In principle, if you observed the isolated gas-filled container for long enough, you would occasionally see a state in which all gas molecules were congregated on the right side of the box. But you would likely have to wait for an exceedingly long time, well in excess of the expected life-time of the visible universe, before it happened for the first time. That is because the number of possible states with no obvious grouping of the molecules vastly exceeds the number of possible states in which the molecules are observably grouped.
As a very simple approximation, consider a box containing a very diffuse gas consisting 10^9 molecules. Take the reasonable approximation that the probability of each molecule being on the left or right side of the box is independent of the positions of the other molecules. Each molecule will be on the right side approximately 50% of the time, so the probability that all the molecules are simultaneously to the right side will be 0.5^(10^9), or about one chance in 10^(10^8). That is an exceedingly tiny probability.
As a more graspable example - if you properly shuffle a pack of cards, then every ordering of cards is equally likely, including the one where the cards are all ordered by rank and suite. But the chance that you will spontaneously sort them into that specific order is very small - one in 52!. Of the remaining 52!-1 possible sortings, most of them will appear 'random'.