Bitcoin Forum

Anyone notice that the hash rate slope is strangely positive recently?  Difficulty is projected to go up and we reached 1% hashrate growth for the first time since August!

I'm guessing it will go up again and exceed the old high eventually when we will see more and more FPGA mining in 2012. If the price goes down again, I assume hashrate will temporarily decrease, yet even with low prices it might stabilze and/or increase due to more effective mining methods.

It'll be clear unitl next diff adjust.

Ofcourse hashrate rising because it's profitable to mine again ;D

Well, profitable for some with super-cheap rigs maybe, with no cases and cheap electricity. The difficulty isn't low enough to give much extra income. The price hasn't really recovered as to anywhere profitable yet. Even merged mining won't do the trick, the nmc income doesn't cut it due to the latest nmc flood and price decrease, perhaps a 10% bonus to Btc at this point, more likely less. And let's not forget that for many people, there is no way to trade at the moment, no SEPA on MtGox nor Tradehill.

  'The Slope is a Lie'
  Observe the same peak just before the recent difficulty change. It is a normal pattern that is constant in the matrix. A sort of 'osscilation' that I've been observing for 6 months now. That's not to say, you are wrong, but there is no where near enough block data this difficulty yet to draw such an increase conclusion.
 Linear
http://bitcoin.sipa.be/speed-lin.png

 Exponential
http://bitcoin.sipa.be/speed.png

its bullish.

Meaningless/sampling bias/small sample size.

According to the numbers at bitcoincharts, it's still going to go down a little, and we're still not mining over 6 blocks an hour.

How are the charts at http://bitcoin.sipa.be/ calculated.

Is it just a function of the difficulty and the average block creation time over some period?

If so does anyone know how many blocks the average is over?

That is the only way any site or pool can estimate hashing power.

The site tells you how large the window is.  1day, 7day, 14 day, 30 days, etc.

It doesn't look like it going up to me.

http://bitcoin.sipa.be/speed-lin-2k.png

Yeah there is a tick up in the 8 hour line but look how much the 8 hour line varies across the chart.  We are still below the 3 day average.

Commenting on predicted difficulty increases/decreases is kind of dumb when it's only been a few days since the last adjustment.  As of  right now I believe it's set to go down, but who know what will actually happen over the next 10-12 days.

7-day estimate for growth is heading up and ready to break 0%.
Estimated next difficulty is notably not down by a lot (only 0.2%).
We might see growth soon...

It has gone up until recent fall back down from $3.8 because the price dropped down all the way close to $2 and many stopped mining, then came back when price recovered back to $3.8.
Also the difficulty for a short time was looking to go up, but it doesn't seem to be the case anymore unfortunately.

All this is showing is the delayed reaction of the miners coming back shortly to mine while it's still profitable, but as the general downtrend continues, it will go back down like it has unless the actual price recovers.

This time around the recovery from losing miner to the previous price dip was amplified by the price falling by about 100% and recovering nearly 80%+ of the fall unlike the previous ones that dropped less in price percentage.

That changes the profitability quite a bit for miners which is showing on the graph. They will either drop or rise again depending on the price of bitcoin. It's certainly not the other way around. The leading indicator is the price, not difficulty or mining power of the network.

Actually the change I am seeing is stability.

The daily/weekly/monthly changes in both difficulty and BTC:USD prices are becoming smaller.  I mean the projection for next difficulty change is <1%.  A far cry from 50% ups and 35% downs.  We will see if this trend holds but I actually stopped looking at Bitcoin prices everyday.  For last week or so I have known I could get ~3 BTC +/- 20% when I sold coins to cover my electrical costs.  Granted 20% is huge for any other price but it sure beats the 100%+ daily volatility.

  its -11% as of right now. Still a bit early to tell with the standard 'frequency' in movement on any given day. Doesn't helpt that BtcGuild was down about 1,000GH, from about 6 hours ago until last check 30 minutes ago though.

  And yea, it would be nice to find some stabalization range on the price. Hopefully that trend holds if nothing else.

See what happens when I open my mouth.   ;D 

Still even 11% would be a rather minor change either up or down compared to the average of the last dozen or so difficulty changes.

  *inserts sock*  ;D

  Much agreed. If I had to speculate it; If price remains stabalized for another 10 days it will be -3% to -10%.  If price continues downward and hits below $2.6~ in the next 10 days then -7% to -18%.  If price rebounds to $3.25+ then -2% to +10%.   Why 10 days?  because any price event too close to next diff change would not have time to reflect fully in the adjustment imho.

    Cheers

It is just a function of the difficulty and timestamps of found blocks. It's however more than just a windowed average.

I use an exponential window (so essentially, using all blocks ever, but those longer ago count significantly less), and perform some extrapolation by fitting a growing speed function through the data. The time factor for the exponential windows is listed in the graphs.

Here is a link showing you that difficulty drop in percentage that is easy to see for people. http://btcserv.net/bitcoin/history/

I guess seeing a difficulty drop percentage less than -18% may serve as a sign we are getting closer to finding the bottom, but then it would be easier to just look at the price to determine that outright instead of trying to forecast it with projected difficulty change.