Van Tharp, one of the two authors who helped me finally get it, actually suggests in his latest book to design one system for each market condition that tends to occur on a regular basis. It's funny because I actually did this without realizing it. I was working on a variant of taking pullbacks in up trending stocks (trading long calls, yikes) and just about got the system nailed, when the market got volatile and sideways. Coincidentally, the trend program presented here tends to explode to the upside during these rough conditions. It takes very few trades in a bull or quiet market. One works when the other doesn't. Both are trend programs. Both have distinct market conditions when they are appropriate. I haven't even started developing the range program. I believe this type of program is the only option when you're moving trillions around.
Minor update for anyone who is curious........ The trend program is holding up nicely in the latest market action and has made new equity highs. It was tested on 13 years of hourly data. I also had a 1 year version and it has failed the forward test with a flat equity result over the past two months. The latest market volatility is too low to trigger any new positions and the system has been flat for a couple weeks. I picked up Trading Systems by Jaekle and Tomasini, which gave a great overview of what I've been discovering in Tradestation while developing this system, as well as walk forward verification and recalibration of a system, and operating a portfolio of systems. Parameter robustness and general system validity is a subject that seems to be lost on many system builders, but it's covered pretty well in the book. I may scrap the range system idea and just run several trend systems on various futures markets when the time is right. This appears to be an old idea that has survived to the present time.