Hard Fork Risks and Why 95% Should be the Standard.


There has never been a hard fork of Bitcoin, as most would define it, where the entire network needs to coordinate a protocol update at the same time.  The closest Bitcoin came was in 2013, due to an unexpected software error and it caused significant disruption, and in 2012 where the protocol made a backwards incompatible change but it only affected clients over 2 years old which as far as developers could tell were not in use.

However a contentious hard fork has never been attempted, which is any fork where less than 95% of the network supports the rule change. A fork of this type could provide major instability to Bitcoin and undermine its unique value proposition.

Whether the fork, is Bitcoin XT, Bitcoin Classic, or even just BIP 102 which calls for a increase to 2MB. Bitcoin must set a high bar for consensus rule changes, for one simple reason. Bitcoin’s existence is based on the belief that it is an immutable store of value. If you do not have confidence that a Bitcoin today will be a Bitcoin tomorrow that economic decisions you make today will hold over years even decades. Then why hold Bitcoin, what is it going to offer except the same empty promises made by other political-financial systems.

Miners securing of the network, the Bitcoin cap fixed at 21 million, and its current large network effect represented by its daily transaction volume and global acceptance all contribute to Bitcoin’s value but these things are only possible because of the decentralized consensus Bitcoin imposes on participants.   Strip away Bitcoin’s decentralized nature and Bitcoin offers no unique value proposition as a payment network, a store of value, or even its person 2 person nature.  Without decentralization all of Bitcoin’s features are open to change.  That is the reality that a Hard Fork brings, it is difficult to argue that a hard fork for one change, does not imply there will ever be a hard fork for another.  The fact is a change of the rules even minor, implies a new system where all participants must rethink their economic position within, and with some winners and some losers.

If 51% of the hashing power was used to conduct a double spend  it would be viewed clearly as an attack on the network.  And so should a 51% fork to change the rules.  If a group can coordinate a rule change to their benefit which the network in unison does not support even if it is a small minority, that is an attack.  It is assumed that while 51% of the mining power can change the rules, they will probably not get the support of the economic majority, and they would invalidate the value of their Bitcoin investments and future rewards.  This risk of self harm is what is believed to keep all participants honest. However raising the threshold to 75% as many keep proposing is still dangerously low, and now blurs the picture of whether it is an economic attack or consensus. Couple this with the fact that voting in bitcoin via miner signaling means nothing in terms of actual implementation since votes are not binding.

This low threshold implies at a minimum 25%  of miners must accept a change that they may view as economically harmful this does not even consider what the economic participants may want since Bitcoin has no way to determine this. Even if just 25% of the Bitcoin population is against the change this would be a significant percentage when discussing a $7 billion network.

Consider that the African American population of the U.S. is only 13%, and no one would question their position as a minority with a significant economic and social stake in the United States.  While they live in a democracy where majority rules there is no doubt that their interests must always be taken into account.  This makes it all the more concerning in Bitcoin’s case because Bitcoin is not a democracy but a consensus system. The rules are set and it is expected that to change them you need the entire system to agree.  To dismiss 25% of Bitcoin as some obstinate hold outs that just want to impose their will for no reason is irresponsible, short sighted and brings into question Bitcoin’s claim of decentralization.


As we discussed in Why a Hard Fork Should Be Fought.

Regardless of how altruistic B claims their intent is, regardless of whether B believes that their solution will bring peace on earth, and caviar dreams and limousines to all, one has to assume in a decentralized system that all participants act out of self interest first, and if one party sees no reason for a consensus rule change and the other party does then more than likely the majority party is looking to increase its benefit at the expense of the minority.

As a result, it is a very real possibility that a contentious hard fork (<95%) will be viewed by many economic participants as a bad sign for the future safety of their investments.  We have seen no positive response from markets to recent proposals to hard fork bitcoin contentiously .

As we approach the likely hood of a contentious hard fork there is a real possibility of instability arising in various forms.

