Bitcoin Mining for Beginners: A Home Miner's Guide

What Bitcoin mining actually does, and what a home miner should understand before buying anything: proof of work, the block reward, difficulty, hashrate, pools, shares and honest solo odds. Figures are dated 27 October 2026, with live numbers on SoloLuck's calculator, and the guide ends with the scams to avoid and a short glossary.

Category: Mining and the Network · Published 2026-10-28

Editorial review complete · technical review pending · sources under review · time-sensitive: revisit periodically

What Bitcoin mining is

Bitcoin mining is the process of adding the next block of transactions to Bitcoin's blockchain by doing a great deal of trial-and-error computing, known as proof of work. Mining machines hash a candidate block over and over, looking for a result at or below a number the whole network agrees on. A miner that finds one gets its block accepted and collects the block reward: newly created bitcoin plus the fees paid by the transactions inside.¹²

Difficulty, hashrate, pools, shares and "best share" are all details of that one race. This guide takes them in the order they matter, with real figures dated 27 October 2026, so that spec sheets, pool dashboards and sales pitches make sense before you spend anything.

It will not recommend a machine or a pool, and it will not pretend that home mining pays. At typical household electricity prices most home setups cost more to run than they are expected to earn, and a solo miner's chance of a block is a lottery ticket. SoloLuck's mining calculator keeps the numbers current.

What a miner actually does

To build a block, a miner gathers waiting transactions, puts a special first transaction at the top that pays the reward (the coinbase transaction) and summarises the lot in a short block header.² Then it hashes the header. A cryptographic hash turns any input into a number that looks random: change any detail and the result is completely different, with no way to steer it.²

The network sets a target, and a header only counts if its hash comes out at or below it.² So the work is a loop: change a field called the nonce, hash, compare, repeat. The whitepaper describes exactly this, noting that the average work grows exponentially with the number of leading zero bits required, while anyone can check a result with a single hash.¹ An Antminer S21 runs the loop about 200 trillion times a second.³

A miner that finds a winning hash broadcasts the block. Other nodes accept it only if every transaction in it is valid, and they show their acceptance by building the next block on top.¹ If two valid blocks appear at almost the same moment, the network follows the chain with the most proof of work behind it, and the other block becomes stale.² Rewriting an old block would mean redoing its proof of work and that of every block after it, which is what makes the history so costly to change.¹ How proof of work secures Bitcoin goes further.

One idea here matters for everything that follows: each hash is a fresh draw. A hash that lands close to the target does not make the next one likelier to succeed, and a machine builds up no credit over hours or years of trying.²

The block reward: subsidy plus fees

The block reward has two parts: the block subsidy, new bitcoin created by the protocol, and the transaction fees paid by the transactions in the block. The coinbase transaction collects both.² Bitcoin Core rejects any block whose coinbase pays out more than the two combined.⁴

The subsidy started at 50 BTC and halves every 210,000 blocks, roughly every four years.⁴ The fourth cut came at block 840,000 on 20 April 2024, and the subsidy has been 3.125 BTC since the April 2024 halving.⁵ Fees move with demand for block space, so totals vary: the October 2026 blocks in SoloLuck's Solo Block Record paid about 3.13 to 3.15 BTC each.⁶ The next halving, at block 1,050,000, cuts the subsidy to 1.5625 BTC. From block 973,413 on 27 October 2026 that is about 76,600 blocks, roughly a year and a half (worked out by us), so any longer plan has to assume half the subsidy.

There is also a wait. Coins created by a coinbase transaction cannot be spent for 100 blocks, a rule Bitcoin Core calls coinbase maturity.²⁷ At ten minutes a block that is about 16 to 17 hours on average (worked out by us). The rule exists because a block can still lose a race and go stale, and its coinbase transaction is then destroyed with it.² Nobody can shorten the wait, whatever they offer. What happens when a miner finds a block walks through that moment.

Difficulty and the two-week adjustment

Difficulty is the target written as a friendlier number: how hard it is to find a hash below the current target, compared with a fixed reference called difficulty 1.⁸ Double the difficulty and the target halves, so a block needs twice as many hashes on average.²

Every 2,016 blocks, the network uses the timestamps in the block headers to measure how long those blocks took. The ideal is 1,209,600 seconds, exactly two weeks, or one block every ten minutes. If the blocks came faster, difficulty rises in proportion, by up to 300%; if they came slower, it falls, by up to 75%.² Nobody votes on it. Every node does the same arithmetic, which keeps Bitcoin close to a ten-minute rhythm as miners switch on and off.¹

As of 27 October 2026 the network difficulty was 140.58 T, about 140.58 trillion.⁹ One unit of difficulty corresponds to about 2³² hashes, close to 4.3 billion, so difficulty also measures the network's hashing: network hashrate ≈ difficulty × 2³² ÷ 600 seconds.⁸ At 140.58 T a block takes about 6 × 10²³ hashes on average, and the network comes to roughly 1,006 EH/s, just over one zettahash per second (worked out by us).

