What Is Cryptocurrency Game Theory: A Basic Introduction
blockgeeks.comSorry all, but this site is banned on Hacker News for having used a spam service to buy fake upvotes.
Normally we err on the side of forgiveness, but this abuse is so outrageous that until we get a proper accounting for it, we're going to keep posting what happened at the top of the threads. I'm pretty sure there's no issue on which HN users feel more strongly, and I agree with them.
Users vouched for the current post, which is fine, but I'm going to write software to prevent that from working on sites that are banned for such a reason.
Ooh, he was doing so well but his explanation for why users use the main chain instead of a forked chain betrays a fundamental misunderstanding of the blockchain.
There is no "blue chain" or "red chain", there are just blocks that point to other blocks. Branches happen all the time, and sometimes those branches become the main chain. So actually, if the red chain was grown faster than the blue chain, it would become "the main chain". But then the miner's transaction in the blue chain would be ignored!
The main chain is chain with the greatest computational value, it's the chain that takes the greatest amount of CPU power to compromise (and therefore it's the most trustable). It has nothing to do with Schelling point or bounded rationality, because that suggests you can tell that a block is a "forked block" the moment it's created, but you can't.
"It has nothing to do with Schelling point or bounded rationality, because that suggests you can tell that a block is a "forked block" the moment it's created, but you can't." - if you mine an invalid block you can tell that it will not be part of the main chain. In this sense mining a valid block is a Schelling point - because it is what you expect others expect.
yes but if you mine a valid block around the same time someone else mines a random block you can't know until another block is mined (and even then, you might not know)
'yes but'? What you are saying is true - but it does not contradict my point in any way. The choice is between making an valid or invalid block - and the Schelling point is making a valid block. Another case is if you see a fork with a shorter and longer branch - then adding to the longer branch is also a Schelling point.
Not really...you're not avoiding the invalid block because it's "familiar" to use valid blocks. You're avoiding the invalid block because anything built on it is also invalid so all of those transactions are ignored.
Those transactions are ignored because nobody will attach to that branch - and they don't attach to it because it is a Schelling point to attach to a valid branch.
It's not a Schelling point to attach to a valid branch if the software itself rejects invalid blocks.
Then it is a Schelling point to use the standard software.
No, they don't just use the software because it "feels familiar". Everyone is acting in their best interests, nobody is going to sit in a sub-optimal position just because they can't be bothered. If you use non-standard software, then everyone else will reject your proposed blocks.
The only way you can use the concept of Schelling point is to explain why so many people use Bitcoin rather than Litecoin or another slightly-tweaked variant. Much like the example in the article, if you ask anyone to pick a cryptocurrency to pay with, they are likely to say Bitcoin because it is familiar.
When people choose to meet at Central Station it is also in their interest - this is the same mechanism. It is a case where it is beneficial for people to choose the same thing - and people choose something because they believe others will choose it as well.
Yes, that makes sense when you're talking about why a person would choose to transact with Bitcoin. It's not "invalid" to choose Litecoin, it's just more familiar and easier to use Bitcoin. Schelling point is about social familiarity.
The reason people avoid invalid blocks is because they would have to convince every other node out there to accept it. It's a coordination issue.
> because that suggests you can tell that a block is a "forked block" the moment it's created, but you can't.
Invalid blocks are invalid, that's one way to tell. Run the bitcoin protocol rules and you'll get the same results as everyone else.
You often get two valid blocks at the same time, and whichever one gets the most blocks built on top of it wins.
He's not talking about a hard fork, where blocks from the new fork are invalid on the old fork. He's just talking about two miners finding a new block at approximately the same time. They're both valid, but only one can become part of the 'main' chain, the other will eventually be lost as nobody continues to mine on top of it.
True, but I'm referring to valid blocks.
There is a great class of Game Theory for free on Youtube. It is from YaleCourses with the professor Ben Polak.
http://oyc.yale.edu/economics/econ-159/lecture-1
https://www.youtube.com/watch?v=nM3rTU927io&list=PL6EF60E102...
Probably my favorite Yale Course to date!
I was really glad to see this article specifically point out that proof-of-work schemes that allow for "specialized ASICs" to be designed to accelerate them--which in the high-level concept of proof-of-work is usually seen as a problem (as it creates an elite class of invested and thereby slightly more centralized miners instead of a diverse population of decentralized users)--has a game theory benefit that is related to proof-of-stake (as it incentivizes people to think past short term gains that can be made by cheating the system, leading to a form of loyalty to the one currency and even a fear of escalating tactics); this is something I (and I am sure many others) had noticed while analyzing the Bitcoin Cash debacle, but it has never really been well stated.
