The Hash Wars: Easy Victory for ABC or just a trap?

in #hashwar6 years ago

Yesterday the chains have split and we now have a ABC and a SV chain. Both chains are at this point incompatible with each other and a public reorg is no longer possible. The fork went very smooth without any complications.
For live information I recommend this site source

So story settled and time to move on? Not yet.

images (5).jpeg
source

Before the fork the SV team claimed to have a massive advantage in hash power and that the miners should decide. They wanted to fight for Nakamoto Consensus and up to this point failed horribly. Will they give up and declare SV dead? I highly doubt it.
Instead I think they were simply bluffing and have some different plan for the war.

Miners vs Community

The Miners make the blocks and that gives them a lot of power, but there are a lot of things they cannot do. For example setting the price. SV is not valued that much by the community so there is really no reason to blow massive hash power into the chain. Even when they would mine twice as many blocks, that would change exactly nothing but their rhetoric. They just have a massively overmined chain where the miners loose money.

So I find it not surprising that giving the lower price of SV the hash rate is lower. That is what a rational actor would do. In the end the miners have to follow the market and their hashing cannot drive up the price.

But miners are not powerless

While miners cannot make the price go up, they certainly can let the price go down. There are some possible attacks that all rely on having 51% hash power. Below that you are just an annoyance but not a threat.
The most devastating one is a secret chain reorg. In that attack dishonest miners try to mine a second chain in secret and publish that one if they feel they have a big enough lead in PoW. This chain is not the official consensus chain and all transactions that were in the old chain could be replayed or double spend. Some will not be able to be replayed when people are quick to move their funds and all of the commerce done on that chain will be left hanging in the air. This is an absolutely disastrous attack that will sink the price.

The big problem with such an attack is that as a shadow miner you need to commit to the attack and cannot exit without taking a total loss. If you misestimate your hash power and end up a few percent short, you waste all your effort and have to eventually give up with nothing to show for your spent electricity. It is a very high risk strategy.

Less risky is a censorship attack. In that case you do not reorg the chain but start attacking at the current head mining only empty blocks and orphaning all other blocks. This attack is much less disruptive and it is immediately public. Everyone knows but you can stop at any time and have caused some disruption. No double spends though, just freezing the network for some time.

In both attacks the attacker will loose money as honest mining is always more profitable. But the bad miners can discourage honest miners and disrupt the chain and that might be worth their loss.

The plan for attack

One of the reasons why SV is far behind mining blocks might be that they are preparing such a shadow reorg. This is certainly possible and I would remain very careful moving ABC tokens right now.

But there is a second much more likely explanation. They are just bluffing this attack and secretly mine BTC with their remaining hash power. As I stated above there is really no reason to sink a lot of hash power into a quite worthless SV chain, so they do something else. And because they have announced all kind of attacks the smartest plan is to not attack and let ABC do stupid thinks trying to protect from the attack that is not actually happening. Instead you just relax and make money.
The real attack might then come in a few month when nobody expects it.

Defending the attack

In the end I think such an attack is not good for either party. Because when SV successfully reorgs ABC, what do they gain. Do they think the people that lost their money will then go and buy SV? Certainly not, they will get BTC which is much more resilient due to the much higher cost of starting a hash attack. Or maybe a POS coin? Everyone may spent a lot of money attacking a chain, but it is not clear to me how they would benefit.

ABC centralisation

But just the threat of such an attack is a good weapon, that costs zero and has an impact. Out of fear of a reorg, ABC has introduced centralised checkpoints, and that may weaken their coin. If SV is just mining bitcoin secretly I would say they played it well so far.

ASIC vs Non-ASIC

This situation leads to a very interesting question. What would happen it the case of a deep reorg and what is the proper way to deal with it.

Quite certainly the attacked coin would fork trying to save the old consensus chain. This is not the official way to deal with it, but it also happed in ethereum and I expect no different reaction here. The threat of such a fork alone is discouraging shadow attacks.

