Don't expect another 8% Diff increase -- expect more

[grod]

So what you're saying is your model can predict bitcoin pricing 20 days out?  That's... amazing.  

Because price is what drives difficulty.  We're one early adopter cashing out or mainstream celebrity adopting bitcoin from exteme price movements one way or another.  Fundamentals trump technicals at this point.

I'd be willing to get on board the betting pool with the stipulation that "nothing material happens between now and 2 difficulties out", as has been the case after the run-up to $30.   Something moves the price to $50/btc?  You can bet your ass the 18.5% increase is woefully conservative.  Nothing happens?  A bit optimistic.  Someone hacks mtgox again and makes off with half a million BTC to sell on other exchanges?  It's the Bitocalypse for capacity.

I'd go with a steady state 8.5% increase month over month myself.

As far as NMC -- good point.  It was being mined at about 200k instant difficulty at peak, so around 1/8th of bitcoin capacity was diverted to NMC (permanent NMC miners account for around 17k difficulty) for around 20% of the difficulty cycle.

[1713992089]

1713992089

[1713992089]

1713992089

[1713992089]

1713992089

[1713992089]

1713992089

[1713992089]

1713992089

[grod]

We're on!  Let's see if your previously amazing predictive record fares well against my ass.  This is assuming a continued 2% monthly price erosion of bitcoin, and both of us have significant weasel room in case the price swings too far from expected.

[bcpokey]

grod kinda beat me to it. I'd be happy to lay out 8.5% for both but now I look like I'm copying him

Network growth rate has almost reached a 14day average of 1%/day due to many 3day average of sub1% (and some sub 0%). If no crazy ass price swings happen I see no reason for this trend not to continue. Like grod though, I'm not saying nothing can change, if some big event occurs all that crap goes out the window, which I readily stipulate to.

But for fun we can do 1990576 for block 141120. Your prediction I believe is 2130690 for the median.

[Mousepotato]

So what you're saying is your model can predict bitcoin pricing 20 days out?  That's... amazing.  

Because price is what drives difficulty.

I'm becoming increasingly convinced that price has absolutely nothing to do with difficulty.  I subscribe to the idea that "awareness" of Bitcoin is what drives difficulty, not including natural net growth of course.

[bcpokey]

So what you're saying is your model can predict bitcoin pricing 20 days out?  That's... amazing.  

Because price is what drives difficulty.

I'm becoming increasingly convinced that price has absolutely nothing to do with difficulty.  I subscribe to the idea that "awareness" of Bitcoin is what drives difficulty, not including natural net growth of course.

Well I hate to say this, but it's of course a congruence of multiple factors. For example, if the whole world knew of bitcoin, but the price were still sitting at $14/coin, the difficulty wouldn't be 175,000 times higher than it is today. By the same token if no one new were to find out about bitcoin but the price were to jump to $1000/coin the difficulty probably wouldn't jump 71 times higher than it is now, simply due to physical limitations and capital limitations.

In both cases it would almost definitely be higher, but more often the two are usually linked, such that when price goes up people tell their friends, or when lots of people hop on board its because lots more people have been made aware, bitcoin gets "hyped" and price goes up, both yielding high difficulties.

I would say however that in my personal opinion, price is a stronger motivator for difficulty increase than simply awareness, looked at individually. Just knowing about bitcoin doesn't really motivate most people stuck in a fiat currency system unable to imagine life any other way (most people aren't all that smart), and the minimal profit made per card per day from mining doesn't really inspire many people (Oh I make $15 a day from my triple radeon rig? so what?). But when price takes off both those who are already invested get gold rush fever, and those who hear about magical money making inventions and ridiculous profits for doing nothing want to jump on board.

I could be wrong, but that's how I see it.

[Mousepotato]

I would say however that in my personal opinion, price is a stronger motivator for difficulty increase than simply awareness, looked at individually. Just knowing about bitcoin doesn't really motivate most people stuck in a fiat currency system unable to imagine life any other way (most people aren't all that smart), and the minimal profit made per card per day from mining doesn't really inspire many people (Oh I make $15 a day from my triple radeon rig? so what?). But when price takes off both those who are already invested get gold rush fever, and those who hear about magical money making inventions and ridiculous profits for doing nothing want to jump on board

Well, of course when you factor in such elements as "human nature", you can draw a circumstantial relationship between price and difficulty.  But let's not forget that the difficulty was raised several times before there were even any exchanges, and that movement was driven primarily by awareness.  We've already seen the scenario you describe where the price of the BTC spikes, and since then we've seen the price progressively diminish, with no corresponding drop in difficulty.  In fact, none of the difficulty forecasts I've seen has any difficulty decreases at all.

