How Bitcoin Mining Pools Work

A mining pool bundles the computing power of many miners and splits every block it finds according to their share. So the pool can measure your share, your device keeps sending "shares": hits at a much lower difficulty than a real block. How much you get for them is set by the payout scheme, such as PPS, FPPS or PPLNS. The block itself is usually built by the pool operator, not by you, and that is exactly the weak spot.

Why mining pools exist

In bitcoin mining every device tries numbers until one of them produces a valid block. Whoever hits gets the reward, since the 2024 halving 3.125 new bitcoin plus the fees in the block. How this works in detail is explained in the overview Bitcoin mining explained.

The problem: a single device's chance is tiny. On Nov 13, 2026 mempool.space estimated the network hashrate at about 930 EH/s. A current ASIC with 200 TH/s has about 0.00002% of that. On its own it would find a block on average every 88 years. If you pay for power every month, you cannot wait 88 years for a hit that might just as well come in 200 years.

A pool solves this. Thousands of miners work on the same blocks together. When any device in the pool finds a block, the reward is split by computing power. The pool finds blocks often, so every participant gets something often, just not much.

Solo or pool: the wait for a block

The difference becomes clear when you do the maths for a home miner. Your chance at the next block equals your share of the hashrate. About 52’560 blocks are found per year.

As of Nov 13, 2026, hashrate 930 EH/s according to mempool.space, reward on average 3.149 bitcoin per block (3.125 new plus 0.024 fees).

Solo has the same expected value, but almost nobody ever experiences it. If the lottery ticket still appeals to you, the article Solo mining: the Bitcoin lottery runs the numbers honestly.

Shares: how the pool measures your work

A pool cannot look inside your device. Yet it has to know how much you computed, otherwise anyone could claim a high output. The solution is shares.

A real block needs a hash below a very low target. The pool additionally gives your device a much easier target. Every hash that meets this easy target is a share. Your device sends it to the pool. A share is not a block, but it proves that on average you made a certain number of attempts to find it.

Because every attempt is random, this is fair. Whoever computes twice as much finds twice as many shares. And now and then one of the shares is so good that it also meets the network's target. Then it is a real block, and the pool sends it to the network straight away.

The pool sets the share difficulty for each device separately, usually automatically ("vardiff"). The aim is for each device to deliver a share every few seconds: often enough to measure its output precisely, rarely enough that the pool is not flooded with messages.

How the network difficulty is readjusted every 2’016 blocks is explained in the article Difficulty adjustment.

The payout schemes: PPS, FPPS and PPLNS

The pool counts your shares. How it converts them into bitcoin depends on the payout scheme. The three most important:

PPS (pay per share). You get a fixed amount for every share, whether or not the pool is finding blocks at the moment. The amount equals the share's expected value: the fraction of a block a share is worth on average, times the block subsidy of 3.125 bitcoin. If the pool has a run of bad luck, it pays anyway. The operator carries the risk.

FPPS (full pay per share). Like PPS, but including transaction fees. The pool estimates how much in fees a block brings on average and pays your share of that too. When network fees are high, you get more. Many large pools settle by FPPS today.

PPLNS (pay per last N shares). You only get paid when the pool finds a block. The reward is then split across the last N shares before the hit, i.e. among everyone who was recently computing. If the pool finds no block for two days, you get nothing for two days. If it finds many, you get more than expected. This costs the pool operator no risk, so the fee is usually lower.

The "last N" has a reason: it prevents pool hopping. With simple older schemes it paid to compute only shortly after a found block and then switch. With PPLNS that brings nothing, because every share is worth the same no matter when in the cycle it came.

There are also variants. PPS+ pays the subsidy per share and the fees by PPLNS. The pool OCEAN uses TIDES, a PPLNS-like scheme in which the payout sits directly in the block's reward transaction.

Worked example: an ASIC with 200 TH/s, one day

For most home miners what matters most is the fee and a low minimum payout. With a large pool the fluctuation of PPLNS is hardly noticeable after a few weeks.

Pool fees and payouts

Pools usually charge between 1 and 4% of the reward, depending on the scheme. There are also pools without a fee, for example for solo mining or as promotion. Two more points are easily overlooked:

• Minimum payout. Many pools collect your balance until it reaches a threshold and then pay out to your address. With a mini miner and 59 sats a day that can take a very long time.
• Balance held by the pool. Until payout your earnings sit with the pool, not in your wallet. That is a relationship of trust, as with an exchange. Pools that pay participants directly from the block avoid this, but usually have higher minimum amounts.

Who builds the block? The role of the pool operator

So far a pool sounded like a mere accounting office. But it does more. Before your device can compute, it needs a template for the block: which transactions go in, which previous block it builds on, and who receives the reward. This template is called the block template.

With the most widely used protocol today, Stratum V1, the pool operator builds the template alone. Your device only receives a finished block header to try out. It does not know which transactions are in the block and cannot influence them.

That means: it is not thousands of miners who decide what goes into blocks, but a handful of pool operators.

How concentrated is it?

mempool.space attributes every block to the pool that found it. In the 7 days up to Nov 12, 2026 it looked like this:

Source: mempool.space, 887 blocks from Nov 5 to 12, 2026. The shares fluctuate from week to week.

Five pools thus built a good three quarters of all blocks. Researchers have also published indications that some smaller pools take over the templates of a large pool. Then the real concentration is even higher than the names suggest.

Why that is a problem

A pool operator who builds many blocks could:

• Censor transactions: Simply not include certain addresses or transaction types, for example under pressure from an authority. A single act of censorship only delays, because other pools include the transaction. But if the large pools fall under the same rules together, for example because they sit in the same country, it becomes serious.
• Collude with others: Pools with more than 50% of the computing power together could ignore new blocks from other miners and rewrite recent history.
• Favour themselves: Include their own transactions for a fee outside the mempool, which disadvantages smaller pools.

What matters is what a pool cannot do: it cannot take bitcoin from other people's wallets, cannot insert invalid transactions and cannot change the 21 million limit. Every own node would reject such a block.

What helps

The computing power does not belong to the pool but to the miners. If a pool misbehaves, they can switch to another within minutes. That has happened before: in 2014 the pool GHash.io briefly reached about half of the hashrate, and miners then left it.

What helps more durably is miners building the template themselves:

• Stratum V2 lets the miner choose the transactions with their own node, while the pool only handles the accounting. How that works is explained in the article Stratum V2: the new mining protocol.
• DATUM, the protocol of the pool OCEAN, pursues the same goal: the miner builds the block, the pool splits the reward.
• Solo mining with your own node is the most radical form: you build everything yourself, but carry the full variance.

How to choose a pool

If you want to mine with your own device, these questions help:

1. Which scheme suits you? FPPS for fixed, predictable payments. PPLNS if a lower fee matters more to you than calm days.
2. How high is the fee really? Compare the fee together with the scheme. 2.5% with FPPS and 1% with PPLNS are not directly comparable.
3. When does the pool pay out? With small devices the minimum payout decides whether you ever see anything in your wallet.
4. Who builds the block? Does the pool support Stratum V2 or another method that lets you build the template yourself? And is it already very large? A mid-sized pool is healthier for the network.