Difficulty moves both ways: SoloLuck's Solo Block Record notes 138.96 T in May 2026, 127.45 T for a block on 10 September and 132.72 T for one on 11 October.⁶ Your hashrate stays the same, so when the network grows your slice of it shrinks and every expected-time figure gets longer. That is why every number on this page carries a date.

Hashrate, watts and efficiency

Hashrate is hashes per second, written with the ordinary metric prefixes: tera (T) is 10¹², a trillion; peta (P) is 10¹⁵; exa (E) is 10¹⁸; zetta (Z) is 10²¹.¹⁰ Home machines are rated in TH/s, while the network's roughly 1,006 EH/s on 27 October 2026 is about a billion TH/s.

Power is the other figure that matters. Watts divided by TH/s gives a machine's efficiency in joules per terahash (J/TH), the energy it spends on each trillion hashes, and lower is better. The rated figures on SoloLuck's hardware pages show the range:³⁹

An S21 spends about half the energy per hash of an S19 and roughly a sixth of an S9's, which is why older machines that still hash perfectly well can cost more to run than they bring in. Each model's page, such as the Antminer S21 or the Bitaxe, shows its odds and running cost at today's difficulty.

Pools: shared rewards or solo

You can mine alone or with others. Solo mining means finding blocks on your own, with the whole reward going to you when you do: large payments, far apart. Pooled mining combines your hashing with other miners' so the group finds blocks more often, and shares the proceeds roughly in proportion to each miner's work: small payments, close together.¹¹

A pool cannot see your machine, so it counts your work in shares. It sets its own target, a few orders of magnitude easier than the network's, and your miner sends in every hash that beats it; each one proves you did a share of the work.¹¹ In a shared-reward pool, shares are what you are paid for. Under pay-per-share (PPS) the pool pays a flat amount for every accepted share straight away, carries the bad luck itself and covers that risk by paying less than the full expected value; full pay-per-share (FPPS) also passes on part of the transaction fees.¹²

A solo pool runs the node and the pool software for you but shares nothing. Your miner still sends shares, which let the pool see that it is working and estimate its hashrate, but they earn nothing. If one of your hashes also beats the network's target, the block pays you, minus any fee the pool takes.¹³ Rules vary: on some pools called solo, the finder keeps only part of a block and the rest is shared.⁶ Read a pool's own payout rules first.

Before fees, both routes have the same long-run average; only the shape differs. As of 27 October 2026 an Antminer S21 was expected to earn about $7.21 a day before electricity, at a bitcoin price of about US$80,400. A shared-reward pool pays that out as a steady trickle, minus its cut. Solo, it arrives as nothing on almost every day and, with a chance of about 1 in 35,000 on any given day, as a whole block of roughly 3.13 BTC.¹⁴

Neither route changes the odds of a single hash. What differs is the fee, the payout rules, how often you are paid and how much work goes stale while new jobs travel to your miner, which usually talks to the pool through a widely used protocol called Stratum.¹¹ To pay you, a pool needs an address, never a key. Wallets receive with public keys and spend with the matching private keys, so no pool ever needs your private key or recovery phrase.¹⁵

Shares and your best share

Every share has a difficulty of its own, measured against the same difficulty-1 reference as the network: the lower the hash, the higher the number.⁸ Pools set a minimum share difficulty so that machines report in at a sensible rate. At a share difficulty of 1,000, a 1.2 TH/s board would send a share about every 3.6 seconds on average (worked out by us with the same 2³² rule).

Your miner may also show the highest share difficulty it has reached. The Bitaxe's AxeOS dashboard calls it Best Difficulty and shows an all-time best and a best for the current session.¹⁶ Miners call it their best share, and a big jump is a real moment of luck: one hash came out unusually low.