I think it would be more accurate to say that difficulty does not follow price, but it is affected by it.

[jamesg]

Do any of these projections take into account the capital outlays needed to create the extra Th/s of computing power and the effects on future difficulty increases? How long do you believe that 10% increases are sustainable?

Right now, we are hovering around 12.5Th/s. To increase this 10% would mean that 1.25Th/s of new computing power would need to come online over the 10-14 day period. That is over 4,150 5830 video cards assuming 300Mh/s per card. That is $600k (for the most cost effective video card) in video cards alone not to mention other outlays of capital. If miners are buying 6850s, 6870s, 6950s or 6970s, the net dollar amounts move much higher.

Also, if we have a 10% difficultly increase for this round, to get to another 10% increase would mean another 1.375Th/s would need to come online over the next 14 day period.

Unless there is a new technology to drive more Gh/s per $$, I would assume that this will not be a sustainable trend and that we will eventually see a tapering off of difficulty increases.

Also, with more difficulty means that smaller players will be moved from being profitable to unprofitable which will have a net decrease of the total hashing power of the network.

Thoughts?

[geek-trader]

Liferealized:  I agree with you.  Especially as the supply of 5xxx series cards runs out.  The 6xxx series is not as cost effective.

At some point the difficulty has to level out.  But, is that point 2,000,000 or 10,000,000  ?

[jamesg]

Liferealized:  I agree with you.  Especially as the supply of 5xxx series cards runs out.  The 6xxx series is not as cost effective.

At some point the difficulty has to level out.  But, is that point 2,000,000 or 10,000,000  ?

First and foremost, I cannot predict the future.

But from the looks of things, I would guess that we will not have too many more 10% increases and that this "new paradigm" of ever growing difficultly will not be sustainable. I would suspect that we might even have a difficulty decrease in the near future (2 - 3 months).

The costs of getting into the 3m difficulty range are in the 10s of millions of dollars and do not take into account the net effect of driving out smaller miners (less than 5Gh).

This assumes no new discoveries / technologies that have any great impact on mining.

Again, I cannot predict the future and have been known to be wrong before.

[bcpokey]

I would say however that in my personal opinion, price is a stronger motivator for difficulty increase than simply awareness, looked at individually. Just knowing about bitcoin doesn't really motivate most people stuck in a fiat currency system unable to imagine life any other way (most people aren't all that smart), and the minimal profit made per card per day from mining doesn't really inspire many people (Oh I make $15 a day from my triple radeon rig? so what?). But when price takes off both those who are already invested get gold rush fever, and those who hear about magical money making inventions and ridiculous profits for doing nothing want to jump on board

Well, of course when you factor in such elements as "human nature", you can draw a circumstantial relationship between price and difficulty.  But let's not forget that the difficulty was raised several times before there were even any exchanges, and that movement was driven primarily by awareness.  We've already seen the scenario you describe where the price of the BTC spikes, and since then we've seen the price progressively diminish, with no corresponding drop in difficulty.  In fact, none of the difficulty forecasts I've seen has any difficulty decreases at all.

I think it would be more accurate to say that difficulty does not follow price, but it is affected by it.

Well that's a fair point. But as I said it's all a very complex mishmash. It is much more difficult for difficulty to decrease than increase. Once you have invested into bitcoin you are invested. If you are making $5 that's still $5 you are making. And the fact is that mining is still ridiculously profitable. Even after all the difficulty increases I see about 1000 USD profit per month off 5GHash in possibly the most electrically expensive area in the country. If I weren't electrically limited I'd be boosting the difficulty up myself. There's simply no reason to see difficulty drop right now. In fact if it weren't for the volatility of bitcoin scaring investors from plopping money into it, and the zero-sum aspect I think you'd see difficulty increasing far faster than it is now even without a price increase, at the current price.

But that still relies on price being the primary motivator, with awareness a subfactor. Difficulty remained relatively constant for the first year and a half or so because both awareness was low, and price was nonexistant. People added value to the coin without exchanges though, there was a thriving "over the counter" trade system created, and this combined with more awareness of the coin caused more people to get on board, raising difficulty.

Aside from the small (now minority) group of enthusiasts who mine bitcoin because they believe in it becoming an eminent player in the world currency, most people playing with mining and trading bitcoin are profit seekers, speculators, and other opportunists, who follow price. And most of the new players who become aware of bitcoin choose to join because of the pricing/profitability.

Basically you will never disentangle price and awareness, but short of a change in human nature as you put it, you will not see awareness be the primary factor in bitcoin mining being taken up. And that's why satoshi constructed it that way in my opinion. To take advantage of human nature.

