Bitcoin Forum

The hash rate doubled to like 4thash/s right after the diffuculty change...deepbit is only 1/4 now o.o
What the...

Funny how every time there's a diff jump the hashrate calculation approximation has to be re-explained :D

lolol

Go ahead, I know you're burning to explain it!

Please explain

Why do we laugh at new people for asking legitimate questions?  Were they supposed to have read the forums a thousand feet deep and figured this out on their own?

They should be laughing at us for having a 50 million dollar "economy" around a website that still doesn't have SSL after all this time, and a program whose user interface is so dinky and devoid of help that it wouldn't even make the Apple App Store as a free app just for its complete lack of intuitiveness alone.

i think this is a miscalculation of the hashing rate right after the difficulty has changed.
it seems to happen each time the difficulty adjusts (at least from what I've seen).

i'm not an expert. someone can correct me if I'm wrong but I think when the difficulty changes the hashrate is re-calculated. it is expected that a new block will be discovered ~every 10 mins. At the start of the difficulty level if some blocks are found faster than that then the calculated hashrate is higher than it is in reality. When more blocks are mined at this difficulty level the time to mine them will average out to a more representative period for the real hash rate.

I will explain this,

When the difficulty changes, the calculation will be grossly inaccurate because it is a calculation based on averages and very few numbers are being averaged.

How you can know if the difficulty "just changed" is if there are more than about 1900 blocks left before next change.  (Each change, this number resets to 2016 and counts down).  When it is still close to 2016, the algorithm has little data to work with, and the numbers can be way off.

Ideally, the site should be more intuitive and just blank or ****** out this number until it can be believed accurate.  Perhaps ****** with a mouseover that gives the hash rate just before the block change, and explain that it might be a bit stale.

It's funny because that question is answered every time the difficulty changes. Maybe these people could use the search function once in a while, you know...


More or less. The network hashrate is established by comparing block resolution speed with difficulty. It's simply an estimate. The block resolution speed is an average taken on whatever period of time. The thing is, when the difficulty jumps up, the average used is still the one established on data from the previous difficulty. Until that sample time spans entirely into blocks found at new difficulty, the estimate will be corrupt.

Thank you

So every new user is supposed to search for something they dont know happens or exists using their mighty psychic powers of past and future knowledge?

Heh, didn't mean to make fun of the OP, I just opened this thread, read the post and was like "holy shit, that's a lot of mining power ki-- oh wait a minute, dammit"

The next difficulty jump is estimated to be 639946. Is this normal? To my new eyes it seems very big.

This again is an estimate that is based on data samples (avg. interval between blocks), you might want to wait a little bit for a larger sample to get more accurate projections.

That estimate too is based on estimate network hashrate estimate. First corruption implies second corruption of estimate.



When I see something odd, I search the forums first. Then if I can't find, I'll post. Just sayin

The next difficulty is estimated based on the current hash rate. Since the hash rate is off, so is the next difficulty.

Thank you all.

I can't for the life of me figure out why the tools are written this way.  It is quite easy to avoid the problem.

The core task is to estimate the hash rate by looking at recent blocks (either the last n blocks or all blocks in a fixed window of time).  The number of hashes that it took to calculate each block can be estimated based on the difficulty in effect for that block.  Add those hash counts and divide by the total time, and you get an estimated hash rate.  It works regardless of changes in the difficulty.  Then use the hash rate to calculate other things like when the difficulty will change and what the new difficulty will be.

Have you tried it?

Challenge:

Give me a search phrase that can answer this question without browsing beyond the 3rd page of results.

"woah". Any further questions ?

My first hit, after looking for 2 minutes:

search "network hashing", match any words, check search in topic subjects only, limit it to the mining sub section. Only one page of results, 3rd from the bottom. Thread if from February.

increased network

stock search

10 seconds

5th result

I got owned

This.

You keep seeing this come up because something is broken.

bitcoin difficulty calculation

second hit...

Now consider that you actually knew what you were looking for and you still had to go through all that.

ROTFL  ;D

Implying the guy looking for it doesn't know? That's kinda odd. There's a wealth of information on the forum and anyone slightly better than me (not hard) with the search function can find answers faster than creating a thread about it.

The other thing is, wouldn't it be nice if 5 out of 10 newcomers wouldn't start threads about deflation being bad or the network being vulnerable to the latest chimera Bitcoin haters come up with? I'm quite certain if people tried and search the forums for their questions instead of making a thread the second they have a question, we wouldn't be seeing such catastrophic headlines on the forums all the time, and we wouldn't have to perform damage control.

I meant the guy looking for an answer doesn't know the answer in advance, unlike you.

I agree that people should use the search function but the simple fact that a question has already been answered once doesn't imply (a) it will be easy to find that answer or (b) they haven't already looked.

Usually, in the time it takes you to tell people to use the search function you could have just answered their question.