Bitcoin Forum

https://tradeblock.com/mining/

The last 2 weeks there was no rise or going down noticeable but now suddenly in 2 days 26 ph+ is someone testing their new chips? Or a big player delivered out 1000s off miners.

I've been watching the hashing rate closely for a while and there have been a few big surges over the last month where large amounts of capacity appears online and then drops off again after a few hours.

Yesterday's was a huge spike though - the hashing rate peaked at over 100 PH/s and was sustaining 90 PH/s for the about 9 hours (91 blocks solved in 9 hours) but it's largely gone now (75 blocks in the last 12 hours). Some of this may be statistical noise but it suggests that someone was running tests on a lot of new hardware. No idea who or why though.

The block withholders saw that people have noticed and have moved off pools and are hashing for real.

So we are screwed with the difficulty climb?

Yes, get yourself more lube. It will continue for a while yet.

Asicminer taking off  :o

Statistical anomaly? You simply do not turn on 26 PH (about 20 MWh  :o) just for test.

Don't forget about Bitmain and the S2s. I'm sure Bitmain sold a lot of these, and they were hashing before shipment on May 8th but are now enroute all around the world.

Most of batch 4 is offline now, but they will come back online once folks start receiving them from UPS.

Obviously. Although I would call it a statistical normality. You are not measuring hashrate, you are measuring the rate at which blocks are found, and we all (should) know thats inherently variable.

Could it just be....luck?

Source: https://bitcoinwisdom.com/bitcoin/difficulty

in about 13.5 hours

I did wonder about this so I've been doing a bit of research. Mining is essentially a Poisson Process so the probability distributions ought to be relatively easy to work out. I found a useful calculator: http://stattrek.com/online-calculator/poisson.aspx (a lot of the online calculators only go to 3 decimal places and so aren't so useful)

First the easy ones. How likely is it to get 10 blocks in an hour instead of 6? The answer is 0.041 so roughly once per day we should see this. In practice while the difficulty levels are increasing then we're probably seeing more like 6.6 blocks per hour and then it's 0.058 (once every 17 hours). How likely is it that we'll see zero blocks in an hour? Assuming 6.6 blocks per hour on average then that's pretty unlikely at 0.00136 (once every 735 hours, or roughly once per month).

What are the odds of 90 blocks in 9 hours, assuming 6.6 blocks per hour normally? That turns out to be 0.0000469 (once every 21.2k hours, or once every 887 days).

If someone didn't switch on a lot of additional hashing capacity then this is definitely an anomaly, not a normality  ::)

I'm thinking that I really ought to plot some graphs!

http://www.coindesk.com/cointerra-unveils-1-petahash-hosted-mining-contracts/

possibly these guys?

It says available for immediate deployment.....

Using your calculator, its actually 0.013% since you should be looking at the cumulative probability, ie, the chance of hitting 90 blocks or more when 59.4 is expected.
So it would happen more than once per year on average.

But Id also like to see which blocks we are talking about, not sure anyone accounted for the fact that time stamps on these blocks can be wildly off, up to two hours IIRC.

Tradeblock current hashrate shows 75phs and bitcoinwisdom says 67phs.

What do I miss here?

Sorry if is a dumb question.

Their estimates are averaged over a different period. The latter averages over 504 blocks, the former Im guessing over 24 hours.

So tradeblock would show more accurate data on spike cases?

Yes, that's definitely a better estimate for the likelihood. The number I'd quoted was for exactly 90 blocks instead of >= 90.

The blocks are numbers 299349 through 299439 - it's actually 91 blocks in the 9 hour period which I guess is really actually a cumulative probability of 0.0000532 assuming the timestamps are accurate. The timestamps seem likely to be reasonably accurate here though as there were blocks 11 minutes before and 4 minutes after. We still appear to be looking at a highly improbable event but one that should occur once every 10 to 11 months and I guess would be all too easy to miss unless you were watching the stats on a very close basis.

What I was interested to understand was the probability that in any 24 hour period the hash rate would be 10% high or 10% low. I couldn't find a good enough calculator (I need to set something up with a bignum library) but for 10% high or low in a 12 hour period things we get that the hash rate will be high 17.5% of the time and low 19.5% of the time. That (to me at least) was surprisingly often and suggests that it would be interesting to find a better way to estimate hashing rates. It's certainly more than enough to explain some of the jitter in the past difficulty increases though.

Right: I am going out on a limb here :-

Please correct me if I am mistaken. But there is no definitive hashrate - it seems that the hashrate is calculated by the speed the last x-number of blocks were solved. This can make it look like the hashrate is fluctuating wildly, when in actual fact, it is just a series of 'lucky' blocks making it 'look' like there is a spike.

can anyone confirm this or explain why I am mistaken....
TY

There is a definitive hash rate but it's not something that can be directly measured because we have no way of knowing who's hashing at any one time and how much capacity they actually have.

Until someone finds a way to break the SHA256 hashing algorithm then hashing is a Poisson Process. There's a most-likely time duration between events (blocks being found) but the block finding essentially occurs at random. That means that in theory we might not find any blocks for 24 hours or we might find 100 blocks within one second, just that both of these are astonishingly unlikely to happen.

The system is memoryless so even though we might like to believe that an improbable event gets cancelled out sometime later that can't be guaranteed either (it's just that things tend to work out that way most of the time). The spike I was looking at the other day is pretty improbable but there's no reason why another one couldn't happen later today, or indeed a much bigger one occur, it's just that both are unlikely to happen; neither is any less likely now though just because we saw one a few days ago.

Originally I assumed the spike had to be down to extra hashing capacity, and in fact if more hashing capacity had been added temporarily then the very unlikely event would become much more likely so that's certainly a possible explanation but without evidence of extra capacity then it's just a guess and it's probably safer to assume random behaviour.

The difficulty changes attempt to keep the system so that the most probable block finding time in the preceding 2016 blocks would have been 10 minutes but we could end up overestimating the difficulty because we randomly found blocks quicker or we could also underestimate it because we found blocks slower. That certainly accounts for why some difficulty changes seem a little larger or smaller than we might expect but over time the highest probability is that things will balance out.

I finally generated and published some results. The most surprising thing to me was that it's quite likely that the difficulty level can end up significantly skewed by random events, but there's no real way to know whether this has happened or not. The full analysis is over at: http://hashingit.com/16-hash-rate-headaches