💻 computer science

Quantifying Bitcoin Network Resilience Through Critical Scenario Discovery: A Dual-Layer Framework for Discovering Contentious Fork Conditions in Decentralized Consensus

This paper utilizes the Patient Rule Induction Method (PRIM) on 1,330 Warnet simulations to demonstrate that Bitcoin's resolution of contentious soft forks is primarily determined by economic weight distribution rather than hashrate majority, identifying specific thresholds for economic support and pool commitment that predict whether a fork will cleanly resolve or persist as a chain split.

Original authors: Peter Foytik, Sachin Shetty, Ross Gore, Eranga Bandara

Published 2026-08-07

📖 7 min read🧠 Deep dive

CC BY 4.0

Original authors: Peter Foytik, Sachin Shetty, Ross Gore, Eranga Bandara

Original paper licensed under CC BY 4.0 (http://creativecommons.org/licenses/by/4.0/). ✨ This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer

Imagine a massive, global digital town square where everyone agrees on the rules of a game called Bitcoin. In this town, the "miners" are like construction crews who build new blocks of the town's history, and the "economic nodes" are the banks, shops, and wealthy families who actually use the town's currency to buy and sell things. Usually, everyone agrees on one single version of the town's history. But sometimes, a group wants to change the rules—maybe to make the game faster or safer. This is called a "soft fork." The big question is: Will the whole town agree to the new rules, or will the town split into two rival villages, each with its own version of history?

For a long time, people thought the answer depended entirely on the construction crews (the miners). If the crews with the most powerful tools (hashrate) wanted the new rules, everyone would follow. But this paper suggests that's like thinking the construction crew decides the price of houses; in reality, it's the people buying and selling the houses (the economy) who really hold the power. The researchers wanted to find out exactly how much economic power is needed to force a split to resolve, and what happens when the miners and the economy disagree. They didn't just guess; they built a giant, virtual simulation of the entire Bitcoin network to watch how it behaves when things go wrong.

The Great Bitcoin Fork Experiment

To understand what happens when a digital town tries to change its rules, the authors built a virtual world using a tool called Warnet. Think of this as a super-advanced video game where they didn't just pretend to be miners and banks; they actually ran the real software that Bitcoin uses, but inside a controlled computer lab. They created 1,330 different scenarios, tweaking the settings like a scientist adjusting knobs on a machine. They asked: "What happens if the miners love the new rules but the banks hate them? What if the banks are split down the middle? What if the price of the coin starts to crash?"

The goal was to find the "tipping points"—the exact moments where a messy disagreement turns into a clean resolution or a permanent disaster. They used a clever math trick called Scenario Discovery to sift through all that data and find the hidden patterns, kind of like using a metal detector to find the specific spot on a beach where all the buried treasure is hidden.

The Big Surprise: It's Not About the Shovels

The most shocking thing the paper found is that having the most powerful mining shovels doesn't guarantee a win. In fact, in the realistic scenarios they tested, the amount of mining power (hashrate) had almost zero effect on who won the fork, as long as there was a decent amount of economic support on one side.

Imagine a tug-of-war where one team has 90% of the strongest, fastest pullers (miners), but the other team has 93% of the people holding the rope (the banks and shops). The paper shows that the team with the rope wins, even if the other team is pulling harder. The miners might be strong, but if the banks refuse to trade on their chain, the miners' work becomes worthless. The simulation showed that if a new version of Bitcoin only had 7% of the economic activity (the shops and banks), it would lose, even if it had 90% of the mining power.

The Three Zones of Forking

The researchers discovered that the outcome of a fork depends on three specific "zones" of economic support, which act like traffic lights for the network:

The Red Light Zone (Below 45–50% support): If the new version of Bitcoin has less than about 0.45 to 0.50 (or 45–50%) of the total economic weight, it simply cannot win. No matter how many miners switch over, the network will snap back to the old version. It's like trying to start a new country when nobody has any money; the economy just won't support it.
The Green Light Zone (Above 78–82% support): If the new version has more than about 0.78 to 0.82 (78–82%) of the economic weight, it wins automatically. The price of the new coin becomes so much better than the old one that even the stubbornest miners are forced to switch. The "economic signal" is so loud that it overrides everything else.
The Yellow Light Zone (The Messy Middle): This is the most interesting part, between 0.50 and 0.78. Here, the outcome isn't decided by the total number of miners or even the total money. Instead, it depends on a very specific, counter-intuitive rule involving the largest mining pool.