It is not progress. A best share of, say, 2 G (two billion) is about 70,000 times short of the 140.58 T a block needed on 27 October 2026 (worked out by us), and the next hash knows nothing about it. A near miss is not saved as progress and does not shorten the wait: a machine that has just set a record is exactly as likely to find a block in the next minute as an identical one switched on a minute ago.²

The flip side is the exciting part. If a share ever comes out at or above the network difficulty, it is not just a share but a block, and each win in SoloLuck's Solo Block Record began exactly that way.⁶

The odds, honestly

The expected time for a machine to find a block on its own is expected seconds = difficulty × 2³² ÷ hashrate, with hashrate in hashes per second.⁸ At the 27 October 2026 difficulty of 140.58 T it gives these averages, with each daily chance from the same formula:⁹³

An expected time is an average over a very long run. It is not a countdown. Every hash is a fresh draw, so the chance of a block on any one day is the same today as on the day the "expected time" runs out. The chance of at least one block within a period t is 1 − e^(−t/T), where T is the expected time.⁹ For an S21, if difficulty and hashrate stayed put (they will not), that is about 1% in a year, and after the full 96 years there would still be about a 37% chance of no block at all (worked out by us). Equally, a block can come on the first day.

That is how small machines sometimes win. As of 27 October 2026 SoloLuck's Solo Block Record lists 44 solo-mined Bitcoin blocks since 2022, found mostly by small and home miners on public solo pools and each checked on the chain. They include a single Bitaxe Gamma (block 957,382, 10 July 2026), a NerdQAxe++ (945,601, 18 April 2026), an Avalon Nano 3S (951,771, 31 May 2026) and an Avalon Mini 3 (913,593, 7 September 2025).⁶ Each was a long shot that came in, and none makes the next one likelier.

A crowd shows the pattern. At the Solo Mining Census reading of 27 October 2026 (15:07 UTC), 17,206 Bitaxes were connected to one public solo pool, Public Pool, averaging 1.06 TH/s each.¹³ Together that is about 18.2 PH/s, enough for the group to expect a block roughly once every 383 days (worked out by us). When one comes, it belongs to the single owner whose board found it; the other 17,205 get nothing, and their odds on the next block are unchanged.

So call solo mining what it is: a lottery ticket paid for in electricity. It is not income, and no hashrate makes a block certain. No setting improves the odds of a hash, whether a worker name, a port or a pool that is nearer or faster; distance can only waste a little work as stale. The solo mining odds calculator shows your own numbers, and how solo-mining probability works goes deeper into the maths.

Electricity, heat and noise

For nearly every home miner, electricity decides the money. As of 27 October 2026 an Antminer S21 was expected to earn about $7.21 a day before electricity, and it uses 84 kWh a day. At $0.10 per kWh that power costs $8.40, a loss of about $1.19 a day; at $0.05 per kWh it would come out about $3.01 ahead.¹⁴ Older machines do worse: on the same figures an Antminer S19 earns about $3.42 a day against $7.80 of power at $0.10 per kWh (worked out by us).

Household prices are another matter. The US average residential price was 18.16 cents per kWh in August 2026.¹⁷ Great Britain's price cap for 1 October to 31 December 2026 works out at an average of 26.32p per kWh on a standard variable tariff paid by Direct Debit.¹⁸ German households paid an average of 37.97 euro cents per kWh in the first half of 2026, taxes and levies included.¹⁹ At those prices an S21's 84 kWh a day costs about $15.25, £22.11 or €31.89 (worked out by us). In the US that is more than twice what the machine was expected to earn.

Desk boards turn this around: tiny bills, tiny odds. A NerdQAxe++ at about 76 W uses under 2 kWh a day, about $121 a year at the US average price, against expected earnings of about $0.17 a day, or $62 a year, at the 27 October 2026 difficulty and price (worked out by us). It makes sense as a hobby or a lottery ticket, not as a way to earn.

Power is also a wiring and comfort question. An S21 draws 3.5 kW from a 220–277 V supply on a single 20 A input, which in most homes means a dedicated circuit, and turns all of it into heat around the clock. Bitmain publishes no noise figure for it; its sibling, the S21 Pro, is rated at 76 dBA with the fans at full speed.¹⁴ Before buying anything, put your own tariff into the calculator.

