Bitcoin & Tragedy of the Commons

[cbeast]

It seems extremely improbable to me that the monopolist could profit from exploiting the ledger. That is regulation enough in my view.

Let me rephrase that.  I would like to see a scheme that regulates a 51% attacker to insure that it does not double-spend.

[DeathAndTaxes]

But even if they are not, this does not automatically mean the death of Bitcoin.  Even if total hashrate is underprovided, it will never fall to zero.  Hash rate is a positive externality of a transaction fee, which some people will always pay for faster processing.

Historically, there are many examples of public goods that are produced in a suboptimal quantity, but they are produced nonetheless.

Just because bitcoin is suboptimal doesn't mean it will fail.   

Look at bittorrent for an analogy. There are a lot of leechers who take more than their fair share and there is a shortage of seeders. But the technology still does its job.

There is a tragedy of the commons you just don't see it.

Nobody is saying hashrate will go to 0.0 MH/s.  Obviously it won't but compensation is what keeps the network strong.  Currently the network costs about $1M annually per TH/s.  If revenue to miners falls then for profit miner hashing power will also fall.  It is possible non-profit and indirect "public benefit" miners may pickup the slack but $10M+ is a lot of slack to pick up.

If hashrate of Bitcoin falls significantly it becomes more vulnerable to a 51% attack and a more tempting target. 

If a single bittorrent seed fails the entire protocol is never at risk.  There is also nothing "lost" except oppertunity.  If seeder/leecher ratio gets out of wack it will fall out of favor among leechers and will reach a new equilibrium.  No "attack" is possible.

Bitcoin will also reach equilibrium but if that equilibrium point is too low and it is successfully attack tens of millions of dollars and millions of hours of work are destroyed.

[cypherdoc]

its not profitable.  all of us are in this thing b/c of the distributed nature of the system.  any move that tries to distort or take advantage of the mining or pool situation would destroy Bitcoin and the value therein.   they would end up destroying themselves along with the huge investments they've made already.

So if mining comes under control of a single entity, everyone will sell out of principle. I don't believe it. Even if they do, it won't be a lasting phenomenon. Confidence will be restored once everyone realizes that the fundamentals haven't changed. At any rate, even if value persistently  crashes to say 33% of its original value, the monopolist would still profit handsomely from the venture.  

i think bittorrent is a useful example.  why don't all the distributed components band together and create an entity called MegaUpload or Napster so they could create a large entity that can force out all the individual nodes and profit off song distribution?

oh, you mean they've tried this already?

[cunicula]

Let me rephrase that.  I would like to see a scheme that regulates a 51% attacker to insure that it does not double-spend.

It seems extremely improbable to me that the monopolist could profit from double-spending. That is regulation enough in my view.

[Steve]

If anyone succeeded in monopolizing mining power, the price of bitcoin will plummet, thus eliminating the very incentive for obtaining that monopoly (unless your incentive is to destroy bitcoin).

If anything, I see mining becoming more decentralized in the near term, not less.  There is simply too much opportunity for people to find electricity that is subsidized in one form or another.  Such subsidies would not be available to a large, centralized miner.  And with dedicated hardware eventually making it as simple as buying a box and plugging it into the wall, chances are that we're going to see an explosion in mining activity (of course, the prospects of such an explosion will have a natural governing effect on the number of people getting into mining).  At $600 & 80w for the butterfly labs box, it's hard not to want to become a miner just for the fun of it.

[cypherdoc]

Other miners who are not trying for monopoly would not be willing to mine at a loss.

there are many reasons why miners mine:

1.  the traditional income/expense analysis.
2.  out of principle to support a belief (even if they lose money)
3.  the hope for price appreciation of their Bitcoins no matter if they are making or losing money based on income/expense.

this creates all sorts of unpredictable human behavior.  a traditional argument around here is that ppl will stop mining if the price drops below cost; lets say $4.  i would argue that some ppl are more apt to mine when the price drops to below 4  b/c they're able to more easily mine coins that have a potential increase in value.

its analogous to a buy the dip mentality.  some ppl like to buy dips and some ppl like to buy at the top.

[cunicula]

If anyone succeeded in monopolizing mining power, the price of bitcoin will plummet,

Why?

[cunicula]

There is simply too much opportunity for people to find electricity that is subsidized in one form or another.  

The subsidized miner will only temporarily need to operate the vast majority of his rigs. Electricity expenses are minimal compared to revenue. I covered this already.

