So my model was super conservative with regards to targets to keep a very high win rate. For example, in the picture below two weeks ago my model gave me a buy signal at the time of the green arrow, and my 75 cent targets were hit the following day on the red arrow. However, due to the unidirectional movement of the Russell 2000 I didn't get another signal for the entire week. This situation happens quite often that I sit out for 5-10 days while the Russell moves continuously in one direction or another. By setting conservative targets I give up all those gains from the unidirectional move. Nearly as worse, whenever I sit out during a huge run my model's results deteriorate because the average amount of signals I get per day drops, as does my average profit per day, all despite not being involved within a trade.
However, it turns out if I merely close out my position upon the next signal my win rate drops down to about 50% but my total profitability improves by over 200%, which due to the exponential nature of the model means about 6 times more money per year. The average gain per win increases by 250% while the average loss increases by about half. What happens is I end up closing out a signal after a huge run of $3-5 or even higher (the month-long run after the 2016 election, for example). A closed trade after a huge run of a few dollars is worth many times the $.575 profit gained from my previous model after a successful signal that hit my $.4 and $.75 targets. The results of the new model from non-volatile periods that don't result in huge runs is not much worse, just about 20% worse.
The new model's rules are a lot more uncomfortable to enact since I don't close out the position when I have hit profitable targets, but I'm excited about it if I can keep my discipline about me.