Bitcoin 51% Attack - Protocol solutions?

[Kakmakr]

So, back in the day when Gavin Andresen was still lead developer of Bitcoin (BTC) ...we had the GHash.IO incident, were in July 2014, the GHash.IO mining pool briefly exceeded the 51% threshold. I still remember Gavin Andresen calming people down, saying that there were failsafe action plans in place to stop 51% attacks, if the miners did not want to switch mining Pools or if one entity managed to get more that 51% of the hashing power.

Was this something written in the Bitcoin protocol or was this something that they would implement with a hard fork if this continued? If this was just a smoke screen answer from him to get the miners to switch pools, I want to ask if it would be possible to implement something like this in the code to circumvent this type of attack? (Protocol level) 

[1715213142]

1715213142

[1715213142]

1715213142

[1715213142]

1715213142

[ranochigo]

I don't remember any concrete proposal regarding the mitigation of 51% attacks. I believe what he meant at that time was for Bitcoin to have a hard fork to switch from SHA256D to something else to make all the ASICs worthless. It's not possible for them to insert any viable safeguards, checkpoint was mistaken by some as one but it doesn't really do anything that is beneficial against said attacks.

[pooya87]

there are a bunch of things that could be added at the protocol level to solve the 51% attack issue to some extent and i believe some altcoins have done some stuff in this area although i'm not really following them closely.

for example a simple solution could be adding a simple rule that prevents any block that is n blocks deep to be reversed. for example if the head is at 100 then block 95 and lower would be irreversible even if 100% of mining power was owned by one entity.

changing PoW algorithm is also a solution as it was mentioned but i don't think it can be considered an actual solution, it is more like sweeping the problem under a rug. the problem doesn't go away, it is simply postponed and there is a good chance that it becomes a lot worse because if we change the algorithm the richest miners or the hardware producers could invest a lot of money buying equipment mining the new algorithm and if before we had 51% of hashrate owned by one pool by then we have 80-90% of hashrate of new algorithm owned by one farm.

[Wind_FURY]

Was this something written in the Bitcoin protocol or was this something that they would implement with a hard fork if this continued? If this was just a smoke screen answer from him to get the miners to switch pools, I want to ask if it would be possible to implement something like this in the code to circumvent this type of attack? (Protocol level) 

I believe protection against 51% attacks isn't only solved in a technical level. There's something else holding everything together in the network. There's also the incentives placed to keep miners/mining pools honest.

[ranochigo]

I believe protection against 51% attacks isn't only solved in a technical level. There's something else holding everything together in the network. There's also the incentives placed to keep miners/mining pools honest.

That is correct. Satoshi actually briefly mentioned the fact that a miner would gain more by being honest than if they were to attack the network. I doubt it would be easy to solve it on a technical level without compromising the trustless and decentralised part of Bitcoin. I would love to read up on solutions that could potentially do it without compromising them but I've yet to come across any.

[DaveF]

I believe protection against 51% attacks isn't only solved in a technical level. There's something else holding everything together in the network. There's also the incentives placed to keep miners/mining pools honest.

That is correct. Satoshi actually briefly mentioned the fact that a miner would gain more by being honest than if they were to attack the network. I doubt it would be easy to solve it on a technical level without compromising the trustless and decentralised part of Bitcoin. I would love to read up on solutions that could potentially do it without compromising them but I've yet to come across any.

Yes, if you attack the chain, the price of BTC would drop as people get out if they no longer feel it's secure. Would not matter if it's another coin or to fiat. So from a business point of view there would not be a reason to do it. If you had enough money you could do it, just to do it. But we are talking about a fair amount of money and time. There would be other cheaper ways of causing problems on the chain.

-Dave

[pawanjain]

I believe protection against 51% attacks isn't only solved in a technical level. There's something else holding everything together in the network. There's also the incentives placed to keep miners/mining pools honest.

That is correct. Satoshi actually briefly mentioned the fact that a miner would gain more by being honest than if they were to attack the network. I doubt it would be easy to solve it on a technical level without compromising the trustless and decentralised part of Bitcoin. I would love to read up on solutions that could potentially do it without compromising them but I've yet to come across any.

