rdyoung
Legendary
Offline
Activity: 994
Merit: 1000
|
![](https://bitcointalk.org/Themes/custom1/images/post/xx.gif) |
July 20, 2014, 07:58:17 PM |
|
... The math there is correct. The chart isn't supposed to be scary. It's just a representation of simulations. ... Lol. ![Roll Eyes](https://bitcointalk.org/Smileys/default/rolleyes.gif) It was your standard ambigous chart with #s that have no relation to each other. These type of charts are used all the time to confuse people. Fox news here in murica is known for it, they will use a standard bar chart representing X versus Y. X is 1000 and Y is 2000 but the bars are nearly the same height. The idea is that people wont read the #s and just see the bars. It works well to keep the populace uninformed and freaked out about shit that doesn't really matter. If you wanted to give some real info instead of spreading fud you would drop the % of pool that is % of network. Just stick with your % of the network as a whole. Unless your in a really small pool or solo mining with very little hardware you are going to earn on average what the math says you will. Using the #s you did was FUD plain and simple.
|
|
|
|
NotLambchop
|
![](https://bitcointalk.org/Themes/custom1/images/post/xx.gif) |
July 20, 2014, 10:05:51 PM |
|
^ Not my chart, Anotheranonlol's. The numbers are unambiguous. Each line represents the same amount of hardware: .01% of network's total. Which can be expressed a "10% of .1%," "1% of 1%," etc., etc.
The math behind those curious lines, and the significance of the number 35... I'm stumped.
Anotheranonlol, share with the rest of the class?
|
|
|
|
Anotheranonlol
|
![](https://bitcointalk.org/Themes/custom1/images/post/xx.gif) |
July 20, 2014, 10:16:51 PM Last edit: July 20, 2014, 10:27:07 PM by Anotheranonlol |
|
... The math there is correct. The chart isn't supposed to be scary. It's just a representation of simulations. ... Lol. ![Roll Eyes](https://bitcointalk.org/Smileys/default/rolleyes.gif) It was your standard ambigous chart with #s that have no relation to each other. These type of charts are used all the time to confuse people. Fox news here in murica is known for it, they will use a standard bar chart representing X versus Y. X is 1000 and Y is 2000 but the bars are nearly the same height. The idea is that people wont read the #s and just see the bars. It works well to keep the populace uninformed and freaked out about shit that doesn't really matter. If you wanted to give some real info instead of spreading fud you would drop the % of pool that is % of network. Just stick with your % of the network as a whole. Unless your in a really small pool or solo mining with very little hardware you are going to earn on average what the math says you will. Using the #s you did was FUD plain and simple. Why on earth would i drop the % of the pool that is % of the network and only leave in % of the network of the whole? When the actual point is to demonstrate the effect the size of the pool you are working on has on your expected earnings. Should you join a pool with a higher % of network hashrate variance is reduced and earnings are more predictable. Absolutely common knowledge. The idea is not for people to not read the numbers at all and get scared by squiggly lines- I've posted the numbers above that you say are incorrect. Once again it is not implying in reality you are guaranteed to earn less with a smaller pool as simulations are not reality it simply means you've got a higher chance of doing so. (and also higher chance of earning more) Are you disputing that? ^ Not my chart, Anotheranonlol's. The numbers are unambiguous. Each line represents the same amount of hardware: .01% of network's total. Which can be expressed a "10% of .1%," "1% of 1%," etc., etc.
The math behind those curious lines, and the significance of the number 35... I'm stumped.
Anotheranonlol, share with the rest of the class?
![](https://ip.bitcointalk.org/?u=https%3A%2F%2Fi.imgur.com%2FepoJzqY.png&t=665&c=k5wskMzXcYrmuQ) ![](https://ip.bitcointalk.org/?u=https%3A%2F%2Fi.imgur.com%2FgNG8WQ2.png&t=665&c=cws-zTMc2A6VAg)
|
|
|
|
rdyoung
Legendary
Offline
Activity: 994
Merit: 1000
|
![](https://bitcointalk.org/Themes/custom1/images/post/xx.gif) |
July 20, 2014, 10:24:56 PM |
|
... The math there is correct. The chart isn't supposed to be scary. It's just a representation of simulations. ... Lol. ![Roll Eyes](https://bitcointalk.org/Smileys/default/rolleyes.gif) It was your standard ambigous chart with #s that have no relation to each other. These type of charts are used all the time to confuse people. Fox news here in murica is known for it, they will use a standard bar chart representing X versus Y. X is 1000 and Y is 2000 but the bars are nearly the same height. The idea is that people wont read the #s and just see the bars. It works well to keep the populace uninformed and freaked out about shit that doesn't really matter. If you wanted to give some real info instead of spreading fud you would drop the % of pool that is % of network. Just stick with your % of the network as a whole. Unless your in a really small pool or solo mining with very little hardware you are going to earn on average what the math says you will. Using the #s you did was FUD plain and simple. Why on earth would i drop the % of the pool that is % of the network and only leave in % of the network of the whole? When the actual point is to demonstrate the effect the percentage of the pool you are working on has on your expected earnings. Should you join a pool with a higher % of network hashrate variance is reduced and earnings are more predictable. Absolutely common knowledge. The idea is not for people to not read the numbers at all and get scared by squiggly lines- I've posted the numbers above that you say are incorrect. Once again it is not implying in reality you are guaranteed to earn less with a smaller pool as simulations are not reality, it simply means you've got a higher chance of doing so. (and also higher chance of earning more) Are you disputing that? ^ Not my chart, Anotheranonlol's. The numbers are unambiguous. Each line represents the same amount of hardware: .01% of network's total. Which can be expressed a "10% of .1%," "1% of 1%," etc., etc.
