How to rank competing blockchain projects’ technologies
This morning I offered some advice to a crypto startup on choosing which platform to build upon (or fork from). While I was thinking up a response, I realised how I had unconsciously been ranking the many blockchain technologies out there, so I thought I’d put it down. Here it is.
- Take the following attributes and put them along a decision tree in succession. Each decision has a “winning” branch.
a) open/closed validator set
b) trustless/trust-laden network actors
c) slow/fast finality of transaction state - Play a game where to win is either to:
a) reach the ultimate level without “cheating” by avoiding having to first win at levels 1 and 2
b) perhaps additionally inventing a new level of a kind that can become level 1, making all the current levels move up by one.
If you accomplish 2a, your blockchain is state of the art currently.
If you accomplish 2b, you advance the entire field and make your competitors look like dinosaurs. - Start on level 1:
a) closed validator set: traditional Byzantine fault tolerance (henceforth “BFT”); private or permissioned chains (e.g. Hyperledger)
b) open validator set: move to the next level - Level 2:
a) trust-laden validator set: DPoS coins like Cosmos, EOS (Tendermint), and Federated BFT in the Ripple-Stellar family
b) trustless validator set: move to the next level - Level 3:
a) slow finality: Bitcoin, Ethereum, PoS coins
b) fast finality: move to next level - Top level: you win!
This includes DAG projects like Team Rocket’s Avalanche and perhaps Swirlds Hashgraph (the latter’s status is unprovable because it’s closed source) - Get disrupted when someone invents a new level 1, like the following:
a) undecideable contract outcomes: Ethereum, Bitcoin, and everything else mentioned above
b) decideable contract outcomes (that is, provable security prior to contract execution): Tau chain (repo is here).
Weaknesses of this ranking system
This is very much a flatus vocis (so much so that I had to think twice before publishing it here); expect it to be a ladder one kicks away after climbing it. Murkiness includes the following:
- It’s far from obvious to me what kind of attribute would qualify, in general, as an invention that can function at level 1 and make every other attribute go up a level.
- It’s far from obvious to me whether some inventions could insert themselves at an intermediate level, pushing only some other ones up a level
- I’m not really sure why my current set of attributes are at the levels I put them at. It makes sense, but what’s the underlying logic? I don’t know. Will a tree of merely binary branches cover all future developments? I don’t know. How could a decision tree necessarily represent something real about the technology? Merely obliquely, that’s how. Treat with caution.
The purpose of this ranking system
This is probably just a loose didactic image, useful only insofar as it facilitates further, better, thinking. It’s a practical tool only — like one of those gimmicky multitools that do 15 things dubiously well and do nothing with provable reliability. I’ll earmark this to see whether it changes over time, and if it must be adapted to handle future advancements, and/or if future advancements in blockchain tech can shed light on the (undiscovered) underlying relations between levels. It may yet be redeemable.