What's going on with this pool's "luck"??
This is ridiculous..
Can someone post the average luck for the last 30 days?
Statistical partI think I had about 80% shares rewarded on April 30. I still had 80% shares rewarded on May 31st. We can then grossly estimate the pool luck was 80% for May 2015.
Is it unexpected? Let's see...
Today, the reported hashing power is 6.7 PH/s. Assuming difficulty was at current rate for this month and hashrate has not moved for simplification, it means there should have been about 86 blocks founds. The estimation of the pool luck tells us that 80% of this expected value was found -> 67 blocks.
Is it statistically surprising? What are the odds that 67 blocks or less are found with what we know?
The number of blocks found can be modelled with a Poisson Distribution with mean 86. A quick Excel calculation tells me that with assumptions told in this post,
there was about a 8% probability to find 67 blocks or less for the month of may. Not that surprising that it can happen.. That's the kind of variation you can expect with the size of the pool.
Let's remember that we are also more sensitive to the bad luck of the pool versus its good luck which can bias our view.
ConclusionIf you want less variation in your reward, you should seek bigger pool.
https://blockchain.info/pools . You can also put your hashing power on many pools which will have for effect to decrease your variance.
For now, all my hashing power is on Eligius. I tried BTCGuild, Ghash, Antpool... But what is important for me is the transparency of the pool and the expected reward value (No fees). This is what you have with Eligius and I'm a big fan of the pool.
I find it unfortunate that it shrinks in hashing power since it increases variance, but the only thing I can do is keep my hashing power there...!