The math behind those curious lines, and the significance of the number 35... I'm stumped.
Anotheranonlol, share with the rest of the class?
![](https://ip.bitcointalk.org/?u=https%3A%2F%2Fi.imgur.com%2FepoJzqY.png&t=665&c=k5wskMzXcYrmuQ) ![](https://ip.bitcointalk.org/?u=https%3A%2F%2Fi.imgur.com%2FgNG8WQ2.png&t=665&c=cws-zTMc2A6VAg) You don't provide the difficulty increases you used to create those charts. Again, just trying to throw fud into the arena. While I have no doubt that there is a ton of hidden hashing power that has yet to come online, we appear to have hit a lull. Its likely because people are dumping their money into the scrypt asics, so therefore less money being put into sha256. Add to that btc being way underpriced for the amount of difficulty increases we have seen and its possible that we will be treading water until something large enough to make a splash hits the news. I have no idea what that would be, newegg and dell both taking bitcoin to me is ginormous news, I personally loathe dell as an IT guy their computers suck, but newegg is the go to place for anything tech related. Is it going to take an amazon announcement to finally get us moving again?
|
|
|
|
NotLambchop
|
![](https://bitcointalk.org/Themes/custom1/images/post/xx.gif) |
July 20, 2014, 10:44:11 PM Last edit: July 20, 2014, 10:57:28 PM by NotLambchop |
|
@Anotheranonlol: Sorry, still don't understand what you are trying to say. What is the significance of the number 35? Please show your work, i.e. the math involved. ty
*Just to make sure we're on the same page, each difficulty adjustment=2016 blocks, each block=25 BTC, 15 difficulty adjustments=756000 BTC mined by 100% of the network, 75.6 BTC mined by .01% of the network. If you were mining in a 100% pool (as in "entire network"), the odds of you receiving 75.6 BTC should be 100% (vertical line @ 75.6 in your chart).
|
|
|
|
dyask
|
![](https://bitcointalk.org/Themes/custom1/images/post/xx.gif) |
July 20, 2014, 11:09:42 PM |
|
PETA is now way over valued going into the difficulty change. If I hadn't already sold I'd be selling at .0031 for sure! I'm amazed that there is so much demand. For the BDD haters, PETA is way over the price of B.Mine now and it pays a lot less dividend than B.Mine.
|
|
|
|
Anotheranonlol
|
![](https://bitcointalk.org/Themes/custom1/images/post/xx.gif) |
July 20, 2014, 11:18:42 PM Last edit: July 20, 2014, 11:37:51 PM by Anotheranonlol |
|
You don't provide the difficulty increases you used to create those charts. Again, just trying to throw fud into the arena.
While I have no doubt that there is a ton of hidden hashing power that has yet to come online, we appear to have hit a lull. Its likely because people are dumping their money into the scrypt asics, so therefore less money being put into sha256. Add to that btc being way underpriced for the amount of difficulty increases we have seen and its possible that we will be treading water until something large enough to make a splash hits the news. I have no idea what that would be, newegg and dell both taking bitcoin to me is ginormous news, I personally loathe dell as an IT guy their computers suck, but newegg is the go to place for anything tech related. Is it going to take an amazon announcement to finally get us moving again?
The diff increase there is 1% daily. I'm a PETA investor. not sure what fear, uncertainty or doubt I am throwing into the arena or what benefit I would get from doing so. That term seems overused though. @Anotheranonlol: Sorry, still don't understand what you are trying to say. What is the significance of the number 35? Please show your work, i.e. the math involved. ty
*Just to make sure we're on the same page, each difficulty adjustment=2016 blocks, each block=25 BTC, 15 difficulty adjustments=756000 BTC mined by 100% of the network, 75.6 BTC mined by .01% of the network. If you were mining in a 100% pool (as in "entire network"), the odds of you receiving 75.6 BTC should be 100% (vertical line @ 75.6 in your chart).
