How "mining" works is at the very heart of Bitcoin. It is often brushed over and simply referred to as "complicated math" in the media, but it's actually quite simple to understand even if it is computationally intensive to solve.
Most of the content in this post comes from a post on Reddit that I have edited, reformatted, and elaborated on. Feel free to read the original post if you prefer.
Understanding hashes is the first step in understanding mining. A hash will take an input of any length, and generate is seemingly randomised output of a specific length. The same input will always generate the same output, but changing just one character will drastically change the output. For example,
a948904f2f0f479b8f8197694b30184b0d2ed1c1cd2a1ec0fb85d299a192a447 is the hash of
hello world, and
30e731839774de9ea08ff1adb8aa6b638e05f64900d005f84aea563cab0092b5 is the hash of
This behaviour makes it very difficult to predict what input gives a particular output. For example, what input gives
aaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaa as a hash? It's effectively impossible to work it out. People will often build lookup tables that map these inputs to outputs in order to perform quick reversals later. These tables are called rainbow tables and rely on the input already having been hashed.
The second step is to get the idea of a proof of work. It might be impossible to find a hash specifically with a string consisting of nothing but the letter "a" but what if we asked for a hash with a single zero at the front?
Altering the last letter of
hello world took 26 attempts to finally get
hello worlC which equates to
Why is this useful? Because it creates a puzzle whose difficulty is measurable and which it's impossible to perform better than blind guessing. That second property is important because it's the only way to create a fair "mining" system. Miners solve such puzzles as above but which are far more difficult. For example, find a hash that looks like this:
Each hash is can be considered to be just a number. For example, the hash
00000000000000a05716b3ab0309c2ccc2952c0b3420b4aabb24ff969a320f8c has a numeric value of
So in mining, the miners have to achieve a hash with a numeric value lower than a specified number. This number is called the target. If your hash attempt gives you a number less than the target, which is the same thing as having a bunch of zeros at the front of the hash, then you win and you get to "mine the block". To find such a small hash takes millions of attempts, or more accurately, the whole mining network, with everyone trying at the same time, needs millions of billions of tries to get it right.
The part of the content that they are hashing and are allowed to change, a single number, in order to try and get a hash beginning with zeros, is called the nonce.
The current block reward of 25 Bitcoins is given to the miner who successfully "mines the block" (finds the appropriate hash). It's not really that mining "generates" the Bitcoin in any sense, it's just that it's written into Bitcoin code that a transaction block starts with a unique transaction called a "coinbase" transaction, which is the only type of transaction with no inputs. It only has an output, consisting of the reward plus the transaction fees.
Byzantine Generals problem
To make any sense of Bitcoin's solution to this problem, you need to understand also what is meant by "distributed timestamp server" and how proof of work hashes can be used to construct this. It is, very briefly, explained in Sections 3 and 4 of the bitcoin whitepaper. You're creating a sequence of blocks, tied to each other by including the hash of the last one in the next one. This proves that the next block knew about the last block (remember, hashes are totally unpredictable), which proves that it came afterwards. However, that's not enough; you might know that block 8 comes after block 7, but what if a different block 8, put in by a different miner, also comes after block 7? Worse still, what if these two competing blocks, 8a and 8b contain different transactions, spending money to different places? Which one is the "true" block of transactions? The reason miners did the complicated proof of work process above is exactly to solve this problem.
In bitcoin, the chain of blocks with the largest total proof of work embedded in it is the "winner".
The reason this is such a good way of deciding is that it makes it incredibly difficult for an attacker (someone, say, who wants to spend the same Bitcoins twice) to create an alternative single block or chain of blocks and try to convince everyone else on the network that theirs is the right one. To be valid, yours would have to have more "proof of work" in it (a lower hash value and/or more subsequent blocks). Since everyone else is working on the "true" chain, they have an enormous amount of CPU power working together to create it. To beat them, you're going to have to have more CPU power than everyone else, hence the "51% attack".
Proof of work makes sense because work cannot be faked.
Lastly, here is Satoshi's explanation of the Byzantine Generals' problem. Hopefully you can see how it connects.
The math problem that these mining computers solve serves no purpose other than to secure Bitcoin's network from attackers wishing to "double spend". Miners are not creating a massive rainbow table or computing the human genome. As more computers are thrown at the problem, and hardware advances, the problem is artificially made more difficult to compensate. This seems incredibly wasteful to me as we start to read about the electrical costs of the Bitcoin network and think about the fact that Bitcoin could easily run on just 3 computers to be considered distributed. This is why I have high hopes for alternative cryptocurrencies, such as Peercoin, that implement proof-of-stake. This will allow us to enjoy the benefits that a cryptocurrency provides, but be able to run the network securely on fewer devices, and not hammering their CPU/electricity whilst doing so. The network could run on multi-purpose devices, such as people's phones and tablets rather than purpose-built and costly ASICs that will be redundant in a few years.