Originally Posted by
regnis
Mick-I easily threw 500,000 trials at home on my table, and probably a lot more. Do you think that didn't give me empirical evidence? Do you really think I woke up one day and decided I was a shooter/ You guys are looking for exact numbers and percentages for each number thrown to determine an exact edge. If you have played craps, and RS should know better, once it was determined that I could avoid the 7 for lengthy periods, and that during those periods I could throw a disproportionate amount of 10s and 8s, why do I need anything more. Why do I need exact percentages. I bet the 8 and 10---duh!!!!!
Now. I don't disagree that I could have possibly optimized my bets better with that exact knowledge, but again, it is craps. There is nothing exact about it.
And don't think there weren't days where my shoulder hurt from hockey or my hand or wrist was sore that I knew immediately I couldn't throw and quickly quit. My edge was my throw--nothing more.
I've come to learn over time that empirical evidence isn't all that great or trustworthy. Last year I played a good amount of blackjack (card counting) and I felt like I was off my game because I was just losing and losing and losing. I didn't know how much I was down in BJ for the current year, but I knew it had to be a lot. I estimated in my head 15-20k loss for the first few months. I opened up my spreadsheet with all my gambling entries (VP, slots, promos, BJ, expenses, etc.) and added up only the BJ sessions....and I was up something like 10-15k (I don't remember off the top of my head, but was in that range). I was astonished. I re-verified all my entries, made sure I wasn't adding extra 0's and all that stuff, and sure enough, even though it felt like all I was doing was losing, I was actually up.
If your results were so astounding such that simple empirical data showed you rolled more 8s/10s and fewer 7s, you would have had a super massive edge and have absolutely incredible influence on the dice. It's been my understanding that dice-influencers can only alter the probabilities by a few percent and not some huge amount which would be required for your empirical data claim to be of significance.
It'd be like flipping a coin then coming to the conclusion using empirical data (and not tracking it) that the coin is biased. For this to happen, you're not going to be able to notice if it's 51/49 in favor of heads or even 55/45 or 60/40. There would need to be a significant difference between the probabilities, more like 75/25.
It just astonishes me that someone would not keep track of their rolls.