speculation on the time of finding new blocks

Rannasha

hero member

Activity: 728

Merit: 500

Quote from: chenchunyu88 on August 16, 2013, 12:26:09 PM

I am kind of wondering if there is method to estimate the time require to find a block when mining a specific block.

I am right now mining using aka Slush's pool. An interesting finding is that sometimes it takes hours (up to 10) to find a new block while other times it takes just few minutes.

If someone can estimate the time of each popular pools to find a new block at the beginning of mining, then we can switch pools to maximize the profit. I am wondering if it even possible to do that. Hope someone who knows clearly how bitcoin source code work and the mechanics of BTC can answer the question.

I am working on my PhD on Statistics so I am ready to write the prediction code once the parameters affecting the time of find new blocks are addressed. Hopefully we together can find some insights here and make a lot of BTC.

Finding a block is a Poisson process, so the expected time to find a new block follows an exponential distribution. There really isn't any arcane voodoo behind it. Each hash calculated is a lottery-ticket, where the difficulty decides the probability of winning. The expected average time to find a block is given by D * 2^32 / H (D = difficulty, H = hashrate).

Many pools use a PPLNS payout system (or some other system) to prevent pool-hoppers from gaining increased returns, so it's unlikely that you can find a way to optimally select a pool. In the long run, they all give you the same expectation value for the profits, bigger pools just have less variance than smaller ones.

DebitMe

legendary

Activity: 2800

Merit: 1012

Get Paid Crypto To Walk or Drive

That is called Pool Hopping and is pretty frowned upon by not only Pool Admin, but people dedicated to individual pools. You are essentially eating into their profits by jumping onto their pools before they find a block. This is countered on most pools by payout structures that limit users who have not been mining for extended periods of time with the pool.

In summary, I doubt the people who actually know these statistics will easily give them up as they will lose profits from the pool they are currently mining at. Plus, as previously stated, it is random luck to find a block. The best way to know who will most probably get a block next is the pool with the highest hash rate.

Bitweasil

sr. member

Activity: 420

Merit: 250

It's pretty much pure luck.

Unless you can find some weaknesses in SHA256 to exploit. But SHA256 is a pretty good approximation of a random oracle from my testing.

chenchunyu88

full member

Activity: 126

Merit: 100

I am kind of wondering if there is method to estimate the time require to find a block when mining a specific block.

I am right now mining using aka Slush's pool. An interesting finding is that sometimes it takes hours (up to 10) to find a new block while other times it takes just few minutes.

If someone can estimate the time of each popular pools to find a new block at the beginning of mining, then we can switch pools to maximize the profit. I am wondering if it even possible to do that. Hope someone who knows clearly how bitcoin source code work and the mechanics of BTC can answer the question.

I am working on my PhD on Statistics so I am ready to write the prediction code once the parameters affecting the time of find new blocks are addressed. Hopefully we together can find some insights here and make a lot of BTC.

Topic: speculation on the time of finding new blocks (Read 808 times)