Bitcoin Forum

If everyone agrees to turn off their miners right before a difficulty change.. they could fake a lower network hash rate!  
In fact I believe this is what has been happening with a few big players..the network hash rate always seems to drop right before a change and this makes mining [temporarily] more profitable.
I don't think this would benefit someone with a small amount of hash power, but a big player like myself might benefit from unplugging a couple hours before the change.  i'll have to simulate this and see if its worth it.   :-* brb k

edit: actually come to think of it, any gains from this technique should be wiped out by a counter affect from the following difficulty change... UNLESS it is pushed down EVERY time.  There would be an accumulation of network hash power that is always hidden from the difficulty changes... could this be why there is an "up and down" trend consistent with the time of the changes?  

I could see how this could cause a problem for the network if everyone decided to manipulate it this way.. We would basically be producing bitcoins at a faster rate than intended by the system.  I guess the only side affect would be that the value of the bitcoin itself would go down (who cares, right?)  

Would never work.  Not only would there be a counter effect like you said, but good luck trying to get other miners to unplug.

You can "manipulate" difficulty lower but it would only cost you.

Lets say 10% of miners (by hashpower) turn of rigs 200 blocks before difficulty period ends.  That will reduce the difficulty post adjustment by roughly 10% * (200/2016) = ~1%.  So yes difficulty will be ~1% lower post adjustment and as a result you will earn about 1% more.  Of course what you are forgetting is that by turning off your rig for 200 blocks you will LOSE 200/2016 = ~10% resulting in a net loss of 9%.

Now for everyone who is good at math and doesn't participate it is just "free money" they mine 100% of the time and difficulty is still 1% lower.  It is merely a wealth transfer from those who do participate to those who don't.  In the example 10% of the miners voluntarily decide to make 9% less in order for the other 90% to make 1% more.

But what if you own 10% of the network?  Then it becomes beneficial to unplug a couple minutes before a change because that would be you're going to make 10% more during the next period.  

You posted this right before i added my last comment..
You even used the same example of 10%... I missed that one fact that 10% <> 10%.  You're right..

edit: what if they turn it off 1 block before it ends?

Such a strategy would be a net loss to the miners who use it. The loss of income is larger than the potential gains of the lowered difficulty.

Also, the size of the difficulty adjustment depends on the time spent on the last 2016 blocks, not the current hashrate at the time of the adjustment. So a temporary dip in hashrate around the last few blocks of a cycle has almost 0 effect.

I think I figured it out.  Miners with new rigs waiting to be deployed would benefit from shutting of some time before difficulty change then when it adjusts BANG turn on old rigs and new rigs.

Does that make sense?

Unplugging a couple minutes (1 block?) will not decrease difficulty by 10%.  The last block will take 10% longer (on average assuming 10% of hashrate is turned off) however the 2016 blocks will only take  0.005% (2016.1/2016) longer in aggregate.  Difficulty is based on the time and thus indirectly the average hashrate, over the entire 2016 blocks.

So yes difficulty will be 0.005% lower in the next period.
Yes you (and everyone else) will make 0.005% over the next 2016 blocks in the next period.
However  to do so it will cost you 1/2016 = 0.05% of your revenue.  
You will lose 10x what you gain.

Yep I get it now.. it was fun to think about though.   :P

It sounds like you're under the impression that the speed is "measured" at the moment of a difficulty change. It is not. It is calculated based on the solving speed of all the blocks since the last difficulty change.  Shutting off 10% of the miners an hour before the difficulty change would have no different effect than shutting them down for any random hour.

And that is one more reason to appreciate the system and how it works.  good job, Satoshi! (if he can hear me)  :P

This can only work if there are less than 20 miners and you are one of the biggest ones. I have done this with an alt named Machinecoin which nobody knew about. I used to stop my miner just before 200 blocks and restart after difficulty dropped.

they already do this ...

https://bitcoinwisdom.com/bitcoin/difficulty

This looks more like testing a veeeeeery huge batch...

I wonder who?      ::)