but - as long as it does not last too long to avoid sending their diff skyrocketing a farm could 'corner the market' so to speak on mining rewards from said coins for a week or so no?
It is a possible scenario, but as Kano mentioned, it should have a little impact, what some people seem to suggest is that miners will shutdown to keep difficulty low, let's imagine a scenario where ALL miners act the same and agree on the same logic.
Let's say we start the year with the following:
Scenario AEpoch #1:
Difficulty at 10t, there are then ten guys mining BTC, each with 1 miner, so a total of 10 gears, each gear gets 1/10 of the total reward
So after 2,016 and 2 weeks, every miner gets 201.6 blocks sent to his wallet.
Now they decide they want to make 201.6 blocks each but in a period shorter than 2 weeks.
epoch #2:
Those 10 guys buy an extra miner each
so
Diff at 10t, 20 miners, each miner gets 1/20 of the total reward, and each person still gets 1/10 ( takes only 1 week)
Epoch #3:
Diff at 20, 20 miners, each gets 1/20 of the total reward, and each person still gets 1/10 (takes 2 weeks) > back to square zero, but they are now burning double the power at epoch number 1.
So total mined coins will be 2,016*3 in 2+1+2 weeks, or 6,048 in 5 weeks.
Scenario B:Epoch #1:
Difficulty at 10t, there are then ten guys mining BTC, each with 1 miner, so a total of 10 gears, each gear gets 1/10 of the total reward
So after 2,016 and 2 weeks, every miner gets 201.6 blocks sent to his wallet.
Epoch #2:
mine the first 1,008 blocks with 20 gears, which will take only 3.5 days, turn off the new 10 gears and mine the remaining 1,008 with 10 gears which will take 1 week, so a total of 10.5 days as opposed to 1 week.
The next diff is going to be a 50% increase in difficulty.
Epoch #3:
diff at 15, 20 miners
mine the first 1,008 blocks with 20 gears, which will take 5.25 days, turn off 5 gears and mine the remaining 1,008 in 1 week with 15 gears
total reward = 6,048 in 2 weeks + 1 week + 3.5 days + 5.25 days + 1 week = 5.25
So 0.25 weeks more time or 5% more time by going with the second method, however, the power consumption difference is a follows
Scenario 1:
Assuming every miner needs 1kwh per day
Epoch 1 = 10 gears = 10khw a day * 2 weeks = 140 KwH total
Epoch 2 = 20 gears = 20 Kwh a day * 1 week = 140 KwH total
Epoch 2 = 20 gears = 20 Kwh a day * 2 week = 280 KwH total
Total power 560 KwH.
Scenario 1:
Epoch 1 = 10 gears = 10khw a day * 2 weeks = 140 KwH total
Epoch 2 = 20 gears for 3.5 days = 70KwH + 10 gears for 1 week = 70 = 140KwH total
Epoch 3 = 20 gears for 5.25 days = 105KwH + 15 gears for 1 week = 105KwH = 210
Total = 490 KwH
which is 12.5% less power.
So more or less, 12.5% saving for 5% less reward
Now assuming all the players here use S19j Pro, and 6 cents per KwH, the miner pays 50% of its profit for the power bill
So if it makes 100$ in 3 epochs with scenario one, it nets only 50$ in profit
in Scenario 2, technically, it makes only 95$ but pays 12.5% less power so only 43.75 for the power which is 51.25$ net profit
50$ vs 51.25$ is 2.43% more profit by playing the diff game.
Keep in mind that this is based on everyone participating in the game and keeping their promise, the chances of anything similar happening is slim to nothing in my point of view.
The incentives need to be more than just direct profit/saving, I would lean more to a theory that suggests, Bitmain wants to grab some coins but not raise the difficulty to keep selling their gears at high prices, that
might sound reasonable to some degree, but them turning off their gear to spare the network 1-2% diff increase, doing that alone is difficult to believe, unless someone thinks that there is really a mutual understanding between 70-80% of all miners to play a game as I explained in scenario 1.