This paper is one of the few that I’ve managed to track down that tries to analyze Nakamoto Consensus from the view of classical consensus algorithms. Most content out there treats the two as though they were completely different beasts.

What’s interesting about this is it describes an oracle they call the PCNELE Oracle that models the Bitcoin protocol properties pretty well. It also eliminates the 50% hashing requirement from the model. What’s left behind is an analysis of the longest fork choice rule.

I wonder what the security of LFW style protocols are then, is there a hidden 2/3 voting requirement underneath Bitcoin’s hashing system as well?