Nakamoto Rationality
Satoshi’s Overpotent Insight
The incentive may help encourage nodes to stay honest. If a greedy attacker is able to assemble more CPU power than all the honest nodes, he would have to choose between using it to defraud people by stealing back his payments, or using it to generate new coins. He ought to find it more profitable to play by the rules, such rules that favour him with more new coins than everyone else combined, than to undermine the system and the validity of his own wealth.[1]
One of the most powerful, brilliant, and impactful insights underpinning the success of Bitcoin (and more generally, all modern cryptocurrency) is the sly note above from §6 of Satoshi Nakamoto’s seminal whitepaper. Here lies a very clever observation with a profound takeaway. Those with power, stake, or privilege in any given system have a natural incentive to support that system, at least insofar as the value that they stand to realize, capture, or otherwise derive from it. I refer to this principle as the tenet of Nakamoto Rationality.
Undeniably Clever
The idea underpinning Nakamoto Rationality makes obvious sense from first principles: if someone were competitively advantaged in a particular money system, they would benefit more from the continued, healthy, “normal” operation of that system than they would from behaving in true Byzantine fashion. From the perspective of someone with the resources required to participate in the ledger-extension consensus process (which, crucially, scale with value accrual), the value proposition of good behavior far outweighs that of attacking, at least in the vast majority of circumstances. This is undeniably clever.
When incentives are oriented in such a powerful way, it can simplify much of our subsequent reasoning and analysis (at least with regards to base-layer network assurances). In a way, Nakamoto Rationality presents a persistent and reasonably-defensible fallback, a sort of “get out of jail free” tucked up the sleeve of the cryptocurrency apologist. Even if you do manage to dream up some sophisticated type of miner attack that could fundamentally harm Bitcoin, in most cases, it’s reasonable to assume that the miners best positioned to make that sort of attack are the ones who would least like such an attack to take place. The incentives point away from hostile miners, in the general sense.
Of course, we shouldn’t let our reasoning be cut short with a “sounds good enough” dismissal. There are always caveats and shades of gray to consider. In many circumstances, the assumption of Nakamoto Rationality might be challenged or rendered irrelevant. For instance, if a government (or coalition of governments) managed to seize direct control over a significant portion of the Bitcoin network’s mining/hashing capacity, the incentive to harm or attack Bitcoin might very well outweigh the incentive to continue mining normally, collecting block subsidies and transaction fees. In minority-hash-share altcoin networks, the potential of the majority-hash-share coin’s miners temporarily disrupting the minority coin’s network represents an ever-present threat, mitigated mostly by the opportunity cost of doing so outweighing the perceived positive effects on the primary network from “squashing” the minority competitor. There are many other out-of-the-ordinary but still-very-possible scenarios in which the assumption that consensus participants have the network’s best interests aligned with their own is violated or undermined. So while it’s a very nice assumption that resonates intuitively, we need to be aware of the limits of Nakamoto Rationality, and prepared to accept that in some circumstances, it might be necessary but not sufficient for robustness or continued healthy operation.
Consensus-Primacy of Nakamoto Rationality
Nakamoto Rationality, as a core principle, is often embraced and elevated in terms of consensual primacy in cryptocurrencies other than Bitcoin. In some cases, in fact, the entirety of a cryptonetwork’s consensus security model is predicated on this axiom.
Bitcoin itself is, in both theory and practice, architected to avoid excessive reliance on the assumption of Nakamoto Rationality; the relevance of the effect was only explicitly remarked upon in the context of miner competition, which is a subcomponent specifically afforded only limited systemic influence. Even so, the fundamental reasoning, namely that the value of the internal, algorithmically-scarce digital tokens themselves can propel the healthy operation of the network and its ledger through incentive-alignment, makes a certain intuitive sense. It seems safe to assume that those who stand to profit from a particular system (or subsystem) functioning normally will probably (at least usually and hopefully almost always) behave in a way that doesn’t undermine it doing so, all else equal. Even more compelling than this a priori approach, though, is the apparent empirical validation of the principle, which has been repeatedly demonstrated in the growth of Bitcoin and the other cryptocurrencies which have followed its momentous 2009 launch.
The Curse of Overpotence
A case could be made, though I hesitate to personally make it, that the principle of Nakamoto Rationality is a strong enough phenomenon in practice, i.e. that the assumption of the existence thereof holds universally true enough, that it acts counter to the potential accrual of network effects in terms of ledger competition. In other words, the premise that participants (and more specifically potential attackers) can reasonably be expected to be Nakamoto Rational in general might be enough to “pick up the slack” on variants of consensus systems which might otherwise be less-than-robust or subtly flawed. This, in turn, might sustain competition, at least in terms of grassroots mindshare, for what could have been a won-before-it-started sort of battle.
The natural questions to follow this, then, are whether this is a recursively applicable phenomenon, whereby the effect might self-mitigate (i.e. self-cannibalize) in negative-feedback-loop fashion,[2] whether it might be curtailed or defended against effectively, and if so whether Bitcoin (or any other coin that exists today) already includes any such defense mechanisms to a sufficient degree. Any effective defense mechanism against this type of loss would require uniqueness or inimitability of some sort. A trust-minimized context renders this a particularly arduous criteria to satisfy.* But it’s worth noting, while we’re here, that one obvious and compelling candidate in this department is the defense mechanism of the Network Effect.
In other words, in a dog-eat-dog world, especially without anything else to go on, you probably want to bet on the biggest, strongest dog.
* For trust-minimization to be feasible, users need to (at least in theory) be able to audit the source code of their clients. As Satoshi put it:
Being open source means anyone can independently review the code. If it was closed source, nobody could verify the security. I think it’s essential for a program of this nature to be open source.
Beyond the code being innately and inescapably available-to-copy, things like patents or IP law enforcement fundamentally do not make sense in a genuinely-leaderless network structure like Bitcoin’s.[3]