41J-1vxyo+L._AC_SY400_.jpeg
source

But looking at the tech part:
-Non-ASIC algorithms have that advantage that everyone can join in and extra hash power may be found to revert the reorg without a fork. It is much harder for the attacker to estimate how much power they need and where they can realistically start their reorg.
-In ASIC coins the miner has to commit to that hash algorithm by investing in specialised hardware. If a reorg happens one may switch the mining algorithm and financially burn the attacker much more, but also the own miners.

At this point I have no idea what is more secure, but both have advantages.

There is another option, introduce a floating 15 block or so checkpoint. Any reorgs that go deeper than that will be rejected by consensus. Of course this brings its own problems that the nodes need to timestamp and that is not decentralised. So the chain may fall out of consensus. In practice this is probably only a little concern as such deep reorgs never happen under normal circumstances.

SV

Let me conclude by commenting on the name SV. If they are actually attacking other coins, their name is a joke. The vision of Satoshi was to create a public network that is naturally resistant to bad actors. If they manage to sink ABC, then they have proven that Satoshis Vision failed and we need to improve on it. They affectively destroy their own legitimacy when they succeed. That sounds like a bad business plan, but for the rest of us it is certainly interesting to learn what happens in this war and find solutions should some other actor eventually plan to attack BTC in the future.
I am hoping for a proper hash war and a secret reorg to learn from it, even though I think I is just a bluff to make ABC do stupid thigs (which they do) and SV was happy from the beginning to fork into two functioning coins.

Sort:  

As BCH and BTC shares hashing algorithm, an attack can easily be pulled off by someone who is currently spending significant mining power on BTC. Even long before there was any talks of a BCH fork I saw speculations that this was "bound to happen".

The total mining revenue will go down if the value of BCH goes down. Though, for someone owning a largely unprofitable legacy mining farm, there is nothing at stake, and it may be profitable spending their unprofitbale legacy-gear on doing a 51%-BCH-attack by first shorting the coin.

With the BCH fork, such an attack is much easier.

Yes you could short the coin and then attack it. But there is always something at stake. Because you need to pay for the electricity even when already having the equipment.

The ability to short the space adds another dimension to the game theory, but for every short there is a long that has an interest to prevent such an attack by adding hash power to the network. In a symmetric case the net impact would be zero.

However, the situation is not symmetric. The attacker has the advantage to choose the moment of attack while the defender has to stay alert all the time. At the same time, the attacker is taking a significant risk for complete failure when wrongly estimating hash rate. Plus the defender can always make a small consensus fork simply rejecting the competing chain by checkpointing. Its not pretty but would have huge support in the community. How much that would then trash the price is quite unclear because we have never seen it.

This is why I am hoping for some action.

the attacker is taking a significant risk for complete failure when wrongly estimating hash rate.

I'm not sure if the risk is that great ... the hash power stays relatively constant, so one can guess pretty well how much hash power is needed to launch an attack. One does the attack covertly, mining an "alternative" branch without publishing it - while doing so one will know pretty well how much ahead of the "real" branch one is. When the secret attack chain is maybe some 20 blocks ahead of the fork point and the known chain is 15 blocks ahead, one can release and watch the havoc. Only way to defend against such using pure hashing power would be to produce some 6 more blocks on the original chain real fast. Check-points ... or, using some other way than hashing power to define which branch is the "real" branch would help, but that's sort of a social solution, I guess there are lots of "code is law"-purists out there that would balk at the idea. I guess the probability that the community will accept the attacking chain depends on how deep the reorg is (number of blocks since the fork point) and how much money will be double spent.

I agree, but I am quite optimistic to defend this attack (e.g. buying a short and then shadow mining) for the two following reasons (and because we have these defences attackers are discouraged to even try):

