The Number Most Bitcoin Users Never Check
Pull up a mining pool pie chart. You see it instantly: a handful of fat colored slices owning most of the circle, a thin rind of smaller operators filling the gap around the edge. Most readers glance at it and move on, which is exactly the wrong instinct. That chart is, arguably, the single most important health metric the Bitcoin network produces, more revealing than price, more honest than sentiment indices, and directly tied to whether Bitcoin can actually do the one thing it promises: process transactions without asking anyone's permission.
So let's talk mechanics, not slogans.
What Censorship Resistance Actually Requires
Bitcoin's censorship resistance rests on one structural guarantee: no single actor can permanently block a valid transaction from being confirmed. A miner can choose not to include your transaction in their block. The protocol permits it. The question is whether they can make that choice stick across the whole network.
They can't. As long as they don't control the majority of hash rate.
Here's the mechanism in plain terms. Suppose a pool running 30% of global hash rate decides to blacklist all transactions touching a particular address, omitting them from every block it finds. Because it finds roughly 30% of all blocks, those transactions still get picked up by the other 70% of the network. Slower confirmation, yes. Permanent exclusion, no. The math is on the user's side.
Now shift the numbers. Give that pool 51% of hash rate and the picture changes entirely: they can out-mine the rest of the network, reorg blocks that included the blacklisted transaction, and keep reorging until it disappears from the canonical chain. That's not a theoretical attack. It's Nakamoto consensus working exactly as designed, just in the wrong hands.
The threshold everyone quotes is 51%. The threshold worth worrying about is considerably lower.
Why 51% Is the Wrong Number to Watch
This is the part most guides skip.
A coordinated censorship attack doesn't require a single pool. It requires colluding pools whose combined hash rate exceeds 50%. Consider a scenario where the top four pools each hold 15 to 18% of global hash rate. No single one crosses any alarming threshold. But if three of those four operate under the same regulatory jurisdiction, share infrastructure providers, or are majority-owned by overlapping investment groups, their effective coordination risk is far higher than the individual numbers suggest.
Call it the jurisdictional concentration problem. Bitcoin's network has, at various points in its history, seen the majority of its hash rate concentrated in a single country, which creates a surface area for state-level pressure that pure pool-percentage numbers don't capture. A government doesn't need to hack a mining pool. It can simply issue a compliance order.
The number to actually watch is this: how many independent entities would need to cooperate, or be coerced, to assemble a censoring majority? When that number is two or three, the network is fragile in ways the pie chart doesn't show. When it's eight or ten, you can sleep easier.
There's also a subtler attack well below 51%. A pool with 30% of hash rate can make censorship expensive for users by selectively mining only fee-maximizing blocks that exclude target transactions. This doesn't guarantee exclusion. Close enough to matter, though.
A Worked Scenario: Two Miners, One Address
Take two hypothetical miners: Carla and Dmitri, both running medium-sized operations, both pointed at the same pool. Carla is in a jurisdiction that passes a financial surveillance law requiring pools to screen transactions against a sanctions list. Dmitri is in a different country with no such requirement.
The pool, based in Carla's jurisdiction, begins filtering. Every block it finds omits transactions from the flagged addresses. The pool holds 22% of network hash rate.
Dmitri's machines are now, functionally, participating in the censorship, even though he never agreed to it and may not even know it's happening. The 78% of the network on other pools keeps confirming the filtered transactions, so censorship fails. But if enough other pools face similar legal pressure and the filtering coalition grows past 50%, the outcome flips.
This is why pool selection by individual miners is not a fee-optimization decision. It is a governance decision, full stop. Miners who point their ASICs at a single dominant pool because it pays 0.1% more in fees are, in aggregate, trading the network's structural resilience for a rounding error on their monthly revenue. The incentive misalignment here is real, persistent, and no one talks about it enough.
The Stratum Protocol and Solo Mining's Slow Comeback
One mechanism that partially offsets pool concentration is the Stratum V2 protocol upgrade, which shifts transaction selection rights from pool operators back to individual miners. Under the original Stratum (V1), the pool operator assembles the block template, choosing which transactions go in, while the miner just provides hash power.
Stratum V2 introduces a mode where miners build their own block templates and submit them upstream. A pool operator running V2 with this feature enabled can no longer unilaterally censor transactions, because the miner, not the pool, is choosing what to include. The pool handles payout accounting. That's it.
Adoption is uneven. Some major pools support it; others don't. And even where it's technically available, many miners haven't configured it, which is a little like installing a deadbolt and leaving the door open. Still, the architecture exists, and it represents a genuine structural improvement to the censorship-resistance picture that doesn't show up in any hash rate pie chart.
Solo mining, practically dead for individual operators due to variance, has found a partial revival through solo mining pools: setups where the miner does their own transaction selection but joins a variance-smoothing arrangement for payouts. Small in aggregate hash rate terms. Worth watching as an ideological counterweight.
What People Get Wrong About This
The most common misconception: that mining pool concentration is a Bitcoin flaw that needs fixing. It isn't, quite. It's a tension inherent to any proof-of-work system where economies of scale in hardware and electricity push toward aggregation. The protocol didn't fail. Rational economic behavior produced a predictable outcome, as it usually does.
The folk remedy that needs to die is the idea that watching the 51% threshold is sufficient due diligence. Censorship risk accumulates well below that line, through coordination, jurisdiction overlap, and infrastructure dependency. A network where six pools hold 95% of hash rate but none individually exceeds 20% is not obviously safe. It is a six-lock door where three of the locks are owned by the same landlord.
So here's the real question: when you look at the current pool distribution, are you counting independent operators or just counting logos?
If eight or more meaningfully independent operators share the top 80% of hash rate, with no single jurisdiction dominating, that's a network in reasonable health. If you're counting three or four, that's a network relying on the goodwill of a very small number of humans.
Goodwill is not a consensus mechanism.