Scams and red flags

Mining is easy to fake from the outside: you cannot see someone else's data centre, and a dashboard can show any number. In 2025 the FBI's Internet Crime Complaint Center logged 181,565 complaints involving cryptocurrency, with reported losses of $11.37 billion; cryptocurrency investment fraud alone accounted for $7.2 billion.²⁰ These are the patterns to know:

• Cloud-mining contracts with fixed daily returns. You pay for hashing you cannot see and watch a balance grow online. HashFlare sold contracts for a share of mined cryptocurrency from 2015 to 2019, with sales of more than $577 million; according to the US Department of Justice, its dashboard showed falsified data and it lacked the computing capacity for most of the mining it claimed.²¹
• Mining apps. In 2021 security researchers found more than 170 Android apps selling cloud mining to an estimated 93,000-plus victims, taking more than $350,000, and no mining took place at all.²² Treat any app that claims your phone is mining bitcoin the same way.
• A balance you cannot withdraw. The FBI describes fake platforms that show an investment growing, then demand taxes or extra fees before anything can be cashed out.²³ A real block reward never needs a release fee.
• Requests for your recovery phrase or private key. A pool, a miner or a support desk only ever needs an address to pay you.¹⁵ Anyone asking for your wallet's backup words wants what is in it; see how to protect a recovery phrase.
• Offers to speed up a reward. The 100-block maturity is a network rule, and nobody can shorten it for a fee.⁷
• Claims to raise your odds per hash. Special firmware, "lucky" worker names, secret ports or a pool that says it finds more blocks per terahash: whether a hash wins depends on the network target and nothing else.²
• Pressure, and numbers that sound too good. The UK's Financial Conduct Authority lists contact out of the blue, pressure to invest quickly and promised returns that sound unrealistic as warning signs, and warns of follow-up scams that offer to get your money back for a fee.²⁴

The safest habits are dull ones: buy hardware from sellers you can verify, keep coins in a wallet you control, run your own numbers in the calculator, and assume that anyone selling a fixed return from mining is selling the promise, not the mining. How to protect yourself from cryptocurrency scams covers what to do if you have been targeted.

A short glossary

• Address: where bitcoin is sent, made from a public key; safe to share.
• ASIC: a chip made for one job. A Bitcoin ASIC does only the hashing that mining needs.
• Best share (best difficulty): the highest share difficulty your miner has reached. Luck, not progress.
• Block reward: the subsidy plus the block's fees, paid by the coinbase transaction.
• Coinbase transaction: the first transaction in every block, which creates the new coins (no connection with the exchange of the same name).
• Difficulty: how hard the current target is compared with difficulty 1; it adjusts every 2,016 blocks.
• Expected time: the average wait for a block, difficulty × 2³² ÷ hashrate. Not a countdown.
• Halving: the cut in the subsidy every 210,000 blocks.
• Hashrate: hashes per second, in TH/s for home machines and EH/s for the network.
• J/TH: joules per terahash, watts divided by TH/s; lower is better.
• Nonce: a header field that a miner changes between attempts.
• Pool: a service that coordinates miners. Shared-reward pools pay by shares; solo pools pay a found block to its finder.
• PPS and FPPS: shared-reward payout methods that pay a fixed amount per accepted share; FPPS adds part of the fees.
• Share: a hash that beats the pool's easier target, proving your miner did some work.
• Stale block: a valid block that lost a race; its reward never becomes spendable.
• Stratum: the widely used protocol miners use to talk to pools.
• Subsidy: the new bitcoin created with each block, 3.125 BTC since the April 2024 halving.
• Target: the number a block's hash must not exceed.

Key takeaways

• Mining is a race to find a block hash at or below the network's target, and the winning block pays the block reward: the subsidy (3.125 BTC since the April 2024 halving) plus fees.
• Difficulty readjusts every 2,016 blocks, about two weeks, to keep blocks near ten minutes apart; it stood at 140.58 T on 27 October 2026.
• Pools change how you are paid, not the odds of any hash: shared-reward pools smooth payouts, while solo mining pays a whole block or nothing.
• A share, or a high best share, is not progress, and an expected time to a block is an average, not a countdown.
• A block reward cannot be spent for 100 blocks, and anyone selling fixed mining returns or a faster payout is a red flag.
• Electricity usually decides the money: at household prices most home setups cost more to run than they are expected to earn.