[cunicula]

Other miners who are not trying for monopoly would not be willing to mine at a loss.

there are many reasons why miners mine:

1.  the traditional income/expense analysis.
2.  out of principle to support a belief (even if they lose money)
3.  the hope for price appreciation of their Bitcoins no matter if they are making or losing money based on income/expense.

this creates all sorts of unpredictable human behavior.  a traditional argument around here is that ppl will stop mining if the price drops below cost; lets say $4.  i would argue that some ppl are more apt to mine when the price drops to below zero b/c they're able to more easily mine coins that have a potential increase in value.

its analogous to a buy the dip mentality.  some ppl like to buy dips and some ppl like to buy at the top.

Okay, price fell and difficulty declined. So we have tested your hypothesis already. It has been rejected.

[kjj]

Attractive is not a sufficient condition.  You also need to have a mechanism whereby the monopoly/oligopoly seekers are able to bar entry into the field.

By rejecting any new blocks generated by anyone else.

Ok, that will work after they've achieved majority, as long as they hold majority, but will not help them become the majority.

The other mechanism is simply the competitive market mechanism. The monopolist would be willing to mine at a loss temporarily in order to achieve dominance. Other miners who are not trying for monopoly would not be willing to mine at a loss. This would help clear the field for a successful takeover. That is the actual amount of hashing power you need to achieve 51% is far less than the hashing power currently at work.

Competition does not help in winner take all games like this. The equilibrium is always one winner. Everyone else gives up.

Mine at a loss?  I thought that was what we were all doing already.

[cbeast]

Let me rephrase that.  I would like to see a scheme that regulates a 51% attacker to insure that it does not double-spend.

It seems extremely improbable to me that the monopolist could profit from double-spending. That is regulation enough in my view.

Somehow, I almost think you are serious. Nah.

[cunicula]

I am serious. Completely.

[DeathAndTaxes]

Let me rephrase that.  I would like to see a scheme that regulates a 51% attacker to insure that it does not double-spend.

It seems extremely improbable to me that the monopolist could profit from double-spending. That is regulation enough in my view.

Somehow, I almost think you are serious. Nah.

There is no economic value to a 51% attack.  The cost of running 51% of the network will always be more than the profit one could make via double spending.  The window is also very short as the double spend will cause a loss of value in Bitcoin undermining the attacker's revenue.

Bitcoin faces 51% attack from a NON-ECONOMIC attacker.  An entity looking to damage/destroy and gaining 51% of hashing power is far more plausible attack scenario then someone trying to profit directly from 51% attack.

To gain 10TH/s right now even w/ a capital cost of $1 per MH would be $10 million.  What possible spending could one do to recoup that cost.  The real cost is even higher when you consider electrical transmission cost, building, labor, security (you going to leave $10 mil in a warehouse unguarded), etc.

How long of a window do you imagine an attacker would have to "profit" before Bitcoin falls 99%?  An hour? a day? a week?  Could you buy $20 million worth of goods with Bitcoin right now in an hour? a day?

[markm]

Is there any incentive for the monopolist to reveal even the existence of a monopoly let alone who/where/which people and/or rigs and/or sites constitute the monopoly?

If there is any truth in the suggestion(s) that people would not like knowing there is a monopoly, why tell them?

-MarkM-

[allten]

All the miners and those that hold bitcoins have a vested interest in making Bitcoin successful.
If this really is an issue, I believe it will become apparent to all before hitting a critical point of failure.
In that atmosphere, a general consensus on a good way to go will be reached.....and we'll move on.

There have been so many proposals for different peer-to-peer currencies and p2p currency improvements, but I believe Bitcoin already has all the best ingenuity that any system of this nature could ever have: it's decentralized and changes are made in democratic way. Everyone has their vote with fiat, bitcoins, and/or mining power. There will be issues, and when they arise then the time will be right for a fix and I believe a consensus will be reached rather quickly.

Bitcoin is successful now and I'm very confident in its future.
It may still be a niche currency, but it will still function as it was designed too... as a currency.

[cbeast]

This conversation has become convoluted. The tragedy of the commons argument addresses a drop in hashrate. This in turn leads to a threat that someone will easily take over control of the blockchain. You can't argue that it is a good thing because it would not be worthwhile to double spend due to the mining requirements when you already said the hashrate is low. You can't have it both ways. If a  monopoly takes over the blockchain who would eve know if double spends are made by the cartel? Seems to me this scenario is much like the current bankster system.