  1. Drop in Exchange Volume: The closer we get to a contentious hard fork, transaction volume and exchange volume could fall as coins are withdrawn to user controlled wallets.   Without clear certainty what the results of a fork will be, users will have an incentive to hold their own coins until the uncertainty is resolved.   Similar behavior is exhibited by  financial markets in anticipation of a major Central Bank decisions.
  2. Drop in Bitcoin Value:  With the uncertainty of a hard fork looming, investors may move to fiat or alternate currencies during this period of high risk.
  3. Rethinking of what Bitcoin is: Long term there will be a revaluation of Bitcoin’s claims of stability and immutability. With a new standard of 75% for forks and centralization of mining in bitcoin, 3 pools contribute to 65% of the hashing power, can one still claim Bitcoin is insulated from the flaws of fiat currency where a majority can be swayed to make changes to the system.
  4. Bitcoin 2.0:  Assuming a low threshold hard fork there is a real risk that a percentage of Bitcoin is actually being kicked off the Bitcoin Island. If we assume every participant acts in their self interest, participants who reject a hard fork do so because it against their economic interest. They maybe motivated to fork to an alternate protocol.  At 6-7 Billion dollars there is significant incentive to create an alternative protocol that may provide a better long term return on investment if it can claim greater decentralization.  Exchanges will have every incentive to run a second Bitcoin and in the end what would happen is the economic interests of the “majority” could be permanently damaged.  A “winning” Bitcoin fork would be left competing with a smaller coin which would position itself as the more decentralized “true” Bitcoin.  This would put the “majority” coin in the position of defending its tactic of kicking out a minority by using a low threshold fork that only needs the support of a handful of mining pools. Mining pools that may directly benefit from ever increasing block sizes.
  5. Fork Day could be a disaster.  As mentioned Bitcoin has never performed a contentious hard fork in its history. If one looks at Bitnodes which is not a 100% view of active nodes many nodes still run versions of the Bitcoin Core software which is several releases back. Only 45% of the nodes run current versions.   BIP66 was a soft fork that used the 95% threshold trigger and resulted in some serious unforeseen circumstances over what was supposed to be a modest change.   A 75% triggered fork, with only a 30 day window to upgrade a global network, poses a real threat of encountering unforeseen problems that once again will bring into question Bitcoin’s stability.


Humans beings have very short-term memories and so the thefts of the past are quickly forgotten. It is why Bitcoin is really a minority attempting to fork from the current economic system.   Many “Bitcoiners” have a vivid understanding of economic history, of men voting for their hamburgers today and pushing the bill off for tomorrow.    This is why Bitcoin’s protocol was to be different because history has shown that getting a majority to vote against their long term self-interest is no difficult task and sadly it may appear that we are witnessing the same in Bitcoin.

If we truly believe a 2 mb block increase is good for Bitcoin then achieving a 95% threshold should not be difficult. It would signal that Bitcoin network is making protocol changes in unison and that no participants are favored over others.   If it cannot be achieved then maybe we should listen to the network.

7 thoughts on “Hard Fork Risks and Why 95% Should be the Standard.”

  1. You do bring up some good points, and I agree there are risks. But many of the Core developers themselves have said that a hard fork will probably be necessary eventually. And wouldn’t it be better to do it now than later?

    Also, if you had read the block size increase BIPs, you would have read why 75% is chosen as the threshold – and I think it’s a very good reason. Here’s a quote: “A 75% supermajority was chosen so that one large mining pool does not have effective veto power over a blocksize increase.”

    You extensively discus how you don’t want mob rule (a very valid concern), and yet you’re willing to allow one big miner veto power?

    1. I am not against a fork. The network traditionally does soft forks at 95% which is the standard. The question is why change it now. Because it is clear that there is no confidence in meeting the 95% standard and this is a serious problem. Again no one can prevent you from doing anything in bitcoin. There is no veto power. You imply because I do not pickup my chessboard and sit down and play your game that I am vetoing something. That is incorrect.

      What you are saying is; I want to create a new network and call it Bitcoin even though I cannot bring over the same participants. My failure to bring along everyone probably implies the change I seek is going to harm a portion of the network and now has changed the dynamics of the network. Instead of analyzing my decision I want to lay blame on those that want to continue with the same rules we all agreed with.

      There is a serious problem in the debate when people imply that a Hard Fork is equivalent to passing a new law. No a Hard Fork is equivalent to eliminating your entire government and replacing it with a new one. Maybe you painstakingly keep every item except one, but it doesn’t change the fact you have replaced the entire government. These things are not majority win, they are constitutional crisis and if you don’t get everyone on board you end up with a civil war.

      The point of hard forks being “HARD” is to protect the integrity of the protocol. It is a self destructive process if done incorrectly.

      1. So, you think that a single person should have the power to stop a change in the Bitcoin protocol. And then there is the possibility that that person wants to run the new code, but is being coerced. Now we have someone outside of Bitcoin possibly making the decision.

        As I understand it the SegWit soft fork, begins working right away for everyone who runs it, so it introduces code without anywhere near 95% agreeing to accept it. What if there is a major security flaw in SegWit when only 60% of the network is using it, and we have a huge number of coins stolen, or some other unforeseen problem? I don’t get any say in that, other than my choice to run the software personally.

        “Again no one can prevent you from doing anything in bitcoin. There is no veto power. You imply because I do not pickup my chessboard and sit down and play your game that I am vetoing something. That is incorrect.”

        Okay so 75% it is then.

        On your last point I think I agree, though I don’t get you’re “eliminating your entire government” analogy. There is significant risk to a hard fork, as there is to no change, and as there is to the Core road map. I have come to the conclusion that the risk of a hard fork for a block limit increase comes out with the best cost-benefit-analysis. I could be wrong. And no matter how it turns out at this point there are going to be hard feelings for a while (pun intended :). Hopefully those feelings don’t fester and poison the Bitcoin community further.

Comments are closed.