Yes that's right. I did read that part where Satoshi mentioned that if any party did get to implement the 51% attack successfully and do try to reverse a transaction or chain of transactions  then the cost of reversal process would itself be so heavy that they would rather choose to continue mining in the longest chain rather than initiating the reversal process.
This will force the miners to remain honest. So the point of doing a 51% attack is actually pointless.

[Wind_FURY]

I believe protection against 51% attacks isn't only solved in a technical level. There's something else holding everything together in the network. There's also the incentives placed to keep miners/mining pools honest.

That is correct. Satoshi actually briefly mentioned the fact that a miner would gain more by being honest than if they were to attack the network. I doubt it would be easy to solve it on a technical level without compromising the trustless and decentralised part of Bitcoin. I would love to read up on solutions that could potentially do it without compromising them but I've yet to come across any.

Yes, if you attack the chain, the price of BTC would drop as people get out if they no longer feel it's secure. Would not matter if it's another coin or to fiat. So from a business point of view there would not be a reason to do it. If you had enough money you could do it, just to do it. But we are talking about a fair amount of money and time. There would be other cheaper ways of causing problems on the chain.

-Dave

It wouldn't actually be that simple to do in Bitcoin. The attacker would have to build their farm, buy all the chips available in the market and more, increasing demand and increasing the price of ASICs. They would spend millons more for the electricity, and what for? For a double-spend, which would then be rejected by the community/army of full nodes, and the attacker would fork into a shitcoin if he/they continue.

Who would follow them to their shitcoin?

BUT if he/they were honest, and do their job in the network, they would earn Bitcoins.

There was a window that an entity could have killed Bitcoin through a 51% attack. I believe that window is closed.

[Wind_FURY]

No one has to buy any chips to 51% attack bitcoin, only morons believe this is the only way to coop bitcoin.

All one has to do is gain temporary coordinated control of the top 3 or 4 bitcoin mining pools,
and one can 51% attack bitcoin.
This destroys the perception of bitcoin security and it's price collapse.

Any decent size criminal organization or corrupt government could accomplish this at anytime very easily.
Plus they could make a killing shorting bitcoin to pay for their efforts.

Or

The governments can just keep seizing bitcoins from anyone with a criminal charge, and selling for fiat until they crush the price to nothing.
Realize bitcoin has no true utility that 1000s of other coins don't already do better.

Or

Governments could just pass a law banning stupid non-essential energy waste , so mining farms can't buy electricity,
and thus ending bitcoin with a whimper. 

"Temporary-coordinated-control", and what? For one double-spend, which will be rejected by the community/network? They'll be rewarded WITH BITCOINS if they simply mine.

Plus your other debates has never, and WILL NOT happen.

[PrimeNumber7]

Unfortunately, there is not anything that can be done on a technical level to prevent an entity with sufficient mining resources from executing a 51% attack. It is also not possible to be guaranteed to detect the same entity controlling 51% of mining resources because one entity could own multiple pools for example, or could publically identify themselves as two or more entities.

As others have mentioned, economic incentives will largely prevent anyone from executing this kind of attack. It would be possible that a government, or other entity that might profit from the failure of bitcoin may execute a 51% attack, but I would find this unlikely.

Some altcoins have been 51% attacked, and were vulnerable because they shared algorithms with other altcoins. This means someone could buy a miner, use it to attack a particular blockchain, and subsequently repurpose their miner to mine on another altcoin's blockchain honestly. An altcoin using the same mining algorithm as other altcoins will reduce its security. The reduction in security for i altcoin will be nominal if the resources going into mining on i's blockchain make up the overwhelming majority of resources being used for that algorithm.

[Wind_FURY]

Unfortunately, there is not anything that can be done on a technical level to prevent an entity with sufficient mining resources from executing a 51% attack. It is also not possible to be guaranteed to detect the same entity controlling 51% of mining resources because one entity could own multiple pools for example, or could publically identify themselves as two or more entities.

As others have mentioned, economic incentives will largely prevent anyone from executing this kind of attack. It would be possible that a government, or other entity that might profit from the failure of bitcoin may execute a 51% attack, but I would find this unlikely.

