"I don't think anyone sensible would label someone "an idiot" just for looking at 'potential explanations for what looks to be a pattern of "luck" '. What often does get one's goat is when the potential explanations are so wrong headed that they just cloud the issue.

To work out if slush's pool's "luck" is unusual or not:

1. Get an average of difficulty 1 equivalent accepted shares / expected shares (mining difficulty for eachblock) for the last  "n" blocks.
2. Using the Erlang distribution, calculate the CDF using x = the average you calculated, k and lambda both = number of blocks.

You can use Wolfram Alpha if you don't have access to statistical software. Here's an example using a 10 block average average accepted/expected of 1.1:

http://www.wolframalpha.com/input/?i=erlang+distribution+Pr%28x+%3C%3D+1.1%29+%2C+k%3D10%2C+lambda%3D10

In terms of bad luck, if the answer is > 0.99 then that is pretty bad and would happen only once in one hundred reruns of k blocks. Likewise, and answer of 0.9999 would happen only once in ten thousand reruns of k blocks. If you'd averaged the last one hundred blocks and you got 0.9999, there would be good grounds to think something funky is going on. Otherwise, maybe not.

Give it a try and report back, and do ask questions if I haven't been clear.

Doing this is the only way anyone will take your concerns about 'bad luck' seriously."



+1 to whatever this guy said, it reads REALLY smart...LoL.