Bitcoin's Third Block Reorg in a Month Raises No Alarm

For a few seconds on Friday, Bitcoin's blockchain briefly had two competing versions of itself. Mining pools AntPool and SpiderPool each found a valid block at height 966,500 within moments of one another, sending two different chains racing across the network before one of them was discarded entirely. AntPool's version carried a timestamp of 11:30:30 UTC and packed in 4,359 transactions; once AntPool went on to mine the next block, 966,501, nodes across the network recognized its branch as the one with more accumulated proof-of-work and switched over, orphaning SpiderPool's block along with its roughly 3.1389 BTC reward.

According to Galaxy Research, whose node had actually received SpiderPool's block first and briefly treated it as the chain tip, the reversal happened only after AntPool's follow-up block gave its branch the heavier cumulative work. That's the entire mechanism behind a one-block reorg: two miners solve a block at nearly the same time, the network temporarily forks, and whichever chain gets extended first wins, with the loser's block and its coinbase reward vanishing from the record as if it had never been found.

What makes this one notable isn't the event itself but the frequency. It's the third time in four weeks that Bitcoin has produced a one-block reorg — the first landed at block height 962,722 on August 16, and the second followed at block height 963,853 on August 24. Three such events in a single month is an unusually tight cluster for a phenomenon that, historically, surfaces only a handful of times a year. Bitcoin's block-discovery process is inherently probabilistic, and any two large pools can theoretically find a valid block within the same few-second window purely by chance, so a short run of coincidences isn't on its own proof of anything systemic.

Galaxy Research was direct about what this isn't: there was no consensus bug and no successful attack behind Friday's event. Short-lived competing blocks are, in the firm's own words, a completely normal part of how a decentralized, probabilistic mining process resolves itself, and Bitcoin's difficulty-adjustment and longest-chain rules exist precisely to settle these races automatically within a single confirmation. No user funds were at risk, and no transaction that had already been confirmed by a second block was ever reversed — the only casualty was SpiderPool's mining reward for a block that, seconds after being found, stopped counting.

Still, the clustering is worth watching rather than dismissing outright. A rising one-block reorg rate can reflect faster block-propagation races between well-connected mining pools, growing hashrate concentration among a handful of large operators, or simply a statistical cluster that will fade as quickly as it appeared. Bitcoin's broader infrastructure has faced its own share of scrutiny this year, and layered systems built on top of the base chain, including sidechains like Liquid that recently had to pause transaction processing of its own, tend to be far more exposed to disruption than the base layer's own occasional orphaned block. For now, miners are the only ones footing the bill, in the form of block rewards that briefly existed and then didn't.

The pools involved are also not minor players, which is part of why the pattern is drawing attention. AntPool currently controls roughly 19% of Bitcoin's total hashrate, making it the network's second-largest pool behind Foundry USA, while SpiderPool sits around 9%, good for fourth place. Foundry USA and AntPool combined now account for close to 60% of all Bitcoin mining, a concentration level that has drawn recurring scrutiny from researchers who track how much of the network's security rests on a small handful of large operators. When two pools of this size are producing blocks within seconds of each other often enough to trigger three separate reorgs in a month, it raises a fair question about whether propagation delays between the biggest pools are widening even as their combined share of hashrate grows, though Friday's event alone isn't enough data to draw that conclusion with confidence.

None of the transactions inside SpiderPool's orphaned block were lost for good. Barring a handful of edge cases involving unconfirmed dependent transactions, the transactions it had bundled simply went back into the mempool and were picked up by the next block that miners found, so users saw nothing more than a brief delay in confirmation rather than any risk to their funds. That's the practical distinction between a one-block reorg like this one and the kind of deep reorganization that would actually worry Bitcoin's security model — the latter requires an attacker to out-mine the honest network over many blocks, a cost that remains prohibitively high at Bitcoin's current hashrate.