I have a question concerning the method with the jump of some pairs from one half to the other half of the list.
As the pairs have a different pip value and a different volatility, I guess that the probability that some pairs "jump" is much higher for some of them than for others. This probability should be somehow linked to the factor volatility*pip value.
Do your practical experience confirms that?

If it is so, have you tried to use a kind of "normalized" list, opening a different lot size depending on the volatility*pip value factor?
It is just one thought because maybe it is anyway better to give more weight to the pairs with higher volatility. Just wanted to know what is your experience about this matter.