Full version of the post will be linked here soon!
TL;DR
With multiple MCP protocols in the design and implementation phases, we think it is timely to discuss how location choice interacts with protocol design to achieve welfare-efficient, fair, and censorship-resistant functioning in these systems. In Price of Decentralization in Block Building, we tackle this problem and answer the question: will the location choices of multiple self-interested proposers, choosing regions to maximize individual utility, improve access for geographically distributed users and tx sources? Modeling decentralized block building as a stochastic coverage game, in which builders choose geographic regions and information sources emit transactions over the course of a block-construction round, we study the welfare and utility concentration properties of this system. To analyze when welfare losses arise in concrete instances, we develop a simulator and compare self-interested equilibrium outcomes with centralized-planner benchmarks across a set of controlled simulation environments. Overall, we provide a first look into interactions among multi-proposer systems, geographic region selection, and welfare and concentration implications, and show why protocol design should treat the location-choice dimension as a first-class citizen.
Interesting framing, but there is a fundamental paradox here. We keep trying to solve “geo-decentralization” at the protocol and game-theoretic level, while completely ignoring the underlying physical reality.
True decentralization cannot exist on centralized infrastructure. It doesn’t matter how well we optimize proposer location choices in a model—at the end of the day, almost all these nodes land on AWS, Google Cloud, DigitalOcean, or Hetzner.
By relying on standard datacenters, we aren’t creating a decentralized network; we are just renting slots on a handful of corporate servers. The moment Web3 relies on Web2 tech giants for hosting, “geographic diversity” becomes an illusion managed by Amazon’s and Google’s routing policies and TOS.
Until the crypto space solves the hardware and physical infrastructure layer (bare-metal, home-hosting, decentralized hardware networks), optimizing server location choices is just trying to achieve decentralization within a fundamentally centralized environment.
I agree with your point. While the modelling choices can also incorporate infra properties of “decentralized hardware networks”, today, the reality is that time-sensitive applications/services run on centralized servers. Perhaps, the design goal of a protocol could be to minimize the incentives for this; or at least, redistribute rewards in a way which does not create significant asymmetries for participants who use decentralized infrastructure.