Quote from: raid_n on November 28, 2014, 11:12:57 AM

Quote from: dinofelis on November 28, 2014, 11:01:55 AM

The *average miner* now still mines 1/10 (8 at 1/8 and 2 at 0), but spends only 1.6 (8 at 2, and 2 at 0).

So the average miner mines the same, and spends less.

Here is your conceptual mistake. Using an arithmetic mean makes no sense here because you want to look at the individual miner and you do not want to include miners that are not mining)
Bob and Alice are still mining by spending 2 resources so they spend exactly the same as before but receive more.
Satoshi is no longer mining so he is spending 0 resources and receives no more coins (not that he needs them ;-) )

You are basically taking satoshi and adding him to your mining calculation which logically makes no sense.

[edit] do you see the problem in your assumption? I can add an infinite set of miners that do not mine (does not affect mining at all) and it would make your average miner spend close to 0...

This becomes nitpicking, but OK.

Consider the fixed group of people who will mine "before" and "after" the difficulty drop. Let X be a randomly chosen miner in that group.

For some choice X (say, Alice), the production will maybe increase, but for another choice of X (say, Bob), the production has to decrease. That was my point. The extreme case is when Bob stops mining all together, and for that special case, you eliminate him from the group. But it could just as well be that Bob switches off 90% of his mining equipment. Then he's still mining, but much less. The case "he switches off 100%" is simply an extreme case of "he switches off part of his mining equipment".

Somebody has to switch of mining equipment to have the difficulty go down, otherwise it wouldn't go down.