Quote from: dub0matic on March 04, 2012, 08:03:43 AM

Quote from: Vanderbleek on March 04, 2012, 05:53:02 AM

Quote from: dub0matic on March 04, 2012, 05:27:04 AM

one thing i dont get is that i have been mining here for a few days now. I have been getting alot of shares in the cgminer window but in the p2pool window i have hardly received any shares

p2pool uses 2 types of shares: the ones you are seeing in CGminer are only used for estimating hashrate, and are a much lower difficulty than actual p2pool share. The high difficulty shares are what are added to the chain, and determine payment.

so the only type of miner that should mine at p2pool are people with alot of hash power. is it even worth it to mine 800mh/s here

This is a TL;DR but anyone interested ...

The reality of it is that you need to understand what's happening and ignore the comments like "I get this amount so it's OK" or "I get his amount and it sux" that clearly don't understand what to expect (since those comments are meaningless)

The pool gets blocks with variance.

The variance is a multiplying factor on the expected time to get a block.
e.g. a 300GH/s pool would expect to get 4.0313724038696 blocks a day on average with the current difficulty

What that means of course is it would expect to get a block on average every 5hr 57min 11.907sec

I'm not going to attempt to calculate the expected variances/standard deviations but however point out examples of big pools getting large variances for finding a block.

DeepBit got a block about a week ago at 824% of the expected shares.
You see high % every few weeks on DeepBit.
Yesterday 3-Mar they got a daily average of about 200% for ALL blocks that entire day.

What this means is of course is that those sort of number are not to be considered improbable.

So firstly, the P2Pool could get a block at 800% of the expected share rate - or 1day 23hr 37min 35.259sec
It happens some times, get used to the idea it will happen as it does for everyone.
It also happens some times that blocks appear in a few seconds or a few minutes.
That's the results of using a pseudo-random function to find blocks.

However ... then there is the next step and that is the probability of you finding a share.
Lets just assume that it is currently 500difficulty to find a share.
That means that at 800mh/s you would expect (on average) to find a share every 44 minutes 44.35 seconds
However, you can find a share in a few seconds or even 800% of that (5 hours 57 minutes 54.8 seconds)
You have exactly the same variance issue as any pool or any person has finding a block.

The bit that seems to be ignored or misunderstood a lot is that the variance is applied to your expected time.
If your expected time is 1 hour than it could take 8 hours for you to find a share.
If your expected time is 10 minutes then it could take 80 minutes for you to find a share.

The variance is a multiplier and thus seems to really suck if you have a low hash rate and are unlucky ... but the reality is that you should expect both good and bad luck to happen since it is random.
(and don't forget it is also a multiplier on the pools expected block finding rate)

Bottom line, with a low hash rate you can expect your share of the 49.25+trn of each block to be VERY unstable.
But over a long time it should get close to the expected value.
The effect of variance on you is better than solo mining but worse than 1 difficulty pool mining.