Bitcoin's Basins
The states Bitcoin returns to after a push, how hard a push it takes to stop it returning, and which of them its designer wrote in.
On this site a basin is the set of conditions a system comes back from. Push a marble around the inside of a bowl and it rolls back to the bottom. The bowl is its basin, and the rim is the edge: the push beyond which the marble does not come back. Looking for basins, and for their edges, is the whole point of this laboratory. Bitcoin has four basins worth naming. Two were built on purpose by its designer, Satoshi Nakamoto. Two formed on their own.
| basin | what it returns to | its edge | where it came from |
|---|---|---|---|
| 1. Ten minutes a block | a new block about every ten minutes | losing most of the miners at once | written in |
| 2. One ledger | a single agreed record of payments | anyone with more than half the computing power | written in |
| 3. Miners who can pay their bills | as many miners as the reward can support | the shrinking reward | formed on its own |
| 4. The rules themselves | the 21 million cap and the ten-minute target | a dispute nobody can settle | formed on its own |
1. Ten Minutes a Block
What it returns to. A new block of payments about every ten minutes, whatever happens to the number of machines mining.
What pushes it. Miners joining or leaving. If the computing power doubles, blocks are found twice as fast; if half the miners switch off, blocks take twice as long.
What brings it back. The difficulty rule. Every 2016 blocks, each copy of the ledger looks at how long those blocks took, compares that with the two weeks they should have taken, and makes the lottery harder or easier by the same ratio. In 2021, when China banned mining and about half the computing power went offline over a few weeks, blocks slowed sharply, the difficulty fell by about 28% on 3 July, the largest drop on record, and the pace returned to ten minutes as the miners moved to other countries. The Difficulty Loop panel on the Bitcoin page lets you try this yourself.
Why the tuning matters. A correction that is too strong or too fast does not settle; it swings back and forth. Bitcoin Cash showed this in 2017. It launched with an aggressive emergency rule for cutting difficulty, miners rushed back and forth between it and Bitcoin to catch the cheap blocks, and for months its blocks came in floods and droughts until the rule was replaced. Bitcoin's rule is slow on purpose: two weeks of data, and never more than a factor of four in one step. It recovers slowly and never swings.
The edge. That same slowness. If nearly all the miners left at once, say nine out of ten, blocks would arrive about every hundred minutes, and if it happened just after an adjustment, the next one would be more than four months away. Recovery would still come, but a network that slow might lose more miners while it waited.
The event: The China Mining Ban, 2021 →
2. One Ledger
What it returns to. A single agreed history of payments, the same on every copy.
What pushes it. Two miners winning a block at almost the same moment. For a few minutes the network holds two versions of the newest block, and different computers may be working on different ones. This happens naturally every so often.
What brings it back. One rule that every copy follows: keep the version of the chain that took the most guessing to build, which is normally the longer one. The next block gets built on one of the two versions, that version becomes longer, every copy switches to it, and the payments in the abandoned block go back into the queue for a later block.
The edge. Half the computing power. Anyone who controls more than half of it can build their own version faster than everyone else combined and pull the whole ledger onto it. That is the attack described on the Spending a Coin Twice page.
The event: More Than Half the Computing Power →
3. Miners Who Can Pay Their Bills
What it returns to. As many mining machines as the reward can pay for.
What pushes it. The price of bitcoin. Miners are paid in bitcoin but pay for their machines and electricity in dollars, so when the price falls, mining pays less.
What brings it back. The least efficient miners switch their machines off. With fewer machines the blocks slow down, the next adjustment lowers the difficulty, and the miners who stayed win more often, so each machine earns more again. When the price rises the same thing runs in reverse: mining pays more, more machines switch on, and the difficulty rises to absorb them. Nobody designed this level. It is set by the market, by what machines and electricity cost against what the reward is worth.
The edge. The reward itself. Every four years the new bitcoin paid per block is cut in half, and eventually miners will have to be paid almost entirely by fees, the small amounts people offer miners to include their payments. Nobody knows yet whether fees alone will pay for enough machines to keep the ledger expensive to attack. This is the most important open question about Bitcoin's future.
The event: The Halvings and the Miners' Bills →
4. The Rules Themselves
What it returns to. The original rules: 21 million coins, the halving schedule, and the ten-minute target.
What pushes it. Proposals to change them. Some have been made in good faith, to make payments cheaper or faster; some would have benefited particular groups.
What brings it back. People. Changing a rule only works if most of the people running copies of the ledger switch to the new rule at the same time. Those who understand that this is an experimental platform in the middle of a life-test, and that the rules are brilliantly placed and executed, have declined to switch, and so the rules have held. Nothing in the code enforces this. It is a basin made of people's judgment.
The edge. A dispute that the community cannot settle. In 2017 the argument over how large blocks should be crossed it. Neither side gave way, and the chain split in two: Bitcoin went on with the old rules, and Bitcoin Cash went off as a separate currency with its own basin.
The event: The Block Size War, 2015 to 2017 →
Designed, Then Left Running
The first two basins were built on purpose. The difficulty rule is a thermostat Satoshi wrote in so that ten minutes a block would be the state the network always returns to, and the rule for choosing the longer chain was written in so that one ledger would be. Since he left, every copy of the ledger has run those rules on its own records with nobody operating them. Running a rule is different from producing it, though. A thermostat keeps regulating a house after the installer has gone home, and the house is still an attractlet, because its setting was chosen from outside. Bitcoin's code and rules, taken alone, are an attractlet.
The other two basins were never written down. The level at which miners can pay their bills comes out of the market, and the rules hold because of people's judgment. Those two were produced by the people taking part, and they are where the open question on the Bitcoin page lives: does Bitcoin, together with those people, keep itself going and repair itself?
The Price Has No Basin
The price of bitcoin has no level it returns to. It has fallen by 70 to 80 percent several times and later risen to new highs, but nothing in the system pulls it toward any particular number. What has returned after each crash is the system itself: the blocks, the ledger, the miners and the rules. That is survival, and it is a different thing from a basin.
The Block (3 July 2021). Bitcoin network undergoes largest difficulty drop ever by nearly
28%.
Bitcoin Core source, difficulty retarget (2016 blocks, two-week target, factor-of-four clamp).