...
Thank You!
May I ask what CDF[ERL] means and how it is calculated and exactly what do you mean by "the probability of getting worse than that result?"
Compared to what result? luck? % difficulty? CDF itself?
Sorry if it is obvious to everyone else, (newbie) but what does the %luck really mean anything as far as the average miner looking at stats if CDF[ERL} is the real key?
CDF[Erl] is the Erlang distribution of the CDF
The value I show allows you to get the probability of getting a result worse than the data shown.
CDF means Cumulative Distribution Function
What it comes from is the integral under the curve of the probability results of a Poission Distributed random event.
In simple terms, a graph of the chances of finding a block looks like a bell shaped curve.
So you have a curve that shows that the chance of finding a block with zero hashes is zero, the chance of finding a block with infinite hashes is zero, and the chances of whatever in-between vary from close to zero, up to some limit then back down - it's a bell shaped curve.
Here's the Wikipedia page about that:
https://en.wikipedia.org/wiki/Cumulative_distribution_functionSo anyway, if say you wanted to know how often would you get a worse block than Diff 200%, the CDF can tell you that.
It adds up the area under the bell shaped curve up to 200%
The CDF for 200% is 0.86466471676339 i.e. 86.47% of the time, you expect blocks to be below 200%.
So how about worse than 200%? That's 1/(1-CDF) = 7.389 so that means, on average over a large number of events (blocks) you'd expect about 1 in 7.4 blocks to be over 200%
Now that's all good and fine except, well, that's just looking at a single block.
As I mentioned, a single sample will tell you next to nothing about a large set of events.
In comes the Erlang distribution -
https://en.wikipedia.org/wiki/Erlang_distributionThis lets you determine something similar to the CDF but for a number of events.
So as the pool stats page currently says:
Last 25 Blocks 13.4wks 118.13% 123.08% 0.8219 84.65% 103.25%
We see that the average Diff of those 25 blocks was 118.13%
So what's the chances that you'd get worse than that?
CDF[Erl] for 118.13% for 25 events is 0.8219
Using the earlier calculation 1/(1-CDF[Erl]) = 5.615 so that means, on average over a large number of events (blocks) you'd expect about 1 in 5.6 sets of 25 blocks to average worse (higher) than 118.13%