The "Flip-Point" Paradox

In the messy middle zone, the paper found a strange "flip-point" at about 0.214 (21.4%) of the committed mining power. This is where things get weird.

Imagine the biggest mining pool in the world is like a giant anchor. If this anchor is on the "old" side, but the economic pressure is high enough, the anchor might get "trapped." The price of the new coin is so good that the anchor has to switch to the new side to make money, even if they didn't want to. This forced switch is powerful because it signals to everyone else that the new side is safe.

However, if that same giant anchor is already on the new side because they really believe in it (ideological commitment), it can actually hurt the new side. Why? Because if the anchor is already there, the "old" side becomes a tight-knit group of true believers who refuse to leave. They hold their ground, and the new side gets stuck in a permanent split.

So, the paper suggests a paradox: Sometimes, having the biggest miner publicly commit to the new rules can actually make the new rules lose. It's better if the big miner is "trapped" by economics and forced to switch, rather than choosing to switch voluntarily.

The Two Layers of the Game

The paper also explains that a fork happens in two separate layers that don't always agree:

Layer 1 (The Miners): This is about who is building blocks. This layer is mostly decided by whether the big mining pools are stuck or free.
Layer 2 (The Economy): This is about who is using the money. Even if the miners agree on one chain, the economy might stay split for a long time if the price difference isn't big enough to force everyone to move.

The study found that regular users (individual people running a computer at home) have almost no influence on this outcome. In their simulation, individual users were so small compared to the big banks and exchanges that they couldn't move the needle at all. It's like a single person trying to decide the price of oil; unless they are a giant corporation, their voice doesn't change the market.

What This Means for the Future

The authors didn't just find these numbers; they turned them into a checklist for anyone watching a Bitcoin fork. Instead of asking "Who has the most miners?", they suggest we should ask:

Is the economic support above 50%? If not, the new version is doomed.
Is the economic support above 82%? If so, the new version has already won.
In the middle zone, is the biggest mining pool "trapped" or "committed"? If the biggest pool is just holding on because they have to, the new version might win. If they are holding on because they love it, the new version might lose.

The paper is careful to say these are results from simulations using specific settings for the year 2026. It's a map of how the network likely behaves under these conditions, not a crystal ball that predicts the future with 100% certainty. But it gives us a much clearer way to understand the invisible forces that decide whether Bitcoin stays as one big town or splits into two rival villages.

Technical Summary: Quantifying Bitcoin Network Resilience Through Critical Scenario Discovery

Problem Statement
Bitcoin's consensus is traditionally analyzed through protocol rules and proof-of-work mechanics. However, the operational reality involves a heterogeneous network of independently configured nodes with divergent economic stakes. While the community qualitatively recognizes that different participant classes (developers, miners, economic nodes, users) hold varying governance powers (as outlined in the BCAP framework), there is a lack of quantitative understanding regarding which actors determine fork outcomes, under what specific conditions, and at what precise thresholds. Existing literature focuses on formal verification or post-hoc historical analysis, failing to answer the operational question of what specific combinations of economic and mining conditions push a network from a clean resolution into a persistent, contentious split. This paper treats contentious fork resolution as an outcome to be discovered rather than assumed, addressing the gap in quantifying infrastructure risk as Bitcoin assumes the role of critical infrastructure.

Methodology
The authors employ Scenario Discovery, a methodology from decision-making under deep uncertainty, applied to a large ensemble of controlled network simulations.

Simulation Platform: The study utilizes Warnet, a Kubernetes-based framework that runs real, unmodified bitcoind nodes (Bitcoin Core) rather than behavioral mocks. This ensures that block validation, chain selection, and propagation follow production consensus code.
Network Model: The network is calibrated to approximately 1:400 scale (as of February 2026), comprising 60 nodes representing eight mining pools, 24 economic nodes (exchanges, custodians, payment processors), and 28 user nodes. Economic weight is defined as $0.7 \times \text{custody}_{BTC} + 0.3 \times \text{volume}_{BTC}$ .
Experimental Design: The authors ran 1,330 valid scenarios across 21 sweep configurations. Parameters included economic weight distribution, mining-pool commitment, pool ideological tolerance, hashrate distribution, and difficulty-retarget regimes (144-block vs. 2016-block intervals).
Analysis Techniques: The study uses the Patient Rule Induction Method (PRIM) to identify axis-aligned parameter regions that maximize the density of contentious outcomes. These findings are supported by random-forest classification for feature importance and logistic regression with interaction terms.
Behavioral Models: Mining pools are modeled with ideological preferences and profitability thresholds. Economic nodes follow an ideology-override logic with inertia. Price divergence is modeled as a function of chain dominance, economic weight, and hashrate, capped at $\pm 20\%$ per chain.