All proof-of-work algorithms can be accelerated by specific hardware of some kind or another. Bitcoin is perhaps the extreme example, with ASIC miners. But coins using other algorithms are mined by using top-end graphics cards, which is still a form of specialized hardware. There's no algorithm that can't improve upon a standard computer.
Other forms of proof-of-work have been mooted, e.g. algorithms that require lots of memory, or lots of storage, but again, specialized equipment will accelerate this too, it's just that the coins aren't worth enough for people to invest in designing the hardware so far.
A top-end graphics card has other uses and other value, and is something a large number of people (myself included) already own; and most importantly: they can always be used to mine a different coin, which entirely removes them from contention as "specialized" as used in this article (as the entire argument was based around how the hardware would not be useful for a different coin).
As for "all things can be accelerated", the question here is more "how unique is the thing being accelerated?" (and so I would agree the wording in my comment was slightly off), as this argument about the grim trigger holds up if and only if the specialized hardware being used is not valuable for other uses (including other coins).
As an example of what you can conclude with this thought process: if you are a small "startup coin" you should seriously considering a proof-of-work scheme that is weirdly unique in that it forces people to not be able to reuse hardware they have from another (particularly if more valuable) coin.
True enough, the graphics cards certainly have alternate uses while a bitcoin miner becomes a very heavy paperweight once the difficulty increases and it turns unprofitable to run.
Your thought process leads to another interesting game theory puzzle: There's a benefit in creating a coin that has a unique proof of work, but as you say, there's also value in letting miners easily switch their existing rigs to mine your coin. I wonder which is actually the best choice?
SIA is making the bet that somewhat ASIC friendly algorithms are better off.
The argument being that because ASICs are worthless for anything but mining that one coin, you can really accurately gauge and maintain a hashrate across the whole network. It would be crazy expensive to hoard a bunch of ASICs for a specific algorithm but not run them in hopes to launch a secret 51% attack.
With GPUs that's not really an option. Because GPUs can be used for so much more, it would be easy to see where you could suddenly have a flood of GPU power hit your chain which could be massively over 51% quite easily.
And they could be gone again just as quickly, leaving the remaining users with an impossibly high difficulty and killing off the chain (without a hard-fork to fix it).
It's a fascinating read which actually flipped my opinions on ASICs in favor of them now, I highly recommend it [0].
[0] https://blog.sia.tech/choosing-asics-for-sia-b318505b5b51
As an example of what you can conclude with this thought process: if you are a small "startup coin" you should seriously considering a proof-of-work scheme that is weirdly unique in that it forces people to not be able to reuse hardware they have from another (particularly if more valuable) coin.
I imagine there is much danger here of thinking oneself smarter than any attackers.
I'm glad to see this. With many systems trying to move to proof-of-stake[0] algos to secure their chains, I've been lamenting the lack of game theory experts in a lot of these projects.
Do you think Satoshi Nakamoto is a "game theory expert"?
Which one of them?
I'm not sure I get your point?
In the crypto world I feel like there is a LOT of talk about theoretical concepts, but nowhere near enough actual implementation. This is really obvious with the ICO craze lately.
I would rather we see some more code written and less talking about game theory and other concepts.
Good article though.
What if a country, like Russia, decided it wanted to destroy bitcoin because of a threat to its currency? They have the resources to build up huge mining power and attack the chain for majority, and cause havoc. Is there any kind of protection for this scenario? Would all the other miners just decide to ignore these new jerky miners?
The quickest solution in such a case would be a POW change that can put Russian miners at a significant disadvantage. Significant could be a factor of 10 or larger.
Also, any attempt to kill one decentralized value system would prove to be a wasteful process in the long run since the cost to create another such system is disproportionately smaller than the cost to attack/destroy them.
> The quickest solution in such a case would be a POW change that can put Russian miners at a significant disadvantage.
As in changing the POW algorithm? That puts everyone at the same disadvantage - the attacker just has to keep outspending the legitimate miners, who will also have lost their mining power. If the miners make new ASICs, the attacker can do so as well, at presumably the same or lesser cost.
> the cost to create another such system is disproportionately smaller than the cost to attack/destroy them.
