My only counter point would be...because influence should really be quantified by numbers, percentages, etc, where you are getting xyz = abc data....using terms like "decent results"....or "majority of the time" or "doing pretty good" are very generalized, vague, non specific terminology. In that case...just about anything can be justified and its left up to the specific person to categorize or validate himself...and or skill level. It's not a big issue...but probably goes to the heart that provability to others is a very difficult undertaking....beyond the idea, why would you want to do that in the first place? If I have 2,000 rolls...feel the data supports influence, wonderful. The catch is what about the forthcoming 2,000 rolls. If we analyze a pattern..that's past tense. It's useless unless replicable in this game. I guess money could validate it to the person. If I ended up the year in the green. I don't track anything beyond that. Is there a valid confirmation there for purposely skewing random if that's what I attempted to do? You tell me.