There is no particular significance to the number 35. In the first chart with solo mining with 0.01% of net hashrate the 50% reward point is around 32 BTC, in the second mining with 10% of a 0.1% pool it's around 35 BTC. That just means that's the earnings outcome approx half of the miners would expect to see. (or if you could go back and repeat the experiment under 'lab conditions', that's the outcome you'd expect 50% of the time) What you write in last paragraph is pretty much correct. (Hope your next question will not be, why is chance of reward > 75.6 greater than 0) ![Smiley](https://bitcointalk.org/Smileys/default/smiley.gif)
|
|
|
|
NotLambchop
|
![](https://bitcointalk.org/Themes/custom1/images/post/xx.gif) |
July 20, 2014, 11:31:33 PM Last edit: July 20, 2014, 11:54:55 PM by NotLambchop |
|
@Anotheranonlol: Sorry, still don't understand what you are trying to say. What is the significance of the number 35? Please show your work, i.e. the math involved. ty
*Just to make sure we're on the same page, each difficulty adjustment=2016 blocks, each block=25 BTC, 15 difficulty adjustments=756000 BTC mined by 100% of the network, 75.6 BTC mined by .01% of the network. If you were mining in a 100% pool (as in "entire network"), the odds of you receiving 75.6 BTC should be 100% (vertical line @ 75.6 in your chart).
Edit: Was there any math behind that chart, or were you just trying out your new french curve??? It looks awful nice... ... ![](https://ip.bitcointalk.org/?u=https%3A%2F%2Fi.imgur.com%2Fly0Ltxz.png&t=665&c=BgWTH3-tbT9ueg) ,,,
|
|
|
|
pedrog
Legendary
Offline
Activity: 2786
Merit: 1031
|
![](https://bitcointalk.org/Themes/custom1/images/post/xx.gif) |
July 21, 2014, 12:39:56 AM |
|
PETA is now way over valued going into the difficulty change. If I hadn't already sold I'd be selling at .0031 for sure! I'm amazed that there is so much demand. For the BDD haters, PETA is way over the price of B.Mine now and it pays a lot less dividend than B.Mine.
Still way cheaper than CEX.io...
|
|
|
|
nwfella
Legendary
Offline
Activity: 1596
Merit: 1000
Well hello there!
|
![](https://bitcointalk.org/Themes/custom1/images/post/xx.gif) |
July 21, 2014, 12:42:57 AM |
|
Way I figure it...all I need to do is hold the partial penny worth of dividends my cryptx shares are earning me and wait till BTC hits the 75-80K range...then I should just about break even ![Wink](https://bitcointalk.org/Smileys/default/wink.gif)
|
¯¯̿̿¯̿̿'̿̿̿̿̿̿̿'̿̿'̿̿̿̿̿'̿̿̿)͇̿̿)̿̿̿̿ '̿̿̿̿̿̿\̵͇̿̿\=(•̪̀●́)=o/̵͇̿̿/'̿̿ ̿ ̿̿
Gimme the crypto!!
|
|
|
jessebtc
Newbie
Offline
Activity: 8
Merit: 0
|
![](https://bitcointalk.org/Themes/custom1/images/post/xx.gif) |
July 21, 2014, 01:04:24 AM |
|
PETA is now way over valued going into the difficulty change. If I hadn't already sold I'd be selling at .0031 for sure! I'm amazed that there is so much demand. For the BDD haters, PETA is way over the price of B.Mine now and it pays a lot less dividend than B.Mine.
I totally agree, using mining ROI calculators shows 4-5 months to breakeven at 0.0028/share with somewhat optimistic difficulty increases (<10%). I sold 90% of my holdings last week at approx that price. All I can think is that there are new customers for PETA stock, or people are using different assumptions for their calculations, or think that 4-5 months is ok. (or I'm doing the calculations wrong, always a possibility.) Or they'll spend a few hundred bucks as part of the fun bitcoin mining game (hey why not). It is a shame that PETA completely changed the gameplan, the reinvest plan was something I really believed in. But the owners moved on to scryptx and whatever else, the hosting fees were prohibitively high, and in the end not making custom pcbs (afaik) ended the path. Maybe they can see the writing on the wall: either you go MegaBigPower route with enormous scale and dirt cheap power, or you design+mfgr your own asics to mine with and sell, or you'll always be playing 2nd fiddle and have an endless hard hill to climb to stay ahead of the vast number of people and small companies trying to do what you're doing with similar tech+financial resources. Anyway, it has been fun to watch. J
|
|
|
|
tragor
Newbie
Offline
Activity: 36
Merit: 0
|
![](https://bitcointalk.org/Themes/custom1/images/post/xx.gif) |
July 21, 2014, 11:14:47 AM |
|
Aaaaand at the current rate I've got a teeny tiny profit (less than 2 mbtc on an initial 49.5 mbtc investment). Today's dividend has been very low but that was expected since we started the week with 2 high days. The graph still shows a predicted increase over last week, I'm pretty sure it will move down (low 0.0018 or even 0.0017) as the days go by. https://i.imgur.com/FpJrUNZ.jpg
|
|
|
|
rdyoung
Legendary
Offline
Activity: 994
Merit: 1000
|
![](https://bitcointalk.org/Themes/custom1/images/post/xx.gif) |
July 21, 2014, 11:35:11 AM |
|
The diff increase there is 1% daily. I'm a PETA investor. not sure what fear, uncertainty or doubt I am throwing into the arena or what benefit I would get from doing so. That term seems overused though.
