool that doesn't ideally meet a need. Of course, you did say "hoped" instead of "expected"...
I think this is more of a gambling simulation than an economic one.
However, even with positive-sum games, the nature of the transaction can have big impact on the wealth inequality. There's a great writeup on that in https://jasoncollins.blog/2020/01/22/ergodicity-economics-a-...