What I find funny is RNG only certifies ( my guess ) that the random # generators Replay uses are on the up & up. Mis-ise of a RNG, can lead to un-randomness even if its 100% legit.
Even live, there is a huge difference if the table has a shuffler built in. In that case 2 decks are used @ all times… so the anomolies described here would be every other hand not every hand. As opposed to replicating live deals with only 1 deck and the dealer shuffle’n… Live there is also the possibility of allowing cut of deck, which usually is never replicated online.
Lets look @ 2 other senarios ( online )
1: as a deck is built, cards are pulled 1 by 1 ( randomly ), till a new deck is built…
2: cards are “shuffled”, and deck is built as cards are folded… including deck last…
How the code treats those 2 situations is very different. How “randomness” is applied is also very different, but thats not saying its fix’d.
You don’t need to be a rocket scientist to do the following experiment… take a brand new deck, replicate a “perfect” shuffle, then review the deck… keep doing this, and in the range of 13-17 times, and the deck almost shuffles itself back to where you started from… a brand new deck… so wouldn’t it make a huge difference if the code @ Replay, shuffled 14 times each time, instead of say 7 ??? Things that make ha go hmmmmmmm.