It's not only the incentives that keeps the network together, it's also the cost that adds to the risk, which also adds to the Game Theory aspect.

Would you attack the network, and pay for the cost to destroy the thing that enriches you, or simply be honest and cooperate with the network to enrich you?

Some altcoins have been 51% attacked, and were vulnerable because they shared algorithms with other altcoins. This means someone could buy a miner, use it to attack a particular blockchain, and subsequently repurpose their miner to mine on another altcoin's blockchain honestly. An altcoin using the same mining algorithm as other altcoins will reduce its security. The reduction in security for i altcoin will be nominal if the resources going into mining on i's blockchain make up the overwhelming majority of resources being used for that algorithm.

OR rent the hashing power, attack, pay for the rent, enrich themselves.

I see the troll believes someone will take the risk for ONE double spend in Bitcoin. The window to 51% attack Bitcoin has closed.

[HeRetiK]

Any decent size criminal organization or corrupt government could accomplish this at anytime very easily.
Plus they could make a killing shorting bitcoin to pay for their efforts.

Or

The governments can just keep seizing bitcoins from anyone with a criminal charge, and selling for fiat until they crush the price to nothing.
Realize bitcoin has no true utility that 1000s of other coins don't already do better.

Or

Governments could just pass a law banning stupid non-essential energy waste , so mining farms can't buy electricity,
and thus ending bitcoin with a whimper. 

If it's so easy and profitable why haven't any of those scenarios happened yet though? Especially the first scenario is something where moving first and as soon as possible would be of utmost importance.

[aliashraf]

Unfortunately, there is not anything that can be done on a technical level to prevent an entity with sufficient mining resources from executing a 51% attack.

Not quite true. It has been discussed lately and an ultimate solution has been proposed: put a cap on the depth of chain-reorg attempts.

it is controversial though: implementing finality in PoW is not welcomed by many of the core devs of because theytraditionally believe in a booting from the genesis idea, aking it as a measure of how trustless the system is or at least looks like. Once the blockchain is finalized in some heigth, individuals have no strong incentive to validate the whole history to become convinced about the state of the machine individually, it is frghtening for many people. Obviously, it turns to be more of a philosophical debate about what trust means and questions like 'is it possible ever to get rid of the social aspects of a monetary system?' Believe it or not it is considered as a Satoshi tradition the idealized bitcoiner who does not trust, verifies! A mathematical notion of a timeless, alone entity called individual user who is surrounded by a bunch of thieves and scammers and adversaries and the super hero fights with all of them by verifying the history of the blockchain they propose from day one, the genesis block.

My point: it is not a pure technical hurdle because the suggested cap would fix it easily, rather it is a political/philosophical debate.

[pooya87]

My point: it is not a pure technical hurdle because the suggested cap would fix it easily, rather it is a political/philosophical debate.

does it really fix anything? it doesn't look like it to me and it is not philosophical.
such solutions are targeting X blocks and not the blocks 1 to X-1 from head and those blocks remain vulnerable (assuming 51% attack were possible) which means it really didn't solve anything. which is why i categorize them under band-aids rather than solutions.

[Wind_FURY]

Any decent size criminal organization or corrupt government could accomplish this at anytime very easily.
Plus they could make a killing shorting bitcoin to pay for their efforts.

Or

The governments can just keep seizing bitcoins from anyone with a criminal charge, and selling for fiat until they crush the price to nothing.
Realize bitcoin has no true utility that 1000s of other coins don't already do better.

Or

Governments could just pass a law banning stupid non-essential energy waste , so mining farms can't buy electricity,
and thus ending bitcoin with a whimper.  

If it's so easy and profitable why haven't any of those scenarios happened yet though? Especially the first scenario is something where moving first and as soon as possible would be of utmost importance.

2nd one is currently happening and at the current time the most profitable as it is a rinse/repeat scenario
The governments can just keep seizing bitcoins from anyone with a criminal charge, and selling for fiat

Will "crush the price" truly "kill Bitcoin"? I believe, not. Plus they can "crush the price", it will be a golden opportunity for us. The wiser move would be to HODL.