Sources

1. Bitcoin: A Peer-to-Peer Electronic Cash System (the Bitcoin whitepaper): proof of work and the nonce, the steps nodes follow, and the block incentive — bitcoin.org
2. Bitcoin Developer Guide, Block Chain: proof of work and the target threshold, difficulty retargeting every 2,016 blocks (ideal 1,209,600 seconds, up to +300% / -75%), the coinbase transaction (block subsidy plus fees), the 100-block spending rule and stale blocks — Bitcoin Project (developer.bitcoin.org)
3. Mining hardware catalog: rated hashrate, power and J/TH for each model, taken from the makers' and designers' own pages — SoloLuck
4. Bitcoin Core source, validation.cpp: GetBlockSubsidy (50 BTC, cut in half every 210,000 blocks, about every four years) and the check that rejects a coinbase paying more than subsidy plus fees — Bitcoin Core project (GitHub)
5. Controlled supply: the halving schedule; block 840,000 reached on 2024-04-20 with a 3.125 BTC subsidy — Bitcoin Wiki
6. Solo Block Record: 44 solo-mined Bitcoin blocks since 2022, each verified on-chain, with rewards and the difficulty at each block (as of 27 Oct 2026) — SoloLuck
7. Bitcoin Core source, consensus.h: COINBASE_MATURITY = 100 (coinbase outputs can only be spent after this number of new blocks, a network rule) — Bitcoin Core project (GitHub)
8. Difficulty: definition against difficulty 1, adjustment every 2,016 blocks, network hashrate = D x 2^32 / 600, and average time to a block = difficulty x 2^32 / hashrate — Bitcoin Wiki
9. Solo mining odds calculator: network difficulty 140.58 T on 27 Oct 2026, expected time to a block by miner, and P(at least one block in t) = 1 - e^(-t/T) — SoloLuck
10. Metric (SI) prefixes: tera 10^12, peta 10^15, exa 10^18, zetta 10^21 — National Institute of Standards and Technology (NIST)
11. Bitcoin Developer Guide, Mining: solo mining and pool mining, payout variance, shares against an easier pool target, and the Stratum protocol — Bitcoin Project (developer.bitcoin.org)
12. Pooled mining: the pay-per-share (PPS) and full pay-per-share (FPPS) payout approaches — Bitcoin Wiki
13. Solo Mining Census: hourly readings of the public solo pools, including the fee each pool takes and the Bitaxes connected to Public Pool (reading cited: 27 Oct 2026, 15:07 UTC) — SoloLuck
14. Antminer S21 model page: 200 TH/s at 3,500 W, expected earnings and net at three power prices, 1 in 35,000 daily solo odds, power input and noise notes — SoloLuck
15. Bitcoin Developer Guide, Wallets: wallets create public keys to receive bitcoin and use the matching private keys to spend it — Bitcoin Project (developer.bitcoin.org)
16. ESP-Miner (AxeOS) dashboard source: the 'Best Difficulty' card showing an all-time best and a session best — bitaxeorg/ESP-Miner on GitHub
17. Electric Power Monthly, Table 5.6.A: average price of electricity by end-use sector, August 2026 (US residential average 18.16 cents per kWh; released 23 Oct 2026) — U.S. Energy Information Administration
18. Changes to energy price cap between 1 October and 31 December 2026: electricity 26.32p per kWh by Direct Debit, average across England, Scotland and Wales — Ofgem
19. Press release No. 345 of 30 September 2026: households paid an average of 37.97 cents per kWh for electricity in the first half of 2026 (in German) — Federal Statistical Office of Germany (Destatis)
20. 2025 IC3 Annual Report: 181,565 complaints involving cryptocurrency with $11.366 billion in reported losses; cryptocurrency investment fraud $7.2 billion — FBI Internet Crime Complaint Center (IC3)
21. Two Estonian Nationals Plead Guilty in $577M Cryptocurrency Fraud Scheme: HashFlare mining contracts 2015-2019, falsified dashboard, missing computing capacity (13 Feb 2025) — U.S. Department of Justice
22. Lookout discovers Android crypto mining scams: more than 170 apps, more than 93,000 victims, more than $350,000 taken, no cloud mining performed (July 2021) — Lookout
23. Public Service Announcement I-100322-PSA, Cryptocurrency Investment Schemes: fake balances that appear to grow, then taxes or fees demanded to withdraw (3 Oct 2022) — FBI Internet Crime Complaint Center (IC3)
24. Crypto investment scams: warning signs and follow-up offers to get money back for a fee (last updated 16 Feb 2026) — Financial Conduct Authority (UK)

Some specific claims in this article are scheduled for further source review; citations may be expanded (see our editorial policy).

Related educational reading

• How Solo-Mining Probability Works
• How Proof of Work Secures Bitcoin
• What Happens When a Bitcoin Miner Finds a Block?
• How Mining Pool Hashrate Estimates Work
• How to Protect Yourself From Cryptocurrency Scams
• How to Evaluate Bitcoin Mining Hardware