My assumption for this has always been I've had time to check off why my first group of ideas won't play out well "like chess moves" which leads me to a better more thought out move. This new idea is created with the reasons why you dot want to do your other idea "cost to much, burns to much CPU , over engineered whatever..". Takes time to narrow down what you really want to do and yelling out 3-4 ideas the first minute is a good start but giving it time is important.