[casascius]

(applicable to this thread as well)

The argument that difficulty will tend toward zero in the absence of a block reward is something I've brought up in the past.

One thing that I had not considered at the time was the effect of merged mining, since I was spouting off this line before merged mining ever existed.

I think that to understand this issue, it's significant to consider that if Bitcoin really does end up in a situation where its difficulty tends to zero due to a lack of economic incentive to mine, something somewhere else is likely to not have that problem (inflatacoin? litecoin?).  Someone's always going to be mining something.  Bitcoin can be secured with merged mining the same way Namecoin benefits from Bitcoin mining.

Of course, it would be a huge OUCH to Bitcoin (specifically, the legitimacy of the main chain as we know it) for it to have to take a back seat to some other alt chain just to survive.  But at least it would remain secure.

[ripper234]

(applicable to this thread as well)

The argument that difficulty will tend toward zero in the absence of a block reward is something I've brought up in the past.

One thing that I had not considered at the time was the effect of merged mining, since I was spouting off this line before merged mining ever existed.

I think that to understand this issue, it's significant to consider that if Bitcoin really does end up in a situation where its difficulty tends to zero due to a lack of economic incentive to mine, something somewhere else is likely to not have that problem (inflatacoin? litecoin?).  Someone's always going to be mining something.  Bitcoin can be secured with merged mining the same way Namecoin benefits from Bitcoin mining.

Of course, it would be a huge OUCH to Bitcoin (specifically, the legitimacy of the main chain as we know it) for it to have to take a back seat to some other alt chain just to survive.  But at least it would remain secure.

I can't imagine Bitcoin becoming a second fiddle to an alt chain far into the future ...

I think the process of discovering whether Bitcoin should be the world's virtual currency, or whether some alt chain is better equipped for this, will take a handful of years, not 40. We will discover "the best" crypto currency, and just go with that.

One of the key selling points of Bitcoins, at least these days, is its fixed money supply.
If the best argument we can find about why Bitcoin can survive well into the future is merged mining, then this undermines this argument, and makes investing into Bitcoin right now now viable. (Why would people mine the alt chain ... it will too become dependent upon yet another alt chain later on). This is sort of equivalent to removing the limit of 21,000,000 coins (which might not be such a terrible idea, when you stop to think about it). If this is the case, a Bitcoin2 without this limit can be started right away, and if it's really a better idea than the market (miners + users + shop owners) will flock there ... if it's not needed, it will dwindle like all the other alts.

Nah, I think other methods can be devised to keep Bitcoin alive, perhaps with some protocol modifications as suggested by Mike Hearn.

[cunicula]

I have several reasons for doubting that assurance contracts will be useful. In order of perceived importance:

1) [lack of proven usefulness] assurance contracts are not widely used to fund public good production. If they were effective, we would see them operating more frequently in the wild. Instead, the vast majority of public goods are funded by large, centralized entities such as governments, churches, and corporations.

2) [amount of money necessary] the amount of donations necessary to sustain a hash rate-currency valuation ratio like the current one is extremely large (maybe 10% of all extant bitcoin need to be donated each year). I am extremely skeptical that assurance contracts could muster this type of support.


3) [uncertainty about security need, boy who cried wolf] it will only be apparent to people that security is necessary if a successful attack takes place. People will quickly weary of contributing to some cause which has no proven usefulness. Once you are able to prove to people that the contract is necessary, it will already be too late. After an attack has occurred, assurance contracts cannot be enforced.

4) [entrepreneurship] establishing assurance contracts is costly for the initiator. Some private individual has to gather information about blockchain security and decide when a contract is necessary. If he initiates a contract when it is not necessary, then he has wasted his money. The incentives for atomistic individuals to perform this entrepreneurial function are extremely weak. They are much more likely to default to not doing anything.

5) [dynamic waiting problem] Under an  assurance contract, there is a marginal contributor whose contribution puts the contract into effect. If you are considering making a contribution, it is always better to wait to the last minute to try to avoid being the marginal contributor. If the contract is fulfilled before the last minute, then you will benefit from withholding your contribution. Thus the contract is unlikely to motivate many contributions until very nearits deadline. Even though incentives to contribute improve at the very end, I doubt that everyone will be able to coordinate at the last minute to fulfill the contract. Last minute coordination is difficult.


