If you create a novel way of solving problems, you should spend some time solving lots of previously unsolved problems with it, rather than trying something new every time. Only start looking for new solutions after exhausting the low-hanging fruit.