Yet another gaping hole in your math. The diff is adjusted every 2016 blocks. If you compute the change on a daily basis you have a compounding effect that throws your #s off even more than they already are and it gets progressively worse over time. The other issue being that you are leaving earnings unaccounted for. Fud? YES, you are trolling tossing out fud. Its 1 thing to give an unbiased view of something, its another to use wild math and loose interpretations. Maybe you aren't tossing fud intentionally, maybe your just worse at mathematical analysis than you think you are. As I and others have said before, the difficulty will level off and looks to be doing so temporarily. We are nearing the end of major improvements to chips other than efficiency of design and even that can only go so far. Is knc still the only one working towards 20nm? Did you know that intel has only recently announced ship dates for 14nm? It will be quite sometime before asic manufacturers can get in on that tech.
|
|
|
|
rdyoung
Legendary
Offline
Activity: 994
Merit: 1000
|
![](https://bitcointalk.org/Themes/custom1/images/post/xx.gif) |
July 21, 2014, 11:42:44 AM |
|
Aaaaand at the current rate I've got a teeny tiny profit (less than 2 mbtc on an initial 49.5 mbtc investment). Today's dividend has been very low but that was expected since we started the week with 2 high days. The graph still shows a predicted increase over last week, I'm pretty sure it will move down (low 0.0018 or even 0.0017) as the days go by. ![](https://ip.bitcointalk.org/?u=https%3A%2F%2Fi.imgur.com%2FFpJrUNZ.jpg&t=665&c=usVEPVfl3RWbZQ) And move down it should. This is my largest issue with cloud mining operations, the cost per ghs paid by customers/investors, IE cex, etc. The price per ghs could and should move when traded between individuals, but if/when more hardware is brought online the price for the "new" hashing power should be lower than the last batch of hardware. If cryptx is still behind this operation with the same zeal as before, I would like to see them launch their own site to buy/sell ghs among each other and directly from them as new hardware is brought online.
|
|
|
|
Anotheranonlol
|
![](https://bitcointalk.org/Themes/custom1/images/post/xx.gif) |
July 21, 2014, 07:04:04 PM Last edit: July 21, 2014, 08:27:35 PM by Anotheranonlol |
|
The diff increase there is 1% daily. I'm a PETA investor. not sure what fear, uncertainty or doubt I am throwing into the arena or what benefit I would get from doing so. That term seems overused though.
Yet another gaping hole in your math. The diff is adjusted every 2016 blocks. If you compute the change on a daily basis you have a compounding effect that throws your #s off even more than they already are and it gets progressively worse over time. The other issue being that you are leaving earnings unaccounted for. Fud? YES, you are trolling tossing out fud. Its 1 thing to give an unbiased view of something, its another to use wild math and loose interpretations. Maybe you aren't tossing fud intentionally, maybe your just worse at mathematical analysis than you think you are. As I and others have said before, the difficulty will level off and looks to be doing so temporarily. We are nearing the end of major improvements to chips other than efficiency of design and even that can only go so far. Is knc still the only one working towards 20nm? Did you know that intel has only recently announced ship dates for 14nm? It will be quite sometime before asic manufacturers can get in on that tech. 1 % daily hashrate increase (sorry, not diff increase) is used in this simulation purely because it's less than we've seen in the past, more than we'll see in the longer term future outlook by slowdown trend and scaling issues and around the level we've seen recently -- it just gives a number to work with With 0% change in hashing 2016 blocks should take nominally 14 days, and you would see diff increase (or decrease) anyway even without change in hashrate because of noise. Actually 10% of the time you will see < 1958 blocks and 10% of the time > 2073 so basically every 5 blocks diff can be 2.8% higher or lower than it should. since its set at 2016 regardless with 1% that will be nominally 12.3 days and diff should be up by ~13%, not 14%.. with 2% hashrate expansion diff typically increase in 11.19 days and diff = 24.8% and so on It's been said by others that a more accurate starting measure to record the actual hashing rate of the network even on the day diff is first set is to multiple the new difficulty by the square root of the difficulty increase, & thats also accounted for but that could of set it at 0% or 50% daily increase and it wouldn't have made a difference..because the point isn't to predict the earnings of peta at all or to give an accurate calculation for how much you will actually earn,Y es of course there's certain things not accounted for; type of share payment, eg we could move to f2pool on PPS, instead of PPLNS we could account or discount for mergemining, could account for latency, doa shares, orphans blabla it's talking about VARIANCE plain and simple and the effects choosing a smaller or larger pool that will have on your probability to earn. I'm well aware that just because a simulation reports X doesn't mean X will happen in reality. Please stop calling it 'FUD'.. That term is overused. It means fear uncertainty and doubt. Variance is well documented and this has got nothing to do with spreading fear or doubt As far as i know knc is the only one to look at less than 28nm.. supposedly 1 manufacturer plan at 16nm already for Q4 2015, but i agree after a bulk S3 shipments there should be a little plateu.
|
|
|
|
rdyoung
Legendary
Offline
Activity: 994
Merit: 1000
|
![](https://bitcointalk.org/Themes/custom1/images/post/xx.gif) |
July 21, 2014, 09:12:32 PM |
|
The diff increase there is 1% daily. I'm a PETA investor. not sure what fear, uncertainty or doubt I am throwing into the arena or what benefit I would get from doing so. That term seems overused though.
