51% attacks - something important people aren't realising

[spooderman]

They have the potential disrupt bitcoin in a large way.

But a pool like GHash.io is demonstrating that it is not doing what that would require. (It may be, but anyone else also may be).

A 51% attack - the type worth actually being afraid of - will logically only be done by a silent entity, apparently solving no blocks, until suddenly it broadcasts a large amount of solved blocks on a chain longer than the current chain, that goes back many blocks, leaving many transactions potentially unconfirmed, or many coins double spent. Anyone can make an alt chain at any time, but having more power than the rest of the network (meaning having over 50% of your network and the rest of the network combined) means you have the ability to make a longer chain - one that will be automatically accepted by the rest of the network.

If a person/organisation has 51% of the total power of the network (which would be greater than double the entire known hashrate of the network due to this needing to be done in secret) they can mine their own chain on which they can conduct whatever mischief they like.

However, if they are solving blocks, and broadcasting them to the network, they are starting from the beginning of their "altchain" each time they broadcast a block - and will have to start a new fork. If they hide all their blocks from the network until they have say 100 blocks, and due to their 51% of the total (secret and known) hashing power (at least 230,000Th) eventually they will have an alt chain of a greater block size than the main chain (say, 99 blocks) then all the nodes will switch over to the longer chain.

I've read over this and it's slightly confusingly put. I will explain again, hopefully better here:

If Ghash.io wants to fork us, and it can given how much power it has, it would have to use all its hashing power to make a silent, parallel chain to the blockchain which it keeps silent for as many blocks as possible, until it has a few more blocks than the main chain. If this chain has not confirmed any transactions since the beginning of the fork, or had double spends within it, it would be very disruptive. However, given that Ghash is braodcasting blocks to the network, it cannot be doing this because it is USING its hashing power to solve blocks that the bitcoin community can see, and has no issues with.

The only risk is that they start being naughty. At which point the community will react, and a solution will be found. This is unlikely, as many have pointed out because it would destroy their business.

A true bitcoin destroying entity does not do what Ghash is doing. It buys up enough hashing power to equal, and then supersede the entire network, then makes a silent, longer blockchain with lots of unconfirmed transactions, most likely transactions that they had made that were included in the shorter chain that they then make invalid, allowing them to scam people as a secondary motive, but their primary motive of course, would be to destroy bitcoin completely by making all transactions in the last x blocks (since the fork) unconfirmed, and by continuing to not allow their confirmation.

Unfortunately, the above scenario is possible, but the chances of a mining pool doing it is just as likely as anyone else doing it, and a large amount of legitimately solved blocks, even an amount greater than 51% is not an attack, it's just a lot of solved blocks. Nothing is being attacked here.

Unless Ghash is only using say 20% of it's hashrate to mine the blocks we see, and is slyly building a powerful 200Petahash rig with which it makes a huge chain that wreaks havoc on the whole community, it is doing nothing wrong, it is just solving lots of blocks.

And to be honest, it is LESS likely to be doing this than anyone else, because if it was, it would be fighting against a network whose main power comes from itself.  

Reasonable conclusion:

People who are attempting 51% attacks are not contributing to the honest part of the network, because if they were, they would be making their own goal harder.

The more legitimate blocks you create, the less likely you are to be attempting a 51% attack (the dangerous kind).

They may be attempting to make a "stupid" 51% attack, where they *shock* double spend a couple of transactions, we all notice and their whole pool comes crashing down.

Be afraid of 51% attacks. They could destroy us but they won't be done by a mining pool we all know about that submits honest blocks to the network.

[1714846749]

1714846749

[1714846749]

1714846749

[spooderman]

edits: made bits clearer, spelling mistakes etc.

[joshraban76]

To be honest, you post made me confused, people say 51% attack will not affect us and you say it will.

I don't know who to trust !!

[spooderman]

Please read again, I've made it much more clear.

Anyone can create an altchain, where they don't confirm anyone's transactions, and they double spend the bejezzus out of the coins they own. And everyone will completely ignore this chain because it will always be shorter than the MAIN blockchain, UNLESS they have more than the rest of the network combined. AT WHICH POINT they can ruin, or at least freeze everything. However to do this, it is unlikely that they will be using 100petahash of power on the LEGIT chain like Ghash.io would have to be doing if it was making a longer, alt chain full of double spends and unconfirmed transactions.

[cyberpinoy]

wow your explanation eased my stress and raised them at the same time.

