Today’s random number generators use such a wide variety of inputs that, not only is it impossible for people to determine how the RNG is going to act at any given moment, but it also offers enough of a variety that you will not see biases like the coin-flipping example above. However, when it comes to betting on real money games, you want something that’s a bit more robust. For many applications, that degree of error isn’t a big deal.
If you have someone who flips a coin to determine the outcome of an event, that might seem pretty random until you realize that they flip the coin the exact same way each time, so you might end up with 55 percent heads and 45 percent tails instead of a true 50/50 split.
There are different degrees of randomness, and some of them are better than others for gambling purposes.