Yet another gaping hole in your math. The diff is adjusted every 2016 blocks. If you compute the change on a daily basis you have a compounding effect that throws your #s off even more than they already are and it gets progressively worse over time. The other issue being that you are leaving earnings unaccounted for. Fud? YES, you are trolling tossing out fud. Its 1 thing to give an unbiased view of something, its another to use wild math and loose interpretations. Maybe you aren't tossing fud intentionally, maybe your just worse at mathematical analysis than you think you are. As I and others have said before, the difficulty will level off and looks to be doing so temporarily. We are nearing the end of major improvements to chips other than efficiency of design and even that can only go so far. Is knc still the only one working towards 20nm? Did you know that intel has only recently announced ship dates for 14nm? It will be quite sometime before asic manufacturers can get in on that tech. 1 % daily hashrate increase (sorry, not diff increase) is used in this simulation purely because it's less than we've seen in the past, more than we'll see in the longer term future outlook by slowdown trend and scaling issues and around the level we've seen recently -- it just gives a number to work with With 0% change in hashing 2016 blocks should take nominally 14 days, and you would see diff increase (or decrease) anyway even without change in hashrate because of noise. Actually 10% of the time you will see < 1958 blocks and 10% of the time > 2073 so basically every 5 blocks diff can be 2.8% higher or lower than it should. since its set at 2016 regardless with 1% that will be nominally 12.3 days and diff should be up by ~13%, not 14%.. with 2% hashrate expansion diff typically increase in 11.19 days and diff = 24.8% and so on It's been said by others that a more accurate starting measure to record the actual hashing rate of the network even on the day diff is first set is to multiple the new difficulty by the square root of the difficulty increase, & thats also accounted for but that could of set it at 0% or 50% daily increase and it wouldn't have made a difference..because the point isn't to predict the earnings of peta at all or to give an accurate calculation for how much you will actually earn,Y es of course there's certain things not accounted for; type of share payment, eg we could move to f2pool on PPS, instead of PPLNS we could account or discount for mergemining, could account for latency, doa shares, orphans blabla it's talking about VARIANCE plain and simple and the effects choosing a smaller or larger pool that will have on your probability to earn. I'm well aware that just because a simulation reports X doesn't mean X will happen in reality. Please stop calling it 'FUD'.. That term is overused. It means fear uncertainty and doubt. Variance is well documented and this has got nothing to do with spreading fear or doubt As far as i know knc is the only one to look at less than 28nm.. supposedly 1 manufacturer plan at 16nm already for Q4 2015, but i agree after a bulk S3 shipments there should be a little plateu. 1% hashrate or 1% difficulty is the same thing. Your agreeing that we should see a plateau after the s3s come online, but your still predicting 14% non compounded over 15 difficulty changes? How is that not fud? If its not fud its just incompetence on your part as to what #s and info to focus on and how to analyze it and then put that analysis in context. Your estimated network growth would put the difficulty at about 123,745,555,366 at the end of your extrapolation on the 8th of jan 2015. The hashrate at that point would also be about 1,009,224,999 thats 1.009Exahashes. Thats a growth of 900petahashes give or take over the next 6 months. While this possible, it would take a few stars aligning to allow it happen. The price would have go parabolic again, and we would have to see a very large gain in efficiency at the 28nm level or a mass shipping of 22/20nm. I don't just mean power usage efficiency, I mean higher clockspeeds at lower voltage. Before its brought up, you can quote the shipping estimates from knc/bitmaintech/asicminer, etc till the cows come home, what actually matters is what is plugged in and hashing. No matter how many PHS of chips asicminer says they plan on producing and shipping in a week or a month, it has to first be used to build hardware and then sold to the end consumer or used in their own mine. Again, please tell me how you throwing up a graph with most of the lines trending down without providing info and context for the information is not fud. You are a shareholder, your invested in an asset that you know over time loses valuation, therefore you could be a long term holder who actually wants the price to drop so you can more in the same fashion as if you were mining at home. If you bought an antminer s3 for 500$ now and in 3 months you could sell it for 250$ OR you reinvest some of your earnings into another s3 for maybe 300$.
|
|
|
|
nwfella
Legendary
Offline
Activity: 1596
Merit: 1000
Well hello there!
|
![](https://bitcointalk.org/Themes/custom1/images/post/xx.gif) |
July 21, 2014, 11:16:24 PM |
|
Also seriously disappointed CryptX wasn't able to follow through with original vision. It all sounded real good "hear him tell it" :p Anyway, still have a couple shares left and enjoying the minimal dividends while they last but suspect i'll be selling these for cheap here soon ![Smiley](https://bitcointalk.org/Smileys/default/smiley.gif)
|
¯¯̿̿¯̿̿'̿̿̿̿̿̿̿'̿̿'̿̿̿̿̿'̿̿̿)͇̿̿)̿̿̿̿ '̿̿̿̿̿̿\̵͇̿̿\=(•̪̀●́)=o/̵͇̿̿/'̿̿ ̿ ̿̿
Gimme the crypto!!
