LOL ... maybe a better term is realistic. IMHO, the numbers are starting to sound more in-line with reality (e.g. 1-2% vs. 8%). BTW: why would I bet 3% if it meant a nearly 50% chance of creating a 50% drawdown? To me, that doesn't make sense, and doesn't appear to be conservative.Forum Mgmnt wrote:If 3% is considered conservative, it's no wonder most traders lose money.
One other triangulation on reality here. I think this optimal value all depends on the system and approach. Afterall, with 4 units scaling, default Turtle is really risking 4% per position. I have one option for "scaling into size" in my trade engine. Using this with "R11" would be betting 0.75% per trade at the above level; less than Turtle. The key difference is that Turtle doesn't always risk 4%, because not every signal scales.
Ok, so far we hadn't formally considered how to optimize systems that scale. Elsewhere in the forum, discussions suggested treating each scaled unit as a separate system [good idea: I just can't recall who mentioned it]. Using that technique would mean creating separate trade sets for each unit (all unit-1 trades, all unit-2 trades, etc), and optimizing on each set. We'd then have an idea of what would be optimal for each scaled unit.