 Someone should not let any big bankers get this info, because if they do I promise you JP Morgan and Barclays will begin their attack.

[spooderman]

It gets harder every day for someone to do. We have an awe inspiringly powerful network. It sucks that so much of it is concentrated in one pool because it looks bad, but we can and will know if they do something funny, meaning they would have to be literally crazy, way beyond MtGox levels of crazy, to actually use their "51%" to attack anything.

At the moment they are making it twice as hard for someone else to do a 51% attack.

Thank you?

(For the record, I mine at eligius).

[SirChiko]

It gets harder every day for someone to do. We have an awe inspiringly powerful network. It sucks that so much of it is concentrated in one pool because it looks bad, but we can and will know if they do something funny, meaning they would have to be literally crazy, way beyond MtGox levels of crazy, to actually use their "51%" to attack anything.

At the moment they are making it twice as hard for someone else to do a 51% attack.

Thank you?

(For the record, I mine at eligius).

Actually yep the ghash.io hashrate  stopped rising 24h chart shows only 33%

[kingscrown]

heres official statement from today from ghash io - http://fuk.io/cexio-ghashio-jeffrey-smith-interview-exclusive/

[RodeoX]

A 51% attack would only benefit an entity who wants to hurt bitcoin.  Mining pools have a greater interest in the health of bitcoin than others and it is not logical for them to use their hash power to destroy the price of bitcoin and render all their expensive equipment useless.

A malicious entity who just wants to make BTC go away could organize a 51% attack. But they would first have to spend millions of dollars buying hardware. Then they would have to compete with legitimate miners until they have most of the hashing power. Finally, they could fork the chain and create chaos. Of course this would not make crypto-currency go away, they would then have to repeat this process indefinitely. Each time spending millions of dollars buying mining hardware then throwing it away.

I see no easy path for a 51% attack to be successful and profitable for anyone. I think the risk is sometimes overstated. 

[spooderman]

A 51% attack would only benefit an entity who wants to hurt bitcoin.  Mining pools have a greater interest in the health of bitcoin than others and it is not logical for them to use their hash power to destroy the price of bitcoin and render all their expensive equipment useless.

A malicious entity who just wants to make BTC go away could organize a 51% attack. But they would first have to spend millions of dollars buying hardware. Then they would have to compete with legitimate miners until they have most of the hashing power. Finally, they could fork the chain and create chaos. Of course this would not make crypto-currency go away, they would then have to repeat this process indefinitely. Each time spending millions of dollars buying mining hardware then throwing it away.

I see no easy path for a 51% attack to be successful and profitable for anyone. I think the risk is sometimes overstated. 

It is the biggest threat we face. I agree with you about how likely it is, but people love worrying about things, and this is the most logical thing to worry about!

You made an excellent point when you said that they would have to repeat this process again with every other crypto-currency.

[silversurfer1958]

saw an interview with Max Keiser and a dark wallet dev, they were saying that it would only cost a Govt a few hundred million to launch a 51% attack on Bitcoin.

[RodeoX]

A 51% attack would only benefit an entity who wants to hurt bitcoin.  Mining pools have a greater interest in the health of bitcoin than others and it is not logical for them to use their hash power to destroy the price of bitcoin and render all their expensive equipment useless.

A malicious entity who just wants to make BTC go away could organize a 51% attack. But they would first have to spend millions of dollars buying hardware. Then they would have to compete with legitimate miners until they have most of the hashing power. Finally, they could fork the chain and create chaos. Of course this would not make crypto-currency go away, they would then have to repeat this process indefinitely. Each time spending millions of dollars buying mining hardware then throwing it away.

I see no easy path for a 51% attack to be successful and profitable for anyone. I think the risk is sometimes overstated. 

It is the biggest threat we face. I agree with you about how likely it is, but people love worrying about things, and this is the most logical thing to worry about!

You made an excellent point when you said that they would have to repeat this process again with every other crypto-currency.

You also make a good point. It is still a big threat, despite the factors I mentioned.

[philipma1957]

heres official statement from today from ghash io - http://fuk.io/cexio-ghashio-jeffrey-smith-interview-exclusive/

\\


I read this  it is interesting.  I find the idea they want to create ghash for scrypt very interesting.

I wonder it that drives ltc up in price or back down.

BTW  the biggest threat they have is deep pockets  they can hire the shit/gear/software guy from China that ran gear with bad software that reported hash but never did a block…  Once they hire that guy they can point him at a smaller pool and kill the luck of that pool.

