Well first post on the forum, so here is an outline of my setup so far:
1 x 7990 watercooled running 1.34 Mh/s
Cgminer --scrypt -o somepool -u username -p x -I 18 --thread-concurrency 20480 --gpu-map 0:1,1:0 -g 1 --gpu-memclock 1679,1679 --gpu-engine 1050,1050 --temp-cutoff 69,67 --temp-overheat 67,66
I was pool mining LTC until the recent difficulty hikes (although with recent events it may come down again soon). I'm currently mining altcoin low difficulties for the reason they outperform LTC in terms of the number of coins I get per day.
I tried the pools for a while and also tried solo mining, I've switched to using cgwatcher (
http://manotechnology.blogspot.co.uk/p/cgwatcher.html) which is magic for keeping things runnning and switching based upon profitablility.
I've now switched to solo mining since I realised I could outperform the pools as they mostly rely on coinchoose which is often incorrect and lagging. I'd quite like to write my own script to interogate the headless wallets and gather data for analysis. I want to average out some figures to decide which coin to mine. Before I do I need to understand something about the probability of getting blocks.
From what I gather:
Coins per 24hr = ((Myhash*coins per day)/difficulty)*(Myhash / networkhash)
Then I assumed that:
Probability of getting reward per block = (Coins per day*(Myhash / networkhash))*Fraction of day per block)/reward per block
Are the above roughly correct and does anyone know of any algorithms that handle statistical averaging? <-I'm completely new to this but a quick learner and was thinking of using python since it is so well documented already ;-)