|
|
|
Anotheranonlol
|
![](https://bitcointalk.org/Themes/custom1/images/post/xx.gif) |
July 22, 2014, 01:38:25 AM Last edit: July 22, 2014, 01:55:20 AM by Anotheranonlol |
|
The diff increase there is 1% daily. I'm a PETA investor. not sure what fear, uncertainty or doubt I am throwing into the arena or what benefit I would get from doing so. That term seems overused though.
Yet another gaping hole in your math. The diff is adjusted every 2016 blocks. If you compute the change on a daily basis you have a compounding effect that throws your #s off even more than they already are and it gets progressively worse over time. The other issue being that you are leaving earnings unaccounted for. Fud? YES, you are trolling tossing out fud. Its 1 thing to give an unbiased view of something, its another to use wild math and loose interpretations. Maybe you aren't tossing fud intentionally, maybe your just worse at mathematical analysis than you think you are. As I and others have said before, the difficulty will level off and looks to be doing so temporarily. We are nearing the end of major improvements to chips other than efficiency of design and even that can only go so far. Is knc still the only one working towards 20nm? Did you know that intel has only recently announced ship dates for 14nm? It will be quite sometime before asic manufacturers can get in on that tech. 1 % daily hashrate increase (sorry, not diff increase) is used in this simulation purely because it's less than we've seen in the past, more than we'll see in the longer term future outlook by slowdown trend and scaling issues and around the level we've seen recently -- it just gives a number to work with With 0% change in hashing 2016 blocks should take nominally 14 days, and you would see diff increase (or decrease) anyway even without change in hashrate because of noise. Actually 10% of the time you will see < 1958 blocks and 10% of the time > 2073 so basically every 5 blocks diff can be 2.8% higher or lower than it should. since its set at 2016 regardless with 1% that will be nominally 12.3 days and diff should be up by ~13%, not 14%.. with 2% hashrate expansion diff typically increase in 11.19 days and diff = 24.8% and so on It's been said by others that a more accurate starting measure to record the actual hashing rate of the network even on the day diff is first set is to multiple the new difficulty by the square root of the difficulty increase, & thats also accounted for but that could of set it at 0% or 50% daily increase and it wouldn't have made a difference..because the point isn't to predict the earnings of peta at all or to give an accurate calculation for how much you will actually earn,Y es of course there's certain things not accounted for; type of share payment, eg we could move to f2pool on PPS, instead of PPLNS we could account or discount for mergemining, could account for latency, doa shares, orphans blabla it's talking about VARIANCE plain and simple and the effects choosing a smaller or larger pool that will have on your probability to earn. I'm well aware that just because a simulation reports X doesn't mean X will happen in reality. Please stop calling it 'FUD'.. That term is overused. It means fear uncertainty and doubt. Variance is well documented and this has got nothing to do with spreading fear or doubt As far as i know knc is the only one to look at less than 28nm.. supposedly 1 manufacturer plan at 16nm already for Q4 2015, but i agree after a bulk S3 shipments there should be a little plateu. 1% hashrate or 1% difficulty is the same thing. Your agreeing that we should see a plateau after the s3s come online, but your still predicting 14% non compounded over 15 difficulty changes? How is that not fud? If its not fud its just incompetence on your part as to what #s and info to focus on and how to analyze it and then put that analysis in context. Your estimated network growth would put the difficulty at about 123,745,555,366 at the end of your extrapolation on the 8th of jan 2015. The hashrate at that point would also be about 1,009,224,999 thats 1.009Exahashes. Thats a growth of 900petahashes give or take over the next 6 months. While this possible, it would take a few stars aligning to allow it happen. The price would have go parabolic again, and we would have to see a very large gain in efficiency at the 28nm level or a mass shipping of 22/20nm. I don't just mean power usage efficiency, I mean higher clockspeeds at lower voltage. Before its brought up, you can quote the shipping estimates from knc/bitmaintech/asicminer, etc till the cows come home, what actually matters is what is plugged in and hashing. No matter how many PHS of chips asicminer says they plan on producing and shipping in a week or a month, it has to first be used to build hardware and then sold to the end consumer or used in their own mine. Again, please tell me how you throwing up a graph with most of the lines trending down without providing info and context for the information is not fud. You are a shareholder, your invested in an asset that you know over time loses valuation, therefore you could be a long term holder who actually wants the price to drop so you can more in the same fashion as if you were mining at home. If you bought an antminer s3 for 500$ now and in 3 months you could sell it for 250$ OR you reinvest some of your earnings into another s3 for maybe 300$. 1% diff increase daily and 1% hashrate increase daily are not the same things no. and 1% per day puts global net hashrate at a growth factor of 6.177x, increase of ~723 PH in 6 month and putting in 860 PH range. The growth rate from 6 months ago is more like 10x I don't think we are going to see perpetual 14% diff increases at all long into the future, and I've argued against those that look at the past increases of say 20% and project we will continue to see exponential increases of the magnitude earlier in the thread. I am invested in this project, but wouldn't be invested if I believed we would continue to see increases of that level because it's simply not profitable without cryptx lowering fees. As it stands for now, yes it can be profitable. No I'm not trying to bring the price down by posting graphs with lines that happen to go down (shock horror, where else would they go?), but I'd certainly be doing a better job than this if i was.. But that's irrelevant, you are misconstruing my point so much it is unreal. this is NOT ABOUT the difficulty increase and it never was, not even slightly. It is not either a specific projection about the earnings of peta and never was intended to be. Once again those charts could of said +500% daily hashrate increase or any equivalent unrealistic number for all that it matters. The point is simple. to compare variance and variance only and the effects choosing a smaller or larger pool has on your expected probability of profiting.
