Puppet
Legendary
Offline
Activity: 980
Merit: 1040
|
![](https://bitcointalk.org/Themes/custom1/images/post/xx.gif) |
September 20, 2014, 10:53:41 PM |
|
no estimate can ever be perfect, unless you have crystall ball, but it helps to understand how these sites make their prediction. AFAIK, bitcoinwisdom extrapolates the growth over the past 1000 or so blocks, and thus implicitly assumes the past growth rate will continue, whereas blockexplorer tends to look only at the current network speed, and assumes no (further) growth.
In reality, the network tends to grow, thus bitcoinwisdom tends to be closer to the truth, but that growth is also beginning to slow down, so BW overestimates the next difficulty to some extend. Blockexplorer otoh, ignoring the historic reality of ever increasing network speed, tends to underestimate it, especially in the beginning of a difficulty period.
|
|
|
|
philipma1957
Legendary
Offline
Activity: 4242
Merit: 8497
'The right to privacy matters'
|
![](https://bitcointalk.org/Themes/custom1/images/post/xx.gif) |
September 20, 2014, 11:12:13 PM |
|
no estimate can ever be perfect, unless you have crystall ball, but it helps to understand how these sites make their prediction. AFAIK, bitcoinwisdom extrapolates the growth over the past 1000 or so blocks, and thus implicitly assumes the past growth rate will continue, whereas blockexplorer tends to look only at the current network speed, and assumes no (further) growth.
In reality, the network tends to grow, thus bitcoinwisdom tends to be closer to the truth, but that growth is also beginning to slow down, so BW overestimates the next difficulty to some extend. Blockexplorer otoh, ignoring the historic reality of ever increasing network speed, tends to underestimate it, especially in the beginning of a difficulty period.
just about a perfect analysis. take your choice blend and match either way you want
|
|
|
|
eoakland
|
![](https://bitcointalk.org/Themes/custom1/images/post/xx.gif) |
September 21, 2014, 01:13:15 AM |
|
In 5 days, the estimated difficulty increase has already gone from ~29,829,000,000 to ~35,259,000,000. At the current increasing rate of network hash power, in 5 more days, it's likely to hit 40,000,000,000. This is really pushing the little guys out of business. Even with the most powerful and efficient miners, there is virtually no way to make a profit unless electricity is free and mining hardware is free. You'd expect with such extreme network hash power that BTC value should have skyrocketed to $1000 USD or more. But instead, it's remained somewhat steady around $500. And now, it's dropping closer to $400. How can this be profitable for anyone, even the huge mega farms?
I'm still doing pretty good. I am not sure where you get mining isn't profitable. Mine for the future, not today. If you're trying to pay bills with today's mining, you're going to lose. This has always been true. Think about the faith the guys in December 2011 had to have (me included) that were running farms of all sizes using gpus when the price was at $2.11 each. Electricity wasn't any cheaper and gpu's are 100X less efficient than Asics. If you can't mine and hold, you need to get out. ~BCX~ easier said than done. it's all relative i guess. it's hard to mine in CA. i know when i was mining "harder" i was paying damn near .30/kilowatt. FYI, you were correct about AUR coin. not sure that you destroyed it like you claimed you were going to, but for the fact that it is a crap coin. cheers
|
|
|
|
fewcoins
|
![](https://bitcointalk.org/Themes/custom1/images/post/xx.gif) |
September 22, 2014, 01:46:41 AM |
|
Soon all these miners will exit! NOBODY said mining would be profitable
|
|
|
|
mikerbiker6
|
![](https://bitcointalk.org/Themes/custom1/images/post/xx.gif) |
September 22, 2014, 08:42:34 PM |
|
|
Participate and Earn on the Letstalkbitcoin forum letstalkbitcoin.com
|
|
|
jwcastle (OP)
|
![](https://bitcointalk.org/Themes/custom1/images/post/xx.gif) |
September 24, 2014, 01:46:35 AM |
|
Soon all these miners will exit! NOBODY said mining would be profitable
Just about EVERYONE has mentioned mining and profit in the same sentence.
|
|
|
|
Waramp22
|
![](https://bitcointalk.org/Themes/custom1/images/post/xx.gif) |
September 24, 2014, 06:40:32 AM |
|
Soon all these miners will exit! NOBODY said mining would be profitable
Mining will never stop as bitcoin is self regulating. Its gotten a bit out of sync with the price lately because miners are very bullish or long on bitcoin and still expect the price to bounce back. We can only go so much higher before the increasing hashrate starts to slow because of miners realizing that it is currently not as profitable as they thought it was going to be. Pre-Orders are slow to arrive and thus the mining community is slow to react to declines in profitability. Much like oil, if the demand/price is dropping over the long term they will slow the investment in their drilling programs. In this case, bitcoin mining. The good part is, that if/when the increasing in hashrate starts to taper off the difficulty to mine a bitcoin will level off as well and profitability will start to return.
|
|
|
|
Puppet
Legendary
Offline
Activity: 980
Merit: 1040
|
![](https://bitcointalk.org/Themes/custom1/images/post/xx.gif) |
September 24, 2014, 06:56:04 AM |
|
Most of you seem to fail to understand that bitcoin mining is still quite profitable. Just not for you who pays retail prices for the hardware and consumer electricity prices.
|
|
|
|
OrientA
|
![](https://bitcointalk.org/Themes/custom1/images/post/xx.gif) |
September 24, 2014, 07:36:02 AM |
|
Most of you seem to fail to understand that bitcoin mining is still quite profitable. Just not for you who pays retail prices for the hardware and consumer electricity prices.
