Somehow I lucked out and missed this thread the other day, but I would like to say that I encourage people to come up with ideas for new stats, etc. If you think you found something that could work, compare it to other stats and advanced metrics, find it's strengths and weaknesses, etc. Think about why it succeeds or fails in certain areas. You may learn something in the process, or you may stumble across something interesting in which case everyone learns something.