In a shoe of 60, I don't bet every hand. Probably only 20-30 out of the 60. Of course there will be shoes where PLAYERS more than BANKERS. It happens, but statistically, BANKERS will still happen more.
I will not move to another table. But just sit and wait for the next shoe. This not only slows down the pace but also slows down the losing(if any). I normally play in a way where I spend 2-3 hours and then I walk around go back to the table play another 2-3 hours. The funny thing is when one is winning, he doesn't feel fatigue. Only when losing, even after an hour, feel like sleeping.
When you dry run on an online free casino, the result is useless as the cards are shuffled every hand. In a 8x52 deck table, card counting can enable when to buy BANKER. That is why I don't buy every hand and only when the counting favours the BANKER.
Even if the table opens more PLAYERS, in long run, BANKERS will still open more. That is why there is 5% comm on Bankers. My friend has calculated a way to detect Tie bets via card counting.
Meanwhile, he doesn't bet on that but just doing manual calculations while betting on Bankers. So far the Tie bets (if he has betted) would have got him favourable profits too. But as he said no point placing so many types of bets when 1 side is enough and also losses are minimized