The revenue is still higher than the electricity cost. But if you factor in the cost of rig, space and labor etc, then it is not profitable for some miners.
|
|
|
|
Puppet
Legendary
Offline
Activity: 980
Merit: 1040
|
![](https://bitcointalk.org/Themes/custom1/images/post/xx.gif) |
September 24, 2014, 08:18:34 AM |
|
The revenue is still higher than the electricity cost. But if you factor in the cost of rig, space and labor etc, then it is not profitable for some miners.
That its not profitable for some is irrelevant. As long as there are companies for whom it is profitable, difficulty will keep going up. And there is no way its not profitable for industrial mines with hardware at variable production cost, such as KnC's farm.
|
|
|
|
bellgarath1
Newbie
Offline
Activity: 49
Merit: 0
|
![](https://bitcointalk.org/Themes/custom1/images/post/xx.gif) |
September 24, 2014, 05:23:21 PM |
|
One must keep reinvesting if you want to make a few bucks mining coin.
|
|
|
|
taipo
Full Member
![*](https://bitcointalk.org/Themes/custom1/images/star.gif) ![*](https://bitcointalk.org/Themes/custom1/images/star.gif)
Offline
Activity: 238
Merit: 100
Kia ora!
|
![](https://bitcointalk.org/Themes/custom1/images/post/xx.gif) |
September 25, 2014, 10:35:21 PM |
|
|
|
|
|
crazyearner
Legendary
Offline
Activity: 1820
Merit: 1001
|
![](https://bitcointalk.org/Themes/custom1/images/post/xx.gif) |
September 26, 2014, 01:14:33 AM |
|
Everyone should team up and stop mining BTC and slow down then all hit it like a boss and get a big spike of rewards xD not much longer before 1TH miners out the window and soon the 10TH are going to be out and soon after that 100TH miners and game over and over and over again. 60GH back in the day was good making 2 to 3 BTC per day now you need like 50TH or more to even brake 1BTC a day . Ill continue to alt mine.
|
|
|
|
mwizard
|
![](https://bitcointalk.org/Themes/custom1/images/post/xx.gif) |
September 26, 2014, 04:32:20 AM Last edit: September 26, 2014, 04:59:57 AM by mwizard |
|
As others have mentioned bitcoinwisdom is the one to use. It looks at the last 2016 blocks solved and should be within about 4% of the actual average network hash rate over the last 2016 blocks 95% of the time. I suspect cryptanalys.is badly in error for a few days after a difficulty change. The problem with cryptanalys is that it either ignores, or does not understand, the statistic variation in the block solve time.. It basically says we are x fraction of the way through finding 2016 blocks and it took x time so we can now calculate the expected difficulty change. But at the start of a new difficulty it is only using a few blocks in its calculation and the error is large. For example after one day cryptanalys is only using about 144 blocks and the solve time calculation error and thus error in hash rate can be up to 20%. The cryptanalys error declines as the next difficulty change approaches as there are more and more blocks included in the calculation until finally it is using 2016 blocks. bitcoinwisdon appears to always use the last 2016 blocks so statistical variation is less important. When the next difficulty change finally occurs both give the same result as both are using the solve time of the last 2016 blocks. Mathematically there is a 10 minute block average solve time with a 10 minute variance (poisson distribution). So the standard deviation percentage error in solve time is 100 / sqrt(number blocks). This gives a standard deviation error of 8.3% for 144 blocks (1 day), 4.5% for 504 blocks, and 2.2% for 2016 blocks (difficulty change). The variation will be within 2 Standard Deviations of the true value 95% of the time. Note - Six months ago, in the high growth rate days, you could have made a case for crptanalys, as on the first day after a difficulty change it used the average of that day, rather than the bitcoinwisdon average for the past 2 weeks. But with a lowering in hash rate growth this is less relevant. The crptanalys error in the first days after a difficulty change is so large it swamps the effect of the data being slightly more up to date. Feel free to improve or correct my understanding of how these sites work.
|
|
|
|
taipo
Full Member
![*](https://bitcointalk.org/Themes/custom1/images/star.gif) ![*](https://bitcointalk.org/Themes/custom1/images/star.gif)
Offline
Activity: 238
Merit: 100
Kia ora!
