So you are using the cryptrade for real trading right now? Let me know if the profit reporting is wrong (though note you have to set slippage level yourself, was added a couple of days ago). Though the implementation DEMA is not 100% the same at CT compared to Gekko's version.
Those stats were running CT, with gekko simulating trading with same settings on the side. They approximately agreed, and what you said about gekko's candles not being from 00sec to 60sec explained (some) variance.
Today I've started running just gekko, letting it trade and simulate at the same time. 1m period, Default DEMA & slippage.
so far simulation says:
2014-01-24 19:05:54 (INFO): (PROFIT REPORT) simulated profit: -0.00385 BTC (-0.62741%)
Which is about right. Hard to measure exactly since I'm currently holding GHS, plus it's been a lame day on the cex curve for dEMA.
(edit: Now I'm back in BTC. Simulated: 0.61022 / Actual: 0.61072 )
(edit again: few hours later Simulated: 0.60987 / Actual: 0.61094 ) increase mostly mining. Trades happened, but variance tiny.
And tiny loss bothers me not a jot, since the bot is there to make sure massive lurches in the graph get acted on quickly. (basically, so I can sleep ever.) I can take a little hit to the balance on flat days that trick EMA slightly, 'cuz the days with movement in the graphs make up for it, and the mined BTC from holding GHS @ cex balances out little hiccup losses. EMA values to 'prefer' holding GHS in flatline and very slow drift situations would be better for cex though to maximise mining.
I'm specifically running sim and real trading at the same time because of the mining from GHS held at cex; sim will show me what I'd have made just trading, whereas actual will in theory be higher due to mining rewards. (edit: knew it.) It's entirely possible at cex that's it's more profitable to sell less often in order to keep mining.
Now I've seen it run and trade, more output suggestions!
'cos as it is now you can see simulation balances but not your actual balance...
So: after every trade, something like
"Bought XX for XX" would help clarify.
And: a portfolio report with % gain/loss much like that of simulated trading would also be good after every trade.
Also: Successful has one L

One more thing: CT ignores amounts below 0.001, (i think) so there was always a bit of balance left in what one doesn't want to be holding. Word on CT's messageboard said that was a cex API limitation. Guess not, cos no such problem with gekko.
Nice.Another difference is that CT will sell mining rewards as they build up, often for unfavourable prices. Because gekko waits until the *beginning* of an EMA upswing, mining reward BTC are sold for GHS at the appropriate time, when they're (relatively) cheap.
Double Nice.Give it a day or two and I'll post how my graph looks after gekko's played with it!
Also: post your cex.io referral link already! Donations are nice but a referral is forever. Plus it's an utter delight having bot-wielding referrals in your cex charts, seeing them rise and fall as their bots trade the EMAs alongside one's own.