Trying to account for random occurrences in chance by looking for swings in the probability of success or disaster when you can't predict a Black Swan occurrence is really futile. The best you can hope for is to limit your disasters or losses and protect your successes or gains. You do this by setting limits and keeping within those limits or at least protecting part of your successes while hoping for a Black Swan miracle.
There are those who advocate continuing to play when you are on a losing streak, because you know you are playing correctly and those who advocate cutting your losses, because psychologically you're just playing head games with yourself. I think the smart thing is to at least take a break to regroup or reassess your play and the players in the game.
Whether it's playing poker (cash games) and looking for your hand to improve, your stack to grow or playing roulette and praying your number hits, if you're calculating the probabilities of success or failure, set limits on both. Losing all your chips is better than losing all your cash and it's always better to limit how much of your profits you are willing to give back, when the tide turns after being on a winning roll.
(The Black Swan: The Impact of the Highly Improbable, by Nassim Nicholas Taleb)
I Never Bluff