The 1st or 3rd scenario will only happen, when they are ready to destroy bitcoin or move another coin to the #1 position.

As they have more lasting repercussions and a greater finality for bitcoin.
1st and 3rd scenario become more probable once the tether scam collapses making 2nd scenario less profitable for the governments.

That will NEVER happen.

The troll's debate is not about a 51% attack.

[zbig001]

It has already been done...
Elastos is a PoW (auxPoW, to be clear) coin with validation performed on an additional DPoS layer.
Which ensures finality (very low probability of a fork) because miners can add only signed blocks to the chain, and a possible attacker would have to dominate both layers of consensus.

But Bitcoin is still well secured, there is no need for change here.
It can be assumed that an investment in Bitcoin mining is the equivalent of about two years of profits from honest mining, isn'it?
So whoever undertakes to enter this business accepts a deep economic commitment.

[HeRetiK]

2nd one is currently happening and at the current time the most profitable as it is a rinse/repeat scenario
The governments can just keep seizing bitcoins from anyone with a criminal charge, and selling for fiat

According to this article by the end of 2018 governments all across the globe had confiscated about BTC 453,000 i.e. 2.6% of the global supply at the time.

https://www.theblockcrypto.com/genesis/2944/analysis-the-u-s-has-seized-nearly-200000-bitcoins-to-date-global-confiscations-are-up-to-453000

Looking at CoinMarketCap the last 24h saw a trading volume of about BTC 3,000,000.

So looking at these numbers the global government holdings of Bitcoin are barely enough to cause a market dip for maybe a day or two, so... just another Tuesday when it comes to crypto.

Of course that was 2018 so maybe the governments already increased their stashes since then.

Except they didn't even held onto the aforementioned amount and have maybe a tenth of that left so like 0.2% of the current global supply.

¯\_(ツ)_/¯

The 1st or 3rd scenario will only happen, when they are ready to destroy bitcoin or move another coin to the #1 position.
As they have more lasting repercussions and a greater finality for bitcoin.
1st and 3rd scenario become more probable once the tether scam collapses making 2nd scenario less profitable for the governments.

"They"

Sure.

[PrimeNumber7]

Unfortunately, there is not anything that can be done on a technical level to prevent an entity with sufficient mining resources from executing a 51% attack.

Not quite true. It has been discussed lately and an ultimate solution has been proposed: put a cap on the depth of chain-reorg attempts.

This will not prevent a 51% attack. Someone with >50% of the network hashrate can successfully execute a reorg of a handful of blocks reliably.

[NotATether]

What a rolling checkpoint does is prevent the damage from a 51% attack.
It does not prevent any reorgs before it.

As long as users wait until the rolling checkpoint passes,
then they can be 100% certain , no doublespending will occur.
As even someone with 100% of the hashrate would be unable to do it.

For small purchases of less than $500, most people would just trust whichever algorithm required blocks for confirmation,
but purchase over $2000,  you wait ½ a day for the rolling checkpoint to be certain.

Rolling Checkpoints are what should be in place, and many coins have implemented it,
bitcoin will never implement it as blockstream has no interest in improving the onchain security for bitcoin users.

It may be able to stop a 51% attack but it also introduces a chain split attack, someone can just mine the number of blocks N needed to lock the rolling checkpoint, broadcast it to several nodes as soon as the main chain hits N blocks after the block with transactions they want to mess up, and the network will be split in half with no way to put it back together as a consequence of how rolling checkpoints work.

51% attacks are more of a practical problem to small altcoins (e.g. bcash), than a coin with exahashes/s total hashrate like bitcoin.

[HeRetiK]

It may be able to stop a 51% attack but it also introduces a chain split attack, someone can just mine the number of blocks N needed to lock the rolling checkpoint, broadcast it to several nodes as soon as the main chain hits N blocks after the block with transactions they want to mess up, and the network will be split in half with no way to put it back together as a consequence of how rolling checkpoints work.

51% attacks are more of a practical problem to small altcoins (e.g. bcash), than a coin with exahashes/s total hashrate like bitcoin.

Thank you for this post.

There was something iffy about reorg prevention that I couldn't quite put my finger on but now I know what it was.