Key Contributions

Ensemble-Simulation Framework: A Warnet-based framework enabling systematic Scenario Discovery across the governance parameter space of contentious soft forks using real consensus code.
Dual-Layer Analysis: A model integrating technical consensus metrics with economic-topology metadata, distinguishing between hashrate resolution and economic adoption.
Structural Quantification of BCAP Actors: The paper provides specific quantitative thresholds for actor classes, including an economic-support floor, an override ceiling, an Economic Self-Sustaining Point (ESP), and a pool-commitment "flip-point."
Independence of Resolution Layers: The establishment that fork outcomes resolve on two independent layers (hashrate and economic adoption) governed by different parameters, and that individual user nodes have no detectable influence on outcomes under modeled economic weightings.
Operational Guidance: Translation of findings into three monitoring questions answerable from publicly observable data.

Key Results

Hashrate Non-Causality: Contrary to common intuition, hashrate majority is not the primary determinant of fork resolution under realistic economic conditions (specifically when economic support is $\ge 0.60$ ). The "Difficulty Adjustment Survival Window" neutralizes the hashrate advantage before the economic cascade resolves.
Economic Weight Dominance: The distribution of economic weight across a partition is the primary determinant of resolution.
- Economic Support Floor ( $\sim 0.45\text{--}0.50$ ): Below this threshold, no cascade is possible; the incumbent chain wins regardless of pool configuration.
- Economic Override Ceiling ( $\sim 0.78\text{--}0.82$ ): Above this threshold, the economic signal forces capitulation of all pools, ensuring the upgrading chain wins universally.
- Economic Self-Sustaining Point (ESP, $\sim 0.74$ ): Between the floor and ceiling, economic momentum becomes self-reinforcing.
The "Foundry Flip-Point" ( $\sim 0.214$ ): A counter-intuitive structural threshold exists regarding pool commitment. If the fraction of committed hashrate on the upgrading chain crosses $\sim 0.214$ (specifically involving the largest pool, Foundry USA), the outcome can paradoxically reverse. Below this point, the largest pool is economically trapped and forced to switch, triggering a cascade. Above this point, the largest pool holds by conviction, purifying the opposing bloc and allowing them to resist the cascade.
Regime Dependence: The dominant causal parameter reverses based on the retarget interval. In a 144-block (accelerated) regime, economic split is dominant. In the operationally relevant 2016-block regime, pool commitment structure becomes the primary driver (52.8% feature importance), as pools must sustain losses through a full 14-day epoch.
User Node Insignificance: Individual user nodes (retail holders, solo miners) collectively hold a $2,197:1$ economic-weight disadvantage relative to institutional actors. The study found zero detectable influence of user nodes on fork outcomes in any configuration, challenging the efficacy of User-Activated Soft Forks (UASF) unless they successfully persuade economic infrastructure.

Significance and Claims
The paper claims to provide the first structural quantification of the BCAP actor framework, moving governance analysis from qualitative descriptions to measurable thresholds. It asserts that the standard framing of fork assessment (focusing on hashrate majority) is actively misleading within the contested "inversion zone."

The significance lies in reframing governance risk:

Monitoring: Governance actors should monitor economic split ratios and the specific ideological assignment of the largest mining pool rather than aggregate hashrate.
Strategic Insight: Publicly committing a large pool to an upgrade can be counterproductive under intermediate economic conditions, as it may prevent the "forced switch" mechanism that drives cascades.
Risk Assessment: The findings identify specific "danger windows" (e.g., the period before the first difficulty adjustment) and structural ceilings on user influence, offering a data-driven basis for assessing infrastructure resilience during contentious upgrades.

The authors maintain modesty, noting that specific threshold values are calibrated to the 2026 landscape and that the mechanisms are general while the exact numbers may require recalibration for materially different network topologies or price divergence dynamics.