All my opinion of course.

Do any of these projections take into account the capital outlays needed to create the extra Th/s of computing power and the effects on future difficulty increases? How long do you believe that 10% increases are sustainable?

Right now, we are hovering around 12.5Th/s. To increase this 10% would mean that 1.25Th/s of new computing power would need to come online over the 10-14 day period. That is over 4,150 5830 video cards assuming 300Mh/s per card. That is $600k (for the most cost effective video card) in video cards alone not to mention other outlays of capital. If miners are buying 6850s, 6870s, 6950s or 6970s, the net dollar amounts move much higher.

Also, if we have a 10% difficultly increase for this round, to get to another 10% increase would mean another 1.375Th/s would need to come online over the next 14 day period.

Unless there is a new technology to drive more Gh/s per $$, I would assume that this will not be a sustainable trend and that we will eventually see a tapering off of difficulty increases.

Also, with more difficulty means that smaller players will be moved from being profitable to unprofitable which will have a net decrease of the total hashing power of the network.

Thoughts?

Just to answer you with some of what I said before, mining is still very very profitable right now. It will have to become about half as profitable as it is today (or about twice as difficult/half the price) before I would find myself squeezed out of mining, and again, that is because I live in literally the most energy expensive area in the united states (its even more costly than many european nations). There is still room for a great deal of growth before people begin to see outlays as being untenable. But yes obviously there is some point in the future where difficulty and expenses meet and you cannot have organic growth at that point.

[molecular]

So what you're saying is your model can predict bitcoin pricing 20 days out?  That's... amazing.  

Because price is what drives difficulty.

I'm becoming increasingly convinced that price has absolutely nothing to do with difficulty.  I subscribe to the idea that "awareness" of Bitcoin is what drives difficulty, not including natural net growth of course.

Price is a large driving factor of awareness, though, right?

[molecular]

Also, if we have a 10% difficultly increase for this round, to get to another 10% increase would mean another 1.375Th/s would need to come online over the next 14 day period.

Unless there is a new technology to drive more Gh/s per $$, I would assume that this will not be a sustainable trend and that we will eventually see a tapering off of difficulty increases.

There are ASIC miners since at least March. FPGA is also to be considered (power-efficient). The new technology is being developed...

[Mousepotato]

Well that's a fair point. But as I said it's all a very complex mishmash. It is much more difficult for difficulty to decrease than increase. Once you have invested into bitcoin you are invested. If you are making $5 that's still $5 you are making. And the fact is that mining is still ridiculously profitable. Even after all the difficulty increases I see about 1000 USD profit per month off 5GHash in possibly the most electrically expensive area in the country. If I weren't electrically limited I'd be boosting the difficulty up myself. There's simply no reason to see difficulty drop right now. In fact if it weren't for the volatility of bitcoin scaring investors from plopping money into it, and the zero-sum aspect I think you'd see difficulty increasing far faster than it is now even without a price increase, at the current price.

But that still relies on price being the primary motivator, with awareness a subfactor. Difficulty remained relatively constant for the first year and a half or so because both awareness was low, and price was nonexistant. People added value to the coin without exchanges though, there was a thriving "over the counter" trade system created, and this combined with more awareness of the coin caused more people to get on board, raising difficulty.

Aside from the small (now minority) group of enthusiasts who mine bitcoin because they believe in it becoming an eminent player in the world currency, most people playing with mining and trading bitcoin are profit seekers, speculators, and other opportunists, who follow price. And most of the new players who become aware of bitcoin choose to join because of the pricing/profitability.

Basically you will never disentangle price and awareness, but short of a change in human nature as you put it, you will not see awareness be the primary factor in bitcoin mining being taken up. And that's why satoshi constructed it that way in my opinion. To take advantage of human nature.

All my opinion of course.

Oh wrd.  I hope you're right.  If so, we can look forward to the possibility of a couple rounds of difficulty reduction, following the big drop from 30+ down to 11'ish we saw recently.  At the very least, we can get some more mileage out of our current HW setups.

[DiabloD3]

Don't make me have to moderate this thread.

[mike678]

Perhaps you should try and keep up with the statements I made in my short and concise post. I said next target is quite easy to do without any modelling of any kind. I said that the trick is making a model that can forecast farther down the road, and used your forecast as an example of how difficult it can be (Your median was about 70% off actual, and even your lower bound was wrong with a huge error bar).

