Profile

What is a 51 percent attack?

What is a 51 percent attack?

A fifty-one percent attack is when a majority of the the miners on the Bitcoin network (as measured by hashing contributions) collude to distort the state of the ...

Transcription

What's up, party people? Chris DeRose here, Community Director of the Counterparty Foundation. On today's video, I wanted to address the question: what is a 51% attack? So a 51% attack is a very much heralded problem or a threat scenario for Bitcoin and something definitely that needs to be understood. Right off the bat, it's not actually a 51% attack per se. There is a point at which the network is overcome by bad actors, and that point is at 50% plus one. But more so than that, we should probably decide what a 51% attack actually is and what it enables people to do. So in Bitcoin, we rely on consensus from the network in order to arrive at truth, in order to perform our settlement and the consensus opinion is typically in some level a vote, or at least some level of representation from stakeholders.

In the Bitcoin network, those stakeholders are minors. When we talk about what a 51% attack is, we talk about a scenario by which of the mining pool, greater than 50% of the people have an intent of perturbing the network, which can take many forms. But certainly the most important of all of those is the hashing [inaudible 00:01:18]. So we talk about what the percentage is. It's not necessarily the number of people, but it is the number of hashing or the number of energy or the amount of energy that they are burning in order to arrive at consensus. For more details on how that mining works, you can see some of my other videos.

But in the scenario that there is a bad actor or there are bad actors that want to pervert the state of the network, they can concentrate their efforts as bad minors. They can take on this percentage of the pool and they can start doing things to the Bitcoin network. Now, it's important to recognize what they can and cannot do. What they can't do is alter transaction. They cannot adjust people's balances. They cannot write bad transactions and they cannot misrepresent transactions.

What they can do is determine which transactions to include or not include. So they can perform a denial of service, if you will, on the network. They can prevent commerce from happening. They can also, if they have a substantial amount of power, roll back the network. They could start to prevent new blocks from being added and actually reverse the prior confirmations, the prior blocks that are already in place by overpowering the legitimate minors with their bad mining. When they do this, in this Bitcoin 51% attack, there are damages to the network.

It would certainly destroy some confidence in the network. It would destroy some faith, but there are a couple of other more subtle issues at work. First and foremost, there is a scale of 51% attack, if you will. There's a vector. It starts at even say 30% or so that you start to see that maybe bad minors can start to affect the pool in this way. If a 30% holder of the network wants to decide what transactions are in their block or not, they can after a time start to perform a minor denial of service on the network.

And at 51%, that doesn't mean that they can totally outnumber the network. There will be a diminished representation from the other minors that are in the network. So they don't perform a full and outright perversion of the network per se. What they do start to do is cause problems and they will hit a critical mass as time goes on. If 51% is accrued, then they could start to perform a more threatening denial of service. But I think that it's really important too that people recognize that in this 51% scenario, it doesn't destroy the Bitcoin network.

It doesn't prevent Bitcoin from working per se. What will happen is alarms will go off everywhere at the time at which we start to see this. Let's say some large bad actor, let's say China decides it doesn't like Bitcoin for some reason and it decides to devote a whole bunch of energy into perturbing the network. Alarms will go off everywhere, people would stop doing transactions. We would see that there's this representation. There are a lot of options we would have at that point but what's certain is that all of the values that are in the network would be preserved.

We would know that a block height of 330,000, that's when it started and for a day thereafter, the network got a little weird and we had some problems and there's a lot of things we can do. Andreas Antonopoulos talks about this. Theoretically maybe we'd change the mining algorithm if we had to. Theoretically we could blacklist those IPs if we had to. But I don't see that being a major problem. The cost to do that is enormous and it begs the question, "What do they really achieve?" If they spend $50 million or something like that, I don't know what it would take.

But if they spend this huge amount of money, for doing what? They wouldn't again alter the balances. They probably couldn't do a double spend. Doing a double spend will require that the network not be aware of this massive entrance of funds, which it would be. And all they could do is kind of stop the network for a little bit. It would be bad for the confidence the other people have in Bitcoin. People would say like, "Oh, bitcoin is failing," and all this other stuff but probably it wouldn't be that big a deal in the scheme of things.

What's important to note is that no balances would be destroyed in terms of your holdings. No transactions could be forged and really all that you can do is kind of be a major pain in the ass and make people hate you. That's about it. It's not a big deal. That's what a 51% attack is. Did I answer your questions? Do you have more? Go ahead and leave comments in the video below.

Tell me what you're looking to hear. I'd love to answer those. My Twitter handle is derosetech and you can write to me at chris@chrisderose.com. Subscribe to the channel if you like this video and want more questions answered. Peace out, party people.