P2POOL and the 51% attack.

[spartacusrex]

Hi,

I understand technically how p2pool operates, and that it is HOP proof and DOS proof.

BUT - can someone explain, technically, why it mitigates the 51% attack.. ?

Is it the fact that since the sharechain used is so much faster, you would need to have 99% of the Hashrate to overrun the sharechain ?

So instead of a 51% attack - you need 99% for a successful attack ?

[etotheipi]

Hi,

I understand technically how p2pool operates, and that it is HOP proof and DOS proof.

BUT - can someone explain, technically, why it mitigates the 51% attack.. ?

Is it the fact that since the sharechain used is so much faster, you would need to have 99% of the Hashrate to overrun the sharechain ?

So instead of a 51% attack - you need 99% for a successful attack ?

The 51% attack is only if one person/entity can decide which transactions to into the blockchain.  This is a concern for big pools where there's a single pool operator making the decision about which tx are accepted, and the miners are just "drones" that follow along.  If that pool operator modifies his software to execute a double-spend, then he gets the mass of miners in the pool to come along with him by default (though, if it was proven he did this, many of those miners would jump ship and mine at a different pool). 

But with P2Pool, each individual miner gets to decide for him/herself which transactions go into the chain, as if each miner was mining solo.  So you get the decentralization of a bunch of independent miners, but they all still get to share the reward of the blocks. 

[spartacusrex]

I see.

SO - If we imagine that p2pool becomes the de-facto pool. let's say 90% or even 99% of the hashrate comes from p2pool. What happens if ONE miner (entity or a private pool of miners) controls 51% of the hashrate of p2pool ?

Can't he still do his attack then ?

I mean the ROLL-Back attack, where he can create a new valid chain, and catchup then overtake the current chain (making his the longest chain), and remove the spends he wants.. ?

[etotheipi]

I see.

SO - If we imagine that p2pool becomes the de-facto pool. let's say 90% or even 99% of the hashrate comes from p2pool. What happens if ONE miner (entity or a private pool of miners) controls 51% of the hashrate of p2pool ?

Can't he still do his attack then ?

I mean the ROLL-Back attack, where he can create a new valid chain, and catchup then overtake the current chain (making his the longest chain), and remove the spends he wants.. ?

The effect of a mining pool is to consolidate all the hashing power under the pool operator's control, in terms of what transactions and blocks are to be included and considered valid.  Miners would prefer to make those decisions themselves, but if you control only 0.0001% of mining power, you'd usually prefer the lower variance of getting a small payout every 30 blocks (by the pool) vs 1 full block every 3 years (mining solo).

P2Pool provides the same variance reduction for miners, without centralizing the decision making process.  Someone who has 51% of all hashing power can do what you are speaking of, regardless of whether they are mining solo or in P2Pool.  If they mine in a regular pool, they are effectively "handing the keys" to the pool operator.  But with 51%+ you have no incentive to use a regular pool or P2Pool.