When developing systems, it's always a good idea to focus on one system until you reach a conclusion as to whether to trade it or scrap it.
As an example, suppose it takes you an average of 50 hours to design and backtest a system. Halfway through your system development (25 hours), you discover a new idea. Lured by excitement, you pause the development of your current system and focus on creating a new system around this idea.
You are now halfway through your second system and another 25 hours have passed. In total, you have consumed 50 hours. What have you to show for it? Two half-complete projects, but nothing tradable to bring in some profit.
On the otherhand, if you had focused on your first system to completion, you may have a trading system ready and running, hopefully generating profit in the background as you start work on the second system.
Looking back at my research folders, I've realised that this has happened to me a few times. The tragedy is that you can completely forget about finishing the first system.