The Bitcoin network experienced a rare two-block reorganization on March 23rd with a block height of 941,880. Foundry mined six consecutive blocks, and AntPool and ViaBTC temporarily extended competing branches.
The chain resolved the fork as designed, following the path with the highest hash rate. Bitcoin worked exactly as designed and that assumption was verified.
Heuristics that no one has labeled
The law of six confirmations is one of those myths that has traveled so far from its origins that most people who repeat it are unable to reproduce why there is a number six.
The answer goes back to Satoshi Nakamoto’s 2008 white paper, which modeled finality as the probability of catching up. When enough blocks pile up on top of transactions, the computational cost of rewriting the history becomes prohibitive for an attacker with limited hashing power.
Six blocks has become community abbreviation for “sufficiently secure,” although the white paper assumed the attacker controlled about 10% of the network’s hashing power.
This assumption has been behind a lot of work for 16 years.
Jameson Ropp clarified its implications in his analysis of confirmation risk. The comfort level built into the six checks depends on who is on your network and how well you run that network.
In Nakamoto’s catch-up model, six confirmations for an attacker with 10% of the hashing power creates a reversal risk of approximately 0.02%. 20%, this number rises to approximately 1.43%. 30%, it reaches about 13.2%.
The share held by Foundry in the most recent pool share snapshot was 32.2%, and the same model would have a reversal risk of nearly 18.9% with six confirmations.
Since mining pools are not organized attackers by default, they do not fit the output of these models. Foundry USA describes itself as an institutional-grade pool built for miners that coordinates many independent operators.
Miners can and do switch pools. Carrying out an overt attack would be economically self-destructive for any reasonable pool operator. Concentrating on block production changes the risk models people use to determine when payments will ultimately be made, regardless of how distributed the underlying machines are.
A 2022 Latency Security Analysis found that even with a 10% attacker and a 10 second propagation delay, the probability of a safety breach is between 0.11% and 0.35% over 6 checks.
Even under far more favorable conditions than today, the Six was never a hard ceiling.
3 conditions at once
The circumstances surrounding the reorganization are important.
The Bitcoin network currently runs three conditions simultaneously that put pressure on the six-confirmation heuristic, but in practice this situation is rarely encountered.
Over the past three days, Foundry has held about 31% of the global hashrate, while AntPool has held about 18.4% and ViaBTC has held 10.5%, according to Hashrate Index data. Together, these three pools account for approximately 60% of block generation.
The concentration of coordinator power has also increased by any reasonable measure over the past few years.
At the same time, the mining economy deteriorated sharply. On March 21st, the difficulty level decreased by 7.76%, the largest negative adjustment in 2026. Hashprice’s daily average per petahash in February was $32.31, down nearly 18% from the previous month, and at one point hit an all-time low of $27.89.
In the last 24 hours of available data, transaction fees were only 0.57% of total block rewards.
As margins compress and fee income dries up, small and medium-sized miners face increasing incentives to pool with the coordinator that offers the best variance reduction. This usually means that an already large pool becomes even larger.
January’s winter storm brought with it a notable counterproposal. Foundry’s hashrate reportedly dropped by about 60%, or nearly 200 exahashes per second, during that period, indicating that pool shares could be quickly redistributed when external conditions change.
Against this backdrop, the six-check rule lacks an automatic adjustment mechanism when pool shares change.
| situation | latest reading | 6 Why confirmation rules are important |
|---|---|---|
| pool concentration | Foundry ~31%; Antpur ~18.4%; Via BTC ~10.5% | The fixed confirmation assumption is not comfortable for large-scale payments, as most of the block generation is concentrated in a small number of coordinators. |
| Top 3 concentration skills | Approximately 60% of total block production | Finality depends not only on the number of blocks, but also on how hashing power is distributed among competing pools. |
| Difficulty adjustment | March 21 -7.76% | Significant negative adjustments indicate stressed mining conditions and weak economics across the network. |
| February hash price | $32.31 per day per PH | Declining miner profits increases the incentive for small miners to seek stability in larger pools. |
| Intra-month hash price is low | $27.89 | The deeper the margin compression, the larger and more attractive the pool for variance reduction. |
| Fee contribution to compensation | 0.57% in the last 24 hours | Weak fee support will make miners even more reliant on a shrinking block subsidy economy. |
| Counterpoint: Redistribution risk | January’s winter storm reportedly caused foundries to drop by up to 60% | Concentration increases, but it is not fixed. External shocks can also quickly rebalance pool shares. |
In fact, the industry’s biggest venues waived the six-check standard in a quiet operational decision made years ago.
Coinbase requires two confirmations to mark a BTC deposit as pending, while Kraken and Gemini each require three confirmations.
None of these thresholds are wrong for your use case. For typical retail deposits, two or three confirmations represent a fully defensible risk tolerance.
The gap between these real-world numbers and the civilian standard of 6 indicates that the “6 confirmations” have always been a cultural product rather than a universal policy.
Ropp’s framework argues that this gap should be widened more deliberately. The required confirmations should scale with the transaction amount and the attacker’s financial situation.
A $500 retail deposit and a $50 million OTC settlement do not share the same risk profile, and the honest version of the finality guidance explicitly states that.
numbers that don’t change
The current hashrate concentration scenario has mixed consequences, which are alarming to users.
The good news is that mining margins will eventually recover and hashrate will be redistributed to a broader pool of coordinators as new entrants compete for a share.
January’s storms have already demonstrated that Foundry’s advantage can quickly disappear under the right conditions. If concentration eases and hash prices recover, 6 confirmations will remain a reasonable default for large BTC payments.
On the contrary, Foundry is likely to maintain above 30%, and its top 3 concentration remains stable. Exchanges, OTC desks, and merchants that handle high-value transfers secretly raise internal thresholds or formalize dynamic hierarchies tied to observable pool sharing data, so it doesn’t take a malicious event for this standard to drop.
Under the Nakamoto model, six confirmations against a fully coordinated attacker of 32.2% would leave a catch-up risk of approximately 18.9%. This number is really hard to reconcile with language like “virtually irrevocable” for transfers in the tens of millions of dollars.
In this situation, all that is required is for the concentration of the pool to remain as it is, but for the gap between the public’s standards and the actual risk to widen enough that those who are betting money stop ignoring it.
Bitcoin’s settlement guarantee has always been “six blocks with a fixed hashpower allocation and a fixed risk tolerance.”
2 The block reorganization created a rare moment in which the gap between Bitcoin’s finality folklore and its underlying mathematics became impossible to ignore.
Given this moment, the days when the six-confirmation rule served as a universal, disqualifying standard are coming to an end.
(Tag Translation) Bitcoin
