Hi Densmirnov
You just described a 1000T with SG, very interesting ^^
Perhaps, it will be possible to calculate the indent between the orders, as well as the Martingale coefficient fully dinamically, depending on the momentum of the price change and current 'p(d)umping speed'.
In the current context of Poloniex, with a flurry of 422/429 or 500 we've seen lately, this part would doubtfully work, unless Gunthar makes a magic, which is also possible... Anyway, if Polo changes back to an stable environment, then this idea might be very very interesting to play ^^
+1