OK, so I didn't want anyone trying to explain probability to me, I get enough of that from my degree. I know to swap doors so I don't get the goat. What I was hoping for was data, experimental or anecdotal to justify (or not) me spending time on this.

Also, Syntaxlies, over a large enough sample yes you can. 27 melds isn't enough, but I think if I took the time to do about 100 at a fixed %, the expectation and meld success % wouldn't coincide well.

I mean, considering they state %s on the melds, my current results seem to disagree with the given success rates on 5% and .01% confidence levels....however, the sample size, etcetera.

Update: 0/14 on the 22% success lol.