   Just think 400 coins were lost at btcguild and 200 coins were lost at eligus pool.  So they can point 200 or 300 th of gear at any pool that is a bit smaller say Bitminter    run it for a month and cause really bad results.

  This harm the smaller pool as luck drops to 80% for a month in a row.People leave the pool and join cex.io   just study the tactics of

Walmart in smaller towns.   This is a real danger and it makes money for them and does not quite hurt BTC like a 51% attack would.

I have been ranting about this asking pools to allow for a solo mining connection at their pools.  
Since the  solo mining can not be hurt with this attack. any pool with really bad luck 80% for a month may be getting attacked as I speak.  Frankly only cex.io has the deep pockets to do these attacks on purpose

 [if they are happening]  Only btcguild and eligus claim this happened and they do not say it was cex.io they also do not say it was done on purpose.

  

[SecureErase]

They have the potential disrupt bitcoin in a large way.

But a pool like GHash.io is demonstrating that it is not doing what that would require. (It may be, but anyone else also may be).

A 51% attack - the type worth actually being afraid of - will logically only be done by a silent entity, apparently solving no blocks, until suddenly it broadcasts a large amount of solved blocks on a chain longer than the current chain, that goes back many blocks, leaving many transactions potentially unconfirmed, or many coins double spent. Anyone can make an alt chain at any time, but having more power than the rest of the network (meaning having over 50% of your network and the rest of the network combined) means you have the ability to make a longer chain - one that will be automatically accepted by the rest of the network.

If a person/organisation has 51% of the total power of the network (which would be greater than double the entire known hashrate of the network due to this needing to be done in secret) they can mine their own chain on which they can conduct whatever mischief they like.

However, if they are solving blocks, and broadcasting them to the network, they are starting from the beginning of their "altchain" each time they broadcast a block - and will have to start a new fork. If they hide all their blocks from the network until they have say 100 blocks, and due to their 51% of the total (secret and known) hashing power (at least 230,000Th) eventually they will have an alt chain of a greater block size than the main chain (say, 99 blocks) then all the nodes will switch over to the longer chain.

I've read over this and it's slightly confusingly put. I will explain again, hopefully better here:

If Ghash.io wants to fork us, and it can given how much power it has, it would have to use all its hashing power to make a silent, parallel chain to the blockchain which it keeps silent for as many blocks as possible, until it has a few more blocks than the main chain. If this chain has not confirmed any transactions since the beginning of the fork, or had double spends within it, it would be very disruptive. However, given that Ghash is braodcasting blocks to the network, it cannot be doing this because it is USING its hashing power to solve blocks that the bitcoin community can see, and has no issues with.

The only risk is that they start being naughty. At which point the community will react, and a solution will be found. This is unlikely, as many have pointed out because it would destroy their business.

A true bitcoin destroying entity does not do what Ghash is doing. It buys up enough hashing power to equal, and then supersede the entire network, then makes a silent, longer blockchain with lots of unconfirmed transactions, most likely transactions that they had made that were included in the shorter chain that they then make invalid, allowing them to scam people as a secondary motive, but their primary motive of course, would be to destroy bitcoin completely by making all transactions in the last x blocks (since the fork) unconfirmed, and by continuing to not allow their confirmation.

Unfortunately, the above scenario is possible, but the chances of a mining pool doing it is just as likely as anyone else doing it, and a large amount of legitimately solved blocks, even an amount greater than 51% is not an attack, it's just a lot of solved blocks. Nothing is being attacked here.

Unless Ghash is only using say 20% of it's hashrate to mine the blocks we see, and is slyly building a powerful 200Petahash rig with which it makes a huge chain that wreaks havoc on the whole community, it is doing nothing wrong, it is just solving lots of blocks.

And to be honest, it is LESS likely to be doing this than anyone else, because if it was, it would be fighting against a network whose main power comes from itself.  

Reasonable conclusion:

People who are attempting 51% attacks are not contributing to the honest part of the network, because if they were, they would be making their own goal harder.

The more legitimate blocks you create, the less likely you are to be attempting a 51% attack (the dangerous kind).

They may be attempting to make a "stupid" 51% attack, where they *shock* double spend a couple of transactions, we all notice and their whole pool comes crashing down.

Be afraid of 51% attacks. They could destroy us but they won't be done by a mining pool we all know about that submits honest blocks to the network.

Thanks Spooderman, your explanation is a bit of sanity in the recent hysteria.