|
|
|
|
rdyoung
Legendary
Offline
Activity: 994
Merit: 1000
|
![](https://bitcointalk.org/Themes/custom1/images/post/xx.gif) |
July 22, 2014, 05:48:05 AM |
|
The diff increase there is 1% daily. I'm a PETA investor. not sure what fear, uncertainty or doubt I am throwing into the arena or what benefit I would get from doing so. That term seems overused though.
Yet another gaping hole in your math. The diff is adjusted every 2016 blocks. If you compute the change on a daily basis you have a compounding effect that throws your #s off even more than they already are and it gets progressively worse over time. The other issue being that you are leaving earnings unaccounted for. Fud? YES, you are trolling tossing out fud. Its 1 thing to give an unbiased view of something, its another to use wild math and loose interpretations. Maybe you aren't tossing fud intentionally, maybe your just worse at mathematical analysis than you think you are. As I and others have said before, the difficulty will level off and looks to be doing so temporarily. We are nearing the end of major improvements to chips other than efficiency of design and even that can only go so far. Is knc still the only one working towards 20nm? Did you know that intel has only recently announced ship dates for 14nm? It will be quite sometime before asic manufacturers can get in on that tech. 1 % daily hashrate increase (sorry, not diff increase) is used in this simulation purely because it's less than we've seen in the past, more than we'll see in the longer term future outlook by slowdown trend and scaling issues and around the level we've seen recently -- it just gives a number to work with With 0% change in hashing 2016 blocks should take nominally 14 days, and you would see diff increase (or decrease) anyway even without change in hashrate because of noise. Actually 10% of the time you will see < 1958 blocks and 10% of the time > 2073 so basically every 5 blocks diff can be 2.8% higher or lower than it should. since its set at 2016 regardless with 1% that will be nominally 12.3 days and diff should be up by ~13%, not 14%.. with 2% hashrate expansion diff typically increase in 11.19 days and diff = 24.8% and so on It's been said by others that a more accurate starting measure to record the actual hashing rate of the network even on the day diff is first set is to multiple the new difficulty by the square root of the difficulty increase, & thats also accounted for but that could of set it at 0% or 50% daily increase and it wouldn't have made a difference..because the point isn't to predict the earnings of peta at all or to give an accurate calculation for how much you will actually earn,Y es of course there's certain things not accounted for; type of share payment, eg we could move to f2pool on PPS, instead of PPLNS we could account or discount for mergemining, could account for latency, doa shares, orphans blabla it's talking about VARIANCE plain and simple and the effects choosing a smaller or larger pool that will have on your probability to earn. I'm well aware that just because a simulation reports X doesn't mean X will happen in reality. Please stop calling it 'FUD'.. That term is overused. It means fear uncertainty and doubt. Variance is well documented and this has got nothing to do with spreading fear or doubt As far as i know knc is the only one to look at less than 28nm.. supposedly 1 manufacturer plan at 16nm already for Q4 2015, but i agree after a bulk S3 shipments there should be a little plateu. 1% hashrate or 1% difficulty is the same thing. Your agreeing that we should see a plateau after the s3s come online, but your still predicting 14% non compounded over 15 difficulty changes? How is that not fud? If its not fud its just incompetence on your part as to what #s and info to focus on and how to analyze it and then put that analysis in context. Your estimated network growth would put the difficulty at about 123,745,555,366 at the end of your extrapolation on the 8th of jan 2015. The hashrate at that point would also be about 1,009,224,999 thats 1.009Exahashes. Thats a growth of 900petahashes give or take over the next 6 months. While this possible, it would take a few stars aligning to allow it happen. The price would have go parabolic again, and we would have to see a very large gain in efficiency at the 28nm level or a mass shipping of 22/20nm. I don't just mean power usage efficiency, I mean higher clockspeeds at lower voltage. Before its brought up, you can quote the shipping estimates from knc/bitmaintech/asicminer, etc till the cows come home, what actually matters is what is plugged in and hashing. No matter how many PHS of chips asicminer says they plan on producing and shipping in a week or a month, it has to first be used to build hardware and then sold to the end consumer or used in their own mine. Again, please tell me how you throwing up a graph with most of the lines trending down without providing info and context for the information is not fud. You are a shareholder, your invested in an asset that you know over time loses valuation, therefore you could be a long term holder who actually wants the price to drop so you can more in the same fashion as if you were mining at home. If you bought an antminer s3 for 500$ now and in 3 months you could sell it for 250$ OR you reinvest some of your earnings into another s3 for maybe 300$. 