I applaud your work on creating a method of prediction, but until you stop patting yourself on the back for being right with the easy stuff, I suspect you will have trouble with the hard stuff. An 8% increase still seems much more likely than a > 15% increase. So we'll see, you are predicting a roughly 17% increase for your next median I believe? Let us see who does better, just for fun, no wagers necessary.

Your mathematical model says 17.5% for median, I will say 8.5% based solely on my own observation and "gut feeling". Let us see who does better for the next retarget. I know you argue against hard target exactitude, but that is not the point of this game, merely to show the ease of next retarget checking. And I haven't even really been paying that much attention lately tbqh.

EDIT: 1834633 to be more clear.

It is much harder to go out two re-targets or more. But it is something I am still working on despite any back patting.

I get it. For block 139104 you are calling 1834633, I am calling 1986157. We will see who is closer.

Care to go out to block 141120?

It's getting close to the next change. Looks like any one who guessed 8-10% was extremely close and your prediction was way off

[bcpokey]

Don't make me have to moderate this thread.

 

Perhaps you should try and keep up with the statements I made in my short and concise post. I said next target is quite easy to do without any modelling of any kind. I said that the trick is making a model that can forecast farther down the road, and used your forecast as an example of how difficult it can be (Your median was about 70% off actual, and even your lower bound was wrong with a huge error bar).

I applaud your work on creating a method of prediction, but until you stop patting yourself on the back for being right with the easy stuff, I suspect you will have trouble with the hard stuff. An 8% increase still seems much more likely than a > 15% increase. So we'll see, you are predicting a roughly 17% increase for your next median I believe? Let us see who does better, just for fun, no wagers necessary.

Your mathematical model says 17.5% for median, I will say 8.5% based solely on my own observation and "gut feeling". Let us see who does better for the next retarget. I know you argue against hard target exactitude, but that is not the point of this game, merely to show the ease of next retarget checking. And I haven't even really been paying that much attention lately tbqh.

EDIT: 1834633 to be more clear.

It is much harder to go out two re-targets or more. But it is something I am still working on despite any back patting.

I get it. For block 139104 you are calling 1834633, I am calling 1986157. We will see who is closer.

Care to go out to block 141120?

It's getting close to the next change. Looks like any one who guessed 8-10% was extremely close and your prediction was way off

Yay! 

I'll regret another hit to my income, but it's always fun to win a bet Come onnn difficulty, daddy needs a new pair of radeons!

[grod]

Yup, looking like 10% on the nose.  And instant difficulty is 8.5% higher already.

It is clear my ass has amazing predictive powers.  I will have a plaster cast of it made, so that you all can order a replica to mount on a pedistal or over your fireplace instead of that gauche moose head.

But I must admit, the growth trend is a bit surprising even accounting for variance.  While we win by price-is-right rules, looks like the next difficulty will be 10% higher although the payout per coin looks to continue trending lower.   My ass says we won't see a slowdown in this trend until BTC hit $5-$6, it's still profitable to add existing video cards to the network.

[ssateneth]

Difficulty increase was more than 10%. It was 11.7%.

[pekv2]

Yup, looking like 10% on the nose.  And instant difficulty is 8.5% higher already.

It is clear my ass has amazing predictive powers.  I will have a plaster cast of it made, so that you all can order a replica to mount on a pedistal or over your fireplace instead of that gauche moose head.

But I must admit, the growth trend is a bit surprising even accounting for variance.  While we win by price-is-right rules, looks like the next difficulty will be 10% higher although the payout per coin looks to continue trending lower.   My ass says we won't see a slowdown in this trend until BTC hit $5-$6, it's still profitable to add existing video cards to the network.

Oh boy, this is not good. I just started mining a week ago. One card. *rolls eyes*

[bcpokey]

Yup, looking like 10% on the nose.  And instant difficulty is 8.5% higher already.

It is clear my ass has amazing predictive powers.  I will have a plaster cast of it made, so that you all can order a replica to mount on a pedistal or over your fireplace instead of that gauche moose head.

But I must admit, the growth trend is a bit surprising even accounting for variance.  While we win by price-is-right rules, looks like the next difficulty will be 10% higher although the payout per coin looks to continue trending lower.   My ass says we won't see a slowdown in this trend until BTC hit $5-$6, it's still profitable to add existing video cards to the network.

Oh boy, this is not good. I just started mining a week ago. One card. *rolls eyes*

Meh, a 10% ish difference after 2 weeks is not much difference for your size. Say you were making .3BTC a day before, now you are making .273 BTC. You missed the days of 50-75% difficulty increases. Those were bad times. But if it was worth it to you to mine on a single card at 1.7mil it's fine to mine at 1.9mil difficulty.