Still fine-tuning it, but on this trade I took 25% off over 25% gains (on a gap it was over 37%) and am 50% out now at a 50% target. So far it's just an arbitrary target, like the Stock Trader's Almanac 'always sell half on a double' rule.

Looking at the historical data from Tumbler's work, the average return with the 20 contracts parameter is about 36% (with the assumption of 5% off per round-trip for commissions/slippage); if you change that to a constant $ investment, it's about 37%. So maybe something like 20% / 40% / 60% / 80% or 30/40/50/60 targets for each 25% sale would fit the historical results better. Or half out by 40% and wait for the signal on the rest.

What I suspect with the historical options trades, given my experience so far, is that many of them hit 50% gains or more at some point, even if they ended up lower in the end -- especially because of overnight gaps. So if this is the way the market works, why not have sell targets to catch the spikes -- especially if they exceed the long-term average of the system? You may miss a few 200% gains, but I suspect you will also turn more 10-20% gains into 40-60%, which would raise the overall average return.