Hacker News Reader

Comment by lottin | Hacker News Reader

That doesn't mean it's a dumb metric. Any efficiency metric must measure cost per unit of output and in a crypto-currency that's power per transaction.

View on HN

why not power per dollar transferred? transactions per 10 minute period is relatively constant over time. The thing that increases the energy usage is the price of bitcoin.

It's a dumb metric.

> The thing that increases the energy usage is the price of bitcoin.

This simply tells us that bitcoin has a dumb cost structure. Apparently its inventors didn't think about that either. But that's not our problem. Our problem is measuring efficiency, and for that we need a measure of efficiency. And a measure of efficiency needs to be chosen in terms of how good it is at measuring efficiency, not in terms of how good it is at hiding that a certain payment system has a dumb cost structure.

Would you measure a car's efficiency based on how much wiper fluid it uses per mile? The two measures aren't meaningfully linked, so it makes little sense to compute a correlation.

No, because wiper fluid contributes nothing to the outcome, which is the car moving from point a to b. The car will move from point a to point b whether it has wiper fluid or not. But it will not move without fuel.

Ok, so what about how many rotations the radiator fan does per mile?

The car will overheat and won't reach point B if the fan isn't running, so does it make sense to compare the efficiency of different cars based on how many rpm's their radiator fans do?

No, because the general idea behind efficiency has to do with the costs that are incurred to achieve a particular outcome. For example, a method is said to be more efficient than another if it achieves the same outcome at a lower cost. The outcome in the case of cars is movement because that's the purpose they're built for, so a measure of car efficiency has to reflect how costly is to move the car some unit of distance. The RPMs the radiators fans do per mile don't tell us anything about how costly it is to move the car, therefore, no, it's clearly not a good efficiency measure.

Reply on news.ycombinator.com