|
![](https://bitcointalk.org/Themes/custom1/images/post/xx.gif) |
September 26, 2014, 05:04:20 AM |
|
https://cryptanalys.is/difficulty/bitcoin_btc.php now shows 34,709,951,920 with just 7 hours on since I made that post. So its predictions this early in the piece are as you say, quite off the mark. Might have quite an interesting result this round I feel.
|
|
|
|
DigitalPirateman
Newbie
Offline
Activity: 56
Merit: 0
|
![](https://bitcointalk.org/Themes/custom1/images/post/wink.gif) |
September 26, 2014, 05:14:36 AM |
|
Kinda worried about it too but we all knew the difficulty will go this way just sooner then expected can't wait to see what it is next year. As the price its buying time!!!!!!!! https://i.imgur.com/J4UUv9a.jpg
|
|
|
|
skuser
|
![](https://bitcointalk.org/Themes/custom1/images/post/xx.gif) |
September 26, 2014, 10:08:27 AM |
|
https://cryptanalys.is/difficulty/bitcoin_btc.php now shows 34,709,951,920 with just 7 hours on since I made that post. So its predictions this early in the piece are as you say, quite off the mark. Might have quite an interesting result this round I feel. I started to parse predictions from all the sites I know about at http://nextdifficulty.com so you don't have to waste time switching between sites ![Cheesy](https://bitcointalk.org/Smileys/default/cheesy.gif) I will also track predictions in time so it'll be possible to see which are useful and which are not. If you know about other sites predicting difficulty let me know I'll add them too ![Wink](https://bitcointalk.org/Smileys/default/wink.gif)
|
|
|
|
skuser
|
![](https://bitcointalk.org/Themes/custom1/images/post/xx.gif) |
September 26, 2014, 10:21:51 AM |
|
As others have mentioned bitcoinwisdom is the one to use. It looks at the last 2016 blocks solved and should be within about 4% of the actual average network hash rate over the last 2016 blocks 95% of the time. I suspect cryptanalys.is badly in error for a few days after a difficulty change. The problem with cryptanalys is that it either ignores, or does not understand, the statistic variation in the block solve time.. It basically says we are x fraction of the way through finding 2016 blocks and it took x time so we can now calculate the expected difficulty change. But at the start of a new difficulty it is only using a few blocks in its calculation and the error is large. For example after one day cryptanalys is only using about 144 blocks and the solve time calculation error and thus error in hash rate can be up to 20%. The cryptanalys error declines as the next difficulty change approaches as there are more and more blocks included in the calculation until finally it is using 2016 blocks. bitcoinwisdon appears to always use the last 2016 blocks so statistical variation is less important. When the next difficulty change finally occurs both give the same result as both are using the solve time of the last 2016 blocks. Mathematically there is a 10 minute block average solve time with a 10 minute variance (poisson distribution). So the standard deviation percentage error in solve time is 100 / sqrt(number blocks). This gives a standard deviation error of 8.3% for 144 blocks (1 day), 4.5% for 504 blocks, and 2.2% for 2016 blocks (difficulty change). The variation will be within 2 Standard Deviations of the true value 95% of the time. Note - Six months ago, in the high growth rate days, you could have made a case for crptanalys, as on the first day after a difficulty change it used the average of that day, rather than the bitcoinwisdon average for the past 2 weeks. But with a lowering in hash rate growth this is less relevant. The crptanalys error in the first days after a difficulty change is so large it swamps the effect of the data being slightly more up to date. Feel free to improve or correct my understanding of how these sites work. If bitcoinwisdom used 2016 blocks to predict difficulty then it should predict 0% difficulty growth everytime at the start of new period. I think they are using less amount of blocks but are incorrectly recalculating solved blocks speed with new difficulty, thats why they are way off at the period start. The simple aproach is just take x last blocks, find average solving time and divide with time it should take. The problem is as you state with variance at lower blocks number and another problem is as I wrote most of the sites are lazy to recalculate blocks which have been solved in previous difficulty period with new difficulty level, which happens at the beginning of new period...
|
|
|
|
Oldminer
Legendary
Offline
Activity: 1022
Merit: 1001
|
![](https://bitcointalk.org/Themes/custom1/images/post/xx.gif) |
September 26, 2014, 10:45:22 AM |
|
I'm still of the opinion BTC will eventually die & that it was simply a prototype.
Its old, out of date, and 14m coins is nowhere near enough not to mention probably millions have been 'lost'.
Looking forward to something (maybe a current coin?) to supersede it...
|
|
|
|
OrientA
|
![](https://bitcointalk.org/Themes/custom1/images/post/xx.gif) |
September 26, 2014, 10:50:38 AM |
|
I'm still of the opinion BTC will eventually die & that it was simply a prototype.
Its old, out of date, and 14m coins is nowhere near enough not to mention probably millions have been 'lost'.
Looking forward to something (maybe a current coin?) to supersede it...
I agree with that it is a prototype. Even 1 m coins is enough because it is dividable.
|
|
|
|
|