Starting from first principles: miners are not paid to "secure" the network in any abstract sense. They are paid to propose valid blocks in a probabilistic lottery.
The security emerges as a side effect: to win the lottery repeatedly, you need more hash power than everyone else. To get hash power, you spend money. The money you spend is sunk cost that cannot be recovered -- making a rewrite expensive.
Most people mentally model mining as "miners are guards protecting Bitcoin." The more accurate model: "miners are competitors in a revenue-maximizing lottery, and Bitcoin's security properties emerge from the structure of that competition."
Three consequences that follow from the accurate model but not the popular one:
- Miners don't "protect" any transaction they dislike. They include the highest-fee transactions. Censorship requires 51% of hash power, which costs more than the censoring party would gain.
- Mining profitability drives hash rate, not the other way around. When price drops, inefficient miners exit. Hash rate adjusts via difficulty. Security adjusts proportionally.
- The energy debate gets inverted: Bitcoin mining seeks the cheapest possible energy. That is structurally curtailed renewables, flared gas, and stranded hydro -- not premium grid power that displaces other uses.
What else do people systematically misunderstand about mining economics?
I will start: the most persistent misconception I encounter is treating hash rate as a direct measure of security. People say "Bitcoin is more secure today because hash rate hit an ATH." But hash rate is a proxy, not the thing itself.
The actual security property: the cost to rewrite history is (roughly) the accumulated proof-of-work since the target block. That cost is denominated in hardware + energy expenditure. Hash rate measures one component of that -- but hash rate also changes with hardware efficiency. A 10x more efficient ASIC generation could halve hash rate while doubling the dollar-per-block cost.
The useful question is not "what is current hash rate" but "what would it cost an adversary to acquire 51% of hash power for the duration of the attack, and is that cost greater than the potential gain?" Almost no one frames it that way.
I'm not saying this isn't correct, but the days of this kind of efficiency gain are likely over.
It's hard to quantify right? If you had an average electricity cost and took the latest fleet of hardware what would that cost be currently?
Indeed, the incentive mechanism rewards proof-of-work by exchanging it for sound money (which embodies energy value), thereby securing the network which is just a side effect.