1% diff increase daily and 1% hashrate increase daily are not the same things no. and 1% per day puts global net hashrate at a growth factor of 6.177x, increase of ~723 PH in 6 month and putting in 860 PH range. The growth rate from 6 months ago is more like 10x I don't think we are going to see perpetual 14% diff increases at all long into the future, and I've argued against those that look at the past increases of say 20% and project we will continue to see exponential increases of the magnitude earlier in the thread. I am invested in this project, but wouldn't be invested if I believed we would continue to see increases of that level because it's simply not profitable without cryptx lowering fees. As it stands for now, yes it can be profitable. No I'm not trying to bring the price down by posting graphs with lines that happen to go down (shock horror, where else would they go?), but I'd certainly be doing a better job than this if i was.. But that's irrelevant, you are misconstruing my point so much it is unreal. this is NOT ABOUT the difficulty increase and it never was, not even slightly. It is not either a specific projection about the earnings of peta and never was intended to be. Once again those charts could of said +500% daily hashrate increase or any equivalent unrealistic number for all that it matters. The point is simple. to compare variance and variance only and the effects choosing a smaller or larger pool has on your expected probability of profiting. You sir have some serious cognitive dissonance going on. You are analyzing and finding meaning in #s that don't add up. Variance is to be expected in a "luck" based system, but with enough data the math adds up and makes sense, if I was buying scratch offs that were designed to win 98% of the time, I would likely only win 10 out of the first 100 I scratched, however if I scratched 1k or 10k my wins would get closer to the 98% average. Unless you or someone else can explain how x% growth in hashrate does not equal x% growth in difficulty, I will see myself out. Maybe I am thinking in logic and your thinking in elvish, I really don't know, but I can't waste anymore of my energy trying to make sense of your 2*2=5 logic "because of variance".
|
|
|
|
Anotheranonlol
|
![](https://bitcointalk.org/Themes/custom1/images/post/xx.gif) |
July 22, 2014, 06:36:19 AM Last edit: July 22, 2014, 09:59:07 AM by Anotheranonlol |
|
You sir have some serious cognitive dissonance going on. You are analyzing and finding meaning in #s that don't add up. Variance is to be expected in a "luck" based system, but with enough data the math adds up and makes sense, if I was buying scratch offs that were designed to win 98% of the time, I would likely only win 10 out of the first 100 I scratched, however if I scratched 1k or 10k my wins would get closer to the 98% average.
Unless you or someone else can explain how x% growth in hashrate does not equal x% growth in difficulty, I will see myself out. Maybe I am thinking in logic and your thinking in elvish, I really don't know, but I can't waste anymore of my energy trying to make sense of your 2*2=5 logic "because of variance".
Variance is to be expected/ The simulations in those charts demonstrated expected probability were done 10 million times and smoothed. A little more than 1k or 10k... and hardly a small run. I don't know if you consider 6 months long but this project hasn't exactly got decades to work with.. I'm sure I've written about 5 times already but I do understand a simulation doesn't equal reality - it shows what can happen and the chance if it happening, not what will or has happened. quick glance at the p2pool global stats (assuming they are accurate, assuming bitfury hardware played nice, assuming optimal conditions and you will see yourself things would of been pretty decent) About the diff= Difficulty is set retrospectively, at a level that would make the preceding 2016 blocks take 14 days to find. So diff lags 5.5 to 7 days after the real hashrate level. This is why It was mentioned taking the square root of the ratio of the new to previous difficulty level and then multiplying it by the new difficulty and using that as a baseline. As was written before even if BTC hashrate stayed exactly the same with 0% growth you would still see diff changes because of how bitcoin mining works (largely; noise) So you can subtract noise from the baseline and end up with a poisson process expecting to oscillate around 0 giving you an actual picture of the hashrate (NOT DIFF) variation and then build a variation probability histogram from there. Double hashrate increase does not mean double % denominated diff increases, again as outlined in a previous post..things become different when actual hashrate is increasing on a 2 year simulation (not accounting for latency, orphans etc) with btc hashrate staying exactly constant you still have 8 instances where diff increases by more than 4% and 7 instances where it decreases by the same amount. Just 14 successive difficulty increases of 5% will put our global hash-rate apparently ~2x % higher..yet the actual hashrate figures have never budged. (Of course that levels out given enough time.) Bitcoin hashrate doesn't move, difficulty does. You can't explain that. I think the real story is you have spent too much time with mining calculators Please, close the door on your way out. ![Kiss](https://bitcointalk.org/Smileys/default/kiss.gif)
|
|
|
|
|