I think it's the opposite. The defenders have to keep mining power up all the time, and may have trouble coordinating on one system - the attacker only has to spend their power while they are attacking, and they can wait until any system gets popular until they mount the attack.
And if the attacker shows they're motivated and capable of outspending the defenders, they don't have to actually attack. The defenders will see that any attempt at mining is futile and never try. At least in game theory, in practice things aren't resolved that easily. :)
This doesn't work when the attacker is a nation state, b/c the PoW change damages all miners equally but the nation state is not playing by the same economic rules, is politically motivated rather than financially, does not need to remain profitable to continue attacking, and is better able to absorb the loss and continue the attack with new hardware.
> The quickest solution in such a case would be a POW change
That may be the quickest solution, but do you think it would be anywhere near quick enough?
Control stratification is a legitimate threat to bitcoin, but I think your fears of threats to a national currency are overblown. There are significant advantages to floating currencies, and there isn't much a cryptocurrency can do to attack it. Russia has magnitudes more control over the Ruble's future than any competing currency ever could.
Good overview and I agree with most of it, but I think the power of the grim trigger is overrated here. In the monarchy metaphor, the threat of defection is death. In the crypto example, it's crypto collapsing. As long as the defectors can cash out to fiat before the collapse they are better for it.
This is roughly the dynamic that seems to have played out with BCH. Every exchange would be better off (according to the grim trigger argument) if Bitcoin never forked, but individually they can profit from being a place to exchange BCH. The fewer exchanges "defect", the more profitable it is to defect. So BCH is now #4 by market cap.
That section is talking about miners, not exchanges; and it specifically notes the limitations and assumptions that even make that possible: in this case, that the Bitcoin proof of work is subject to acceleration using ASICs. Even then, that was a response to a specific attack: using >50% of the mining power to do a takeover of the blockchain, which was itself made more possible (as otherwise it would fizzle) using the takeover contract specified by Vitalik. It was an interesting analysis of a specific scenario that the model seems to provide protect against.
In the very different case of Bitcoin Cash, what you saw were miners perceiving segwit as a protocol fork which would devalue the future potential of their investment in specialized mining equipment, as it changes the proof-of-work parts of Bitcoin into a form of contract verification for payment channels rather than as the one true way a transaction can be performed, so we would have expected them to revolt to maintain the rules they had invested assuming (and in fact once they were already dealing with a fork, they went ahead and made their own rule changes to benefit them: larger blocks).
Fair point, the example of BCH may not have been a good one. I stand by my more general point that grim trigger isn't a strong incentive in crypto because a rational, selfish player can profit from pulling the trigger.
I think a better grim trigger analogy is the market cap of Ethereum surpassing Bitcoin, it proves the digital scarcity of Bitcoin doesn't matter and that opens the possibility of many kings/protocols to be dominant
> So BCH is now #4 by market cap.
BCH is still so thinly traded that this is a meaningless number.
It's 4th by volume also.
There is a mistake in the wording of a sentence (the table is fine):
"If they both confess, then the payoff matrix says that the outcome is (4,4)."
Is actually talking about the scenario where neither Rob nor Ben confess.
This is not "Cryptocurrency Game Theory", it is some game theory applied to some cryptocurrency problems.
What else could cryptocurrency game theory be?
The prisoner's dilemma example is wrong. It should be globally optimal for both not to confess, but knowing the other player can get a better sentence by confessing, each player will also confess in order to at least get the somewhat reduced sentence.
Game solved: the only winning move is not to play.
I never realized just how useful game theory could be to cryptocurrency. Strategy is really important when evaluating the security of a particular crypto, or trying to optimally invest.
There is a new field called cryptoeconomics that studies the application of game theory to cryptocurrencies. Here's an interesting talk about it: https://youtu.be/pKqdjaH1dRo
its entirely game theory because the only way to secure these distributed systems is if the participants all benefit most by enforcing the code and not deviating
All protocols of the world were thought with "game theory" in mind. They just didn't call it by that fancy name.
In a way but I believe it is less direct in other examples. TCP was not built in a way with game theory in mind to give reward to the participants. I believe BTC will thrive because the users and miners are incentivized to strengthen and enforce the system instead of attack it, this was an important and purposeful part of the design.
So true. You have to think of every little action a potential malicious attacker could use and find a way to counter it. It's like a chess game, but with much higher stakes.