On another note: Ripper, you asked about proof-of-stake. I provide some more details about my ideas for proof-of-stake in this thread: https://bitcointalk.org/index.php?topic=55184.20
There are ideological objections (read FUD) to proof-of-stake, but I am not aware of any logical flaws. I think the ideological concerns would fall away in the face of a proof-of-concept. After all, most people have ideological objections to bitcoin when they first hear about it.

[ripper234]

I like this discussion.

I have several reasons for doubting that assurance contracts will be useful. In order of perceived importance:

1) [lack of proven usefulness] assurance contracts are not widely used to fund public good production. If they were effective, we would see them operating more frequently in the wild. Instead, the vast majority of public goods are funded by large, centralized entities such as governments, churches, and corporations.

It is possible that the lack of real world examples of assurance contracts is because of the overhead and inefficiency. With Bitcoin, they can be very efficient and simple to use, and this might make the difference. Lack of existing prior cases is not a disproof of the method's effectiveness.

2) [amount of money necessary] the amount of donations necessary to sustain a hash rate-currency valuation ratio like the current one is extremely large (maybe 10% of all extant bitcoin need to be donated each year). I am extremely skeptical that assurance contracts could muster this type of support.

You speak with high confidence ... I just don't know yet. If Bitcoin's valuation becomes significantly larger than it is today, then a relatively minor portion of it will suffice to incentive miners.

3) [uncertainty about security need, boy who cried wolf] it will only be apparent to people that security is necessary if a successful attack takes place. People will quickly weary of contributing to some cause which has no proven usefulness. Once you are able to prove to people that the contract is necessary, it will already be too late. After an attack has occurred, assurance contracts cannot be enforced.

Large Bitcoin stakeholders are likely to be well informed of its mechanics. Remember, this is a good few years into the future ... this is ample time for plenty of detailed economics studies to be conducted, and for the "correct" security requirements to be very well known. Institutionl investors will understand the value of security ... the knowledge will "seep in" and ample warning will be given when hash rate approaches critical levels. The key thing here is that (I believe) BTC holders will really understand this aspect, and will known it's in their best interest to fund security.

4) [entrepreneurship] establishing assurance contracts is costly for the initiator. Some private individual has to gather information about blockchain security and decide when a contract is necessary. If he initiates a contract when it is not necessary, then he has wasted his money. The incentives for atomistic individuals to perform this entrepreneurial function are extremely weak. They are much more likely to default to not doing anything.

There are many people with a large stake in Bitcoin. Those people will devote the ample time and resource to initiate assurance contracts and educate others about them. I believe we'll actually see a strong "non-profit" Bitcoin Foundation form pretty soon.

5) [dynamic waiting problem] Under an  assurance contract, there is a marginal contributor whose contribution puts the contract into effect. If you are considering making a contribution, it is always better to wait to the last minute to try to avoid being the marginal contributor. If the contract is fulfilled before the last minute, then you will benefit from withholding your contribution. Thus the contract is unlikely to motivate many contributions until very nearits deadline. Even though incentives to contribute improve at the very end, I doubt that everyone will be able to coordinate at the last minute to fulfill the contract. Last minute coordination is difficult.

This is an iterated game (compare it to iterated prisoners dilemma vs single round). In an iterated game of assurance contracts, people will understand it's really in their best interest to get a mutual buy in, and will not necessarily do the short term selfish thing of withholding from investing. Also, we could blur the line with "vague contracts" - some amount of BTC and a deadline are chosen in advance, but the exact numbers are not published beforeahand, but rather only some estimates. "I donate 10 BTC, provided X amount of BTC, between 1000 and 1500 is gathered up by block Y, between 2,000,000 and 2,150,000". This vagueness might reduce the possibility of knowing whether you are the marginal contributor, and might solve this problem completely. This is like in prisoner's dilemma, if the number of rounds is known in advance, the rational analysis suggests you should betray at round 1, while if the exact number of rounds is not known, this is not the case. Also, as Mike and others have commented, game theory is just a model of real people's behavior, and is often very very wrong, as numerous studies have shown.

On another note: Ripper, you asked about proof-of-stake. I provide some more details about my ideas for proof-of-stake in this thread: https://bitcointalk.org/index.php?topic=55184.20
There are ideological objections (read FUD) to proof-of-stake, but I am not aware of any logical flaws. I think the ideological concerns would fall away in the face of a proof-of-concept. After all, most people have ideological objections to bitcoin when they first hear about it.

Thanks, I'll take a look at that thread.