The seller puts up 1x price of item as a deposit and buyer puts up 1x for deposit + 1x for payment. These can only be released when both agree on payment or refund. This keeps incentives aligned without the need for escrow agents or reputation systems.
That's a good scenario to consider.If my motivation is to lock up funds and destroy trust in the marketplace, and if I did not care about resources, this system might not work.
I'm curious if there is some obvious reason that others see that makes this attack vector unlikely, or impossible to pull off?
True, you could use reputation systems. But I don't think that fixes the issue. Let's assume that these reputation systems are 100% perfect (ie the reputation rating cannot be gamed).One solution is that sellers can choose to reuse an identity and people that have purchased from them could favorite them and avoid attacks by buying from sellers they've used before.
This is a pretty interesting scenario. I'm curious if you think the scenarios you mentioned are any *worse* than a 3-party escrow sytsem, where a government party, or well funded adversary, could lock up funds, or even harm your reputation by ensuring that you 'lose' on all your contracts as the 3rd party arbiter.For a 'dark' marketplace, where a government with alterior motives would act 'irrationally' from the market's perspective, I am not sure.
Imagine this scenario:
You and I are making a trade. I am the "seller", you are the buyer. I am, in fact, a government agent intent on doing reputation damage to the system and I have no product and no intention to ship. You put up 2 BTC, and I put up 1 BTC. Then, I have a 2x multiplier, because you trusted me to have the product and to ship it.
If my motivation is to lock up funds and destroy trust in the marketplace, and if I did not care about resources, this system might not work.
I am not attempting to compare the two systems side-by-side. I am brain-dumping my first thoughts on the subjectThis is a pretty interesting scenario. I'm curious if you think the scenarios you mentioned are any *worse* than a 3-party escrow sytsem, where a government party, or well funded adversary, could lock up funds, or even harm your reputation by ensuring that you 'lose' on all your contracts as the 3rd party arbiter.
But you are of course correct that the solutions offered by this system should be weighed against the competition and not considered in isolation.No. Because you would not re-use identities. As such, all new identities are 'unreliable' and 'risky'. But that is solved by letting the market decide the required escrow ratios. In a normal market, these ratios would be different than a market that is 'under attack'.Additionally, I'd say that in the 2-party, or 3-party escrow system, if such an adversary were attempting to do this, wouldn't they quickly get a bad reputation.
Exactly. And that is a barrier to entry to the market for legitimate users with no reputation. Barriers to entry are an obstacle to free markets.Most escrow-based markets seem to have a notion of reputation, even if they remain anonymous, there is some ID which can accumulate reputation. Of course an adversary could create lots of accounts and fraud people from different accounts, but then other 'legitimate' users would become ever more wary of accounts/Identities with no reputation at all, and the value of good reputation would go up even more, so as to ensure that almost no one does business with those adversaries over time. Wouldn't this be possible to play out in the 2-party or 3-party escrow system?
Thoughts?
Yeah, that is interesting. Even with no abuse, there is *some* barrier to entry. If you have 0 reputation, you are always going to have to sacrifice something else to begin the process of building up your reputation, to attract people to give you a chance. So maybe to build reputation, you have to take less escrow (than say the 50%, or whatever), giving the buyer an advantage/incentive to using you. I think the market would find an equillibrium there that works, over just freezing in place with only the existing reputation-holders being able to facilitate transactions in the system with no competition.Exactly. And that is a barrier to entry to the market for legitimate users with no reputation. Barriers to entry are an obstacle to free markets.
Slick software!I'd like to hear more people's thoughts on the two party escrow system described here: https://voluntary.net/bitmarkets/
It seems very interesting, but there is obviously a lot of game theory involved. Do people think this kind of system can work?
Why, or why not? Is it superior to a standard 3 party escrow?
The seller puts up 1x price of item as a deposit and buyer puts up 1x for deposit + 1x for payment. These can only be released when both agree on payment or refund. This keeps incentives aligned without the need for escrow agents or reputation systems.
No "reputation system" was suggested outside of buying from a seller you've already done business with. A government could start with proxy attacks on sellers but now it gets into mass selling of illegal items (assuming they are attacking sales of items banned in their territories e.g. banned books in a police state), and this leads us to the question of how likely any of this is.True, you could use reputation systems. But I don't think that fixes the issue. Let's assume that these reputation systems are 100% perfect (ie the reputation rating cannot be gamed).One solution is that sellers can choose to reuse an identity and people that have purchased from them could favorite them and avoid attacks by buying from sellers they've used before.
Thanks. We're working on a Javascript port. It won't run in a browser because the whole idea is that this is a decentralized app and not something connecting to centralized web servers, but the port will allow us to have a single code base that runs on OSX, Windows and Linux..Slick software!
Is it not possible to make this web-based?
No, they would not need to sell anything. They would only need to pretend to sell something. They only need to offer it for sale, and wait for someone to buy. This happens all the time in "real life", where undercover agents posing as dealers make a deal - and when the customer confirms the deal to the point that a judge would accept intent, they arrest the guy. They never actually sell anything. This attack vector is no different.No "reputation system" was suggested outside of buying from a seller you've already done business with. A government could start with proxy attacks on sellers but now it gets into mass selling of illegal items (assuming they are attacking sales of items banned in their territories e.g. banned books in a police state), and this leads us to the question of how likely any of this is.
I think this premise is probably accurate. But look at how the government went after the Silk Road. If they consider this a treat, it might not take a billion dollars per year in sales to put it on their to do lost.A given marketplace wouldn't be on anyone's radar until it is big (say, $1B/year in sales). At that point, an attack might require hundreds of millions of dollars. Can you imagine the dialog of a government agency meeting deciding if it should spend it's budget in this way and who will be the responsible party?
This is a completely different argument. I fail to see the connection.The same questions were raised about governments doing 51% attacks on the Bitcoin network.
Two questions:I prefer 3rd party escrow. There is scammers on sellers and buyers side. Sometimes real persons view as 3rd party is needed
In which case users sticking to sellers they've already used avoids the attack. I suggested the proxy attack as a way of a government avoiding this defense.No, they would not need to sell anything. They would only need to pretend to sell something. They only need to offer it for sale, and wait for someone to buy. This happens all the time in "real life", where undercover agents posing as dealers make a deal - and when the customer confirms the deal to the point that a judge would accept intent, they arrest the guy. They never actually sell anything. This attack vector is no different.
I didn't make that claim, only that it's long been suggested that a government would stage a 51% on the Bitcoin network, yet it doesn't appear to have happened so far. Can we assume you've chosen to avoid using bitcoin because of the possibility of such an attack or have you come to think of it as an unlikely scenario and if so, for what reason if not the one's I've mentioned?I also do not share your optimism that a government would never do something that is illegal for regular citizens to do.
Yes! They would. And this is the core of the problem.In which case users sticking to sellers they've already used avoids the attack.
You can be never sure.Two questions:I prefer 3rd party escrow. There is scammers on sellers and buyers side. Sometimes real persons view as 3rd party is needed
1) How does a 3rd party fairly resolve a dispute if they only have one party's word against another?
2) In a pseudonymous reputation system, how can you be sure that the agent and other party are not the same entity?
Return to “Bitcoin Discussion”
Users browsing this forum: No registered users and 6 guests