  • the shadow miners that have a stake to attack the network spend computing power to remine x old blocks and publish when they are y blocks ahead. In order to do this they have to use much more computational power than just mining x+y blocks because the honest miners dont sleep. But once they publish, all honest miners (contested mining rewards), the stakeholders and the long positions have an interest to remine the original chain. And to do this they do not need to remine x old blocks, they only need to get y blocks. After the shadow miners have already committed to remining x+y blocks, it becomes a race where they have only y blocks advantage over the honest miners. In addition their stake is always smaller than the stake of the honest miners since every short is also a long plus the marketcap and mining rewards (difference in stake of x blocks). So it seems that the shadow miners are at a double disadvantage and are set up to fail and it is only a matter of time until the honest chain is restored.
  • This brings me to the second point. Since it is clear that the shadow chain will almost certainly end up orphaned, some consensus nodes will immediately reject it by checkpointing. This may be done via a fork, where no sane investor would give any real value to the shadow chain, or as a 'temporary' service, to let people see what happens on the 'real' chain and continue to do transactions there in the meanwhile. In both cases, the value would be with the original chain. Even the code is law people may use this, since they know that the shadow chain will be orphaned and there is not need to wait for it if it will happen anyway.

Then if I am correct, the attack will not end up doing much. But what we need is seeing this happing to discuss the strategies and solutions. This is why I am still hoping ABC will be attacked.

Also note that a fork introducing a checkpoint to dismiss a shadow chain is still decentralised. Nobody is forced to use this side and people are free to use the coin supported by the shadow miners. But why would they, what is the value proposal of the shadow chain. And if the shadow miners do not also put in their own checkpoint to prevent a future reorg, it will be overtaken anyways.

It comes down to saying "Okay you mined a lot of blocks that are compatible with our consensus rule, but you miners are there to provide a service to us and clearly you did not do that. So thanks for mining all these block, but we dont need them. And then moving on with life while leaving the attackers ruined. "

This post is promoted by @reversed-bidbot as an upvote lottery with a jackpot of 0.300 steem!

There are still up to 1500 tickets waiting to be distributed.

Tickets are distributed according to the value of your upvotes with a multiplier of 1500.0.
Upvote the post to participate and claim your tickets!

Excluded from participation are: bidbots, autovoters.

The jackpot is paid in redfish_shares by @curatorbot, giving you lifetime upvotes while supporting the redfish community. Find more information here.

Follow me to also earn steem on my other promoted posts; or use @reversed-bidbot to hold your own raffles. For more details please visit my discord server or read my introduction post

Hi @frdem3dot0!

Your post was upvoted by @steem-ua, new Steem dApp, using UserAuthority for algorithmic post curation!
Your UA account score is currently 3.431 which ranks you at #6794 across all Steem accounts.
Your rank has dropped 5 places in the last three days (old rank 6789).

In our last Algorithmic Curation Round, consisting of 276 contributions, your post is ranked at #179.

Evaluation of your UA score:
  • You're on the right track, try to gather more followers.
  • The readers like your work!
  • Try to improve on your user engagement! The more interesting interaction in the comments of your post, the better!

Feel free to join our @steem-ua Discord server

I was called to perform a provably fair random drawing for @reversed-bidbot!

There are a 687 tickets participating in this round.
@inversionistam holds tickets number 0 - 7
@curatorbot holds tickets number 8 - 20
@tobixen holds tickets number 21 - 521
@freebornsociety holds tickets number 522 - 536
@sternblitz holds tickets number 537 - 537
@johnvibes holds tickets number 538 - 540
@smartcoins holds tickets number 541 - 557
@frdem3dot0 holds tickets number 558 - 575
@definethedollar holds tickets number 576 - 664
@darkflame holds tickets number 665 - 678
@reversed-bidbot holds tickets number 679 - 686
The merkle root of the block in which this post will appear determines the winner.

The block in which the above post is included has a merkle root of 9dc092b56f97d59cc3d9e88e296ceb57b6c9b808.
To find the winner we:

  1. Convert the merkle root into base 10:
    9dc092b56f97d59cc3d9e88e296ceb57b6c9b808 = 900606074230647811564867103607495766542132033544.
  2. Compute the residual after dividing by the number of tickets: 900606074230647811564867103607495766542132033544 mod 687 = 184.
  3. This is the winning ticket, held by @tobixen.

Congratulations!

Coin Marketplace

STEEM 0.19
TRX 0.15
JST 0.029
BTC 63498.69
ETH 2645.91
USDT 1.00
SBD 2.80