11

Is it 51% attack or 50.0000...1% attack?

Bitcoin is getting to a point where 1% is a lot of hashing power...

My understanding is >50% due to the nature of the protocol.

user664833
  • 103
  • 2
nehz
  • 213
  • 1
  • 6
  • 2
    Actually it's possible attempt an attack with less. You could be successful in punctual double-spends with less than 45% of the total hashing power, but it requires a bit (if not a lot) of luck. – Stéphane Gimenez May 15 '13 at 07:02
  • 1
    50-51% is the point where it becomes essentially trivial to have continued [control](http://bitcoin.stackexchange.com/questions/658/what-can-an-attacker-with-51-of-hash-power-do) over block creation. – Stéphane Gimenez May 15 '13 at 07:11
  • this actually makes a lot of sense. you should post this as an answer! – nehz May 15 '13 at 07:31

1 Answers1

8

When performing a 51% attack, you will need to find new blocks faster than the remainder of the network combined. This way, your block chain will become the longest and clients will consider it the main chain.

Theoretically, this is achieved (on long term) when you have >50% of the hash power.

But as we know, a lot of luck is involved in mining. It's possible that someone with 55% or even 60% of the total hash power will only succeed in his 51% attack after many weeks because he has bad luck. (Note that with >50%, statistically you will always succeed on long term.) But it is also possible that someone with less than 50% can succeed on short term.

Steven Roose
  • 11,473
  • 8
  • 44
  • 71
  • 6
    In most contexts referring to majority rule, 50% + 1 of votes is generally the right way to describe the majority. 51% is shorter and less confusing though. – Stephen Gornick May 15 '13 at 12:11