A Race Between Pools

Bitcoin experienced a two-block reorganization at block height 941880, with Foundry USA Pool emerging as the winner over AntPool and ViaBTC. The competing chains diverged when both Foundry and AntPool broadcast valid blocks for height 941881 within approximately 12 seconds of each other. Network propagation timing determined which block individual nodes saw first, creating a temporary fork.

AntPool's block at height 941881 (hash 00000...eb2d9f3) and ViaBTC's block at height 941882 (hash 00000...f9687b2e) both became stale as Foundry mined three consecutive winning blocks in rapid succession. Fork.observer data confirmed the arrival gap between the competing blocks at height 941881 was roughly 12 seconds, comfortably within the window where propagation delays create natural forks. As one analyst noted, blocks arriving within 15 to 30 seconds of each other will reliably produce a temporary fork.

Foundry extended its run to seven consecutive blocks, resolving the chain split decisively. Transactions from the stale chain were returned to the mempool for inclusion in subsequent blocks. No confirmed transactions were permanently lost.

A Pattern of Rarity

Two-block reorgs are uncommon but well-documented in Bitcoin's 16-year history. Previous known instances occurred at heights 788687/788686 in May 2023, heights 656478/656477 in November 2020, heights 358999/358998 in June 2015, and heights 261199/261198 in October 2013.

The pattern suggests these events occur roughly every 150,000 blocks, or approximately every three years. The gap since May 2023 was about 153,000 blocks, fitting neatly within the historical range. Single-block stale blocks happen far more frequently, sometimes multiple times per month, but extending the race across two consecutive heights requires a narrower timing coincidence.

Bitcoin's difficulty adjustment and block propagation characteristics make occasional short-lived forks a statistical certainty, not a design flaw. The protocol accounts for them explicitly through its longest-chain rule.

The Centralization Red Herring

Foundry USA Pool currently commands roughly 30% of Bitcoin's total hashrate, a figure that surfaces centralization concerns whenever the pool mines extended block sequences. The distinction between pool hashrate and pool control matters here.

Foundry's hashrate is aggregated from thousands of independent mining operations that freely choose to direct their hashpower through Foundry's infrastructure. These miners can and do redirect to competing pools at any time, often based on fee structures, payout methods, or geographic latency. The 30% figure reflects a collective, voluntary contribution, not a centrally directed resource.

The two-block reorg was a statistical outcome, not a display of concentrated power. Whichever miners happened to be contributing hashrate to Foundry at that moment collectively solved two blocks faster than the miners contributing to AntPool and ViaBTC. The same result could have favored any pool with a comparable share.

Why It Matters

When events like this occur, expect noise from two predictable directions: sh!tcoin promoters claiming Bitcoin's proof-of-work is broken, and smaller mining pool operators leveraging the moment to spread fear about larger competitors and funnel hashrate toward their own operations. Both narratives serve commercial interests, not technical understanding.

A two-block reorg is Bitcoin's consensus mechanism doing exactly what it was designed to do. Nodes detected the fork, evaluated chain weight, and converged on the heaviest chain. No coordinator stepped in. No committee voted. No emergency patch was deployed. The protocol resolved the conflict autonomously, as it has reliably since 2009. Bitcoin's resilience is not the absence of temporary forks. It is the automatic, trustless resolution of them every single time.