We have already noticed these differences.
It seems you can't estimate your payout only by taking some dudes' payouts with the same hashrate.
There are so many parameters that you don't know from the server: accepted shares, stale shares (and btw do you submit your stale shares?), discarded works, gpus used, PSU, drivers, OS, settings in config files, mod miners, network connections...
I've been mining on 2 different servers (eu and east) these last days: I didn't notice huge differences (same coins mined at -almost- the same time on the 2 servers). I may be wrong but I don't think the differences depend on which server you are.
That just does not seem fair or right. Hardware should be irrelevant once the server accepts my work and tells me I have XX.XXX hashrate. The server is accepting (as verified in stats) 3.6mhash or more of my work an I get paid for less than peers that do the same work.
I decided to split my rigs apart with one on a different address shortly after payout time. The main rig (2.2mhash) is sitting at .013/.020 BTC overnight, the other rig (1.4mhash) got .0047/.0025
http://i.imgur.com/7w1StXs.pngIf youse got some latency, try adding --no-submit-stale to your .bat file. Ain't no use in chasing after shares you can't get if others can get em first because of better latency. h2o says to leave it on because the staled shares might be due to a coin switch but that's only 1 possible reason and if you're getting those type of reject rates, you're wasting hashing power chasing after stales.
