Mining Bitcoin involves ongoing and unavoidable costs, foremost among which is machine power consumption. These expenses are incurred independently of any results, even though revenues from mining are, by their very nature, rare and random. The discovery of a block depends exclusively on the miner's share of hashrate, which makes earnings all the more unpredictable the smaller that share is. It is precisely this practical problem that rapidly led to the widespread use of mining pools. In this final chapter of the MIN 101 course, I offer an introduction to the principles and operation of mining pools in Bitcoin.

A mining pool is an organization (often an online service) that aggregates the computing power of many independent miners, in order to increase the frequency with which their group finds blocks. When the pool finds a block, the block reward is then redistributed among the participants according to internal pool rules (which will be covered in the MIN 201 course, as they are too complex for MIN 101).

Participants in a mining pool are then often referred to as "hashers", rather than as "miners", as they no longer carry out all the mining work, but simply hash the data transmitted to them by the pool.

Be careful not to confuse the mining pool with the mining farm. These are two different concepts. As we saw in the previous chapter, a farm is a physical site where a single entity operates numerous mining machines. A pool, on the other hand, is above all a virtual grouping: thousands of machines, often geographically dispersed, work under a common coordination. However, a farm can, and often does, participate in a pool.

So why join a pool? Quite simply because the result of mining activity is probabilistic: with each hash attempt, there is a very small chance that it will meet the difficulty target and therefore produce a valid block.

Over the very long term, your average earnings should be proportional to your share of the overall hashrate. This principle follows directly from the laws of probability: each hash calculation constitutes an independent random draw, and, by the law of large numbers, the frequency with which you discover blocks converges mathematically towards your fraction of the network's total hashrate. In the short to medium term, however, your actual earnings can be extremely irregular. It's this discrepancy between theoretical average and random reality that we call variance in mathematics.

Here's a simple example to illustrate the principle:

But this value corresponds only to a mathematical expectation: the distribution of discovery times follows a random law, so it's perfectly possible, in practice, never to discover a single block, even over a very long period. You can be unlucky and find nothing for a very long time, while paying recurring costs (electricity, maintenance, equipment depreciation...).

The mining pool changes the nature of this problem: by combining hashrates, the pool finds blocks more often. Each participant then agrees to receive only a fraction of each block found, but much more frequently. It's a transformation from a highly volatile, widely-spaced income to a more regular one, at the cost of sharing the rewards and paying service fees.

The higher your computing power, the higher your expected frequency of found blocks. More importantly, the more frequent the events, the closer the observed results are to the statistical average over a given period.

On a solo basis, a small-scale miner may go for years without a single block, then get a big payout one day, then nothing. In a pool, the same probabilistic reality still applies, but it's smoothed out at the collective scale: the pool finds blocks more often, and redistribution converts these events into more regular payouts for each participant. The mining pool therefore sells predictability on the mining activity.

As we saw in the previous chapter, at the very beginning, mining could be done solo with a conventional computer, as the difficulty was very low. But as the global hashrate exploded (GPU, then ASIC), solo mining became a very time-consuming gamble for most participants.

The first pools were created precisely in response to this new reality. Braiins Pool (formerly Slush Pool / Bitcoin.cz) is the first Bitcoin mining pool: it mined its first block on December 16, 2010. The success of this first mining pool was rapid, as in just a few days it captured nearly 3.5% of the global hashrate.

On the technical side, pools were then structured around specialized communication protocols between the pool and the miners (e.g. Stratum, then Stratum V2), in order to efficiently orchestrate distributed work. We'll be taking a closer look at these concepts in our MIN 201 training course.

At the time of writing (early 2026), the global Bitcoin hashrate is at an order of magnitude of zetta-hash per second (= 1,000 EH/s = 1,000,000,000,000,000,000,000 hashes per second), and almost all the blocks found come from mining pools.

Here is a ranking, to date, of the main mining pools and their respective share of hashrate. This ranking is likely to change by the time you read this course. For up-to-date data, please visit mempool.space.

In a more advanced course, we'll go further into the internal workings of the pools (shares, network protocols, payment methods...), because this is where the details that determine both the miner's profitability and the potential implications for Bitcoin come into play.

We've now come to the end of MIN 101. Thank you for following it through to the end. If you'd like to assess the skills you've acquired throughout the course, a final exam awaits you in the next section.

With the basic knowledge you've just acquired, you can now take more advanced courses on mining on Plan ₿ Academy, whether in theory, like this one, or more practical courses, so that you too can start participating in Bitcoin mining!