The Oligarch Game: use coin-tosses to demonstrate "winner take all" and its power to warp perceptions

Originally published at:


over the duration of the very short games, the winners arrive at a “feeling of righteous empowerment based on being successful” and players experience class divisions.

That’s definitely the most interesting part, given that the game mechanic is a textbook illustration of dumb luck. If you’ve ever wondered how someone can fall for grifts like “prosperity gospel”, the path of inquiry leading from this game is a good one to follow.


Take this game, add in the ability of the ruthless or the politically connected to weight the coin flip in their favour, and you have an almost perfect description 1990s Russia, and how the wealth of an entire superpower was stolen in a few years.


I’ve always found it very odd that “luck” isn’t more often brought up as a pretty significant reason people succeed or fail – at various aspects of life, but certainly including economic opportunity. And that’s not even counting the dumb luck of who you happened to pop out of, and what their own circumstances are, which obviously you had nothing to do with when you emerged.


It is, just not by the people trying to preserve the fallacy that it isn’t.


True. But I even mean in economics classes and such. I don’t remember “dumb luck” coming up in my intro to political economy class, although it was a while ago, now… :slight_smile:


Look, the winner of the game is blessed by god and has superior coin flipping abilities. Really a super person like that needs to be admired and with a bit of gumption one can learn to have amazing coin flipping ability that society can rally behind. Such amazing abilities need to be cheered and held up as brilliant examples of the best people our country can produce! Not like those horrible people that couldn’t flip a coin property and just look to leech off of society.

It’s only logistical.


The people on the bottom are just too lazy to pull themselves up by their bootstraps. /s


At neighborhood gatherings we sometimes play a game called Left-Right-Center which has very similar outcomes.

1 Like

Cute. So, once you’re behind, you get stuck on a runaway exponential track that looks like 2^(t*(log(1.5)+log(0.5))/(2log(2))) ~ 2^(-t0.21). (‘t’ is the number of turns) I suspect one sees a lot of similar death spiral dynamics in the business world…


Economic growth is by definition not zero sum, and yet this game is; as such, the correct conclusion to draw is that economies must grow to survive.

1 Like

However, economies also don’t have people stop playing when they hit zero - they can, and do, go negative. Often very negative, so in many ways reality is even worse. Even growing economies don’t do enough to counteract this ability to go negative.


GDP is a measure of money being spent (i.e. a flow). So GDP growth is a measure of acceleration. The economy grows when money flows faster through the system after each transaction. If the rules of the game have a baked in “winner takes all” effect, all that happens when you accelerate it, is you get to the end state quicker.


Thank you, Cory (for the post), and Neal (for the automation of the game).

Cool things about the Scientific American article: it explains the US wealth inequality (extreme and unstable) and European (stable-ish) based on a simple model that matches results within a 1/3 of 1%. We now know what to do to bring back stability and I would say productivity to the US: slight redistribution.

And if we do not then we get civil war, as explained in this mind-blowing book about Debt Forgiveness in the near east bronze age: … And forgive them their debts : lending, foreclosure and redemption from Bronze Age finance to the Jubilee Year by Michael Hudson.


Oh sure, GDP growth is an imperfect way to estimate economic growth (“an increase in the amount of goods and services produced per head of the population over a period of time”).

True, but irrelevant. However you operationalize it, by that definition, you’re still measuring the derivative of a flow, i.e. an acceleration.


I’d love to see the experiments repeated under growth assumptions, or a proof that growth doesn’t matter.

Are you sure superior coin flipping can be learned?

If I give my offspring $10,000 we can test if they also inherited my superior coin flipping abilities.


Stanford Prison experiment, anyone?