"Gambler's Fallacy" is a term used to describe the belief that a certain trend will continue just because that has been the experience in the past.  Some poker teachers and authors teach that this belief is correct and that you should therefore respond to a string of bad luck hands by tightening up on your play or folding down continuously until the cards show you that the trend is breaking up.  Other authorities on the matter correctly point out the fact that since you can't determine what will happen in the future by looking at what has happened in the past in the case of a random series which does not have the attribute of "memory",(such as the random series of cards which are dealt in Texas Hold Em), then responding to bad luck in this way will only your hurt your game in the long run.  It would be interesting to see the results of studies in which both methods of response were tried.

My own feelings about the matter are this:  From the point of view of experience, there seems to be something to the continuation of bad luck trends that holds up in ways which seem to go beyond the simple factor of one mathematical or theoretical perspective, such as the correctness or incorrectness of "Gambler's Fallacy".  I have just seen too many players whittled down to a short stack through strangely bad luck and then get beat out of their last 10$ when they are holding aces against their opponent's 4,8 off suit, to think that there is not something going on there other than the outcome of a random draw. 

I have to stress that as I say this, I am not citing "Gambler's Fallacy" as my reasoning, or that non-memory based trends tend to continue themselves, reduce themselves, or increase themselves, simply because they have happened in the past.  I can not argue against the fact that "Gambler's Fallacy" is indeed in itself false.  I am only suggesting that the continuation of bad luck trends may be affected or prolonged due other subtle factors which people have not been able to as yet define.  Testing random strings of data produced by random number generators may not be an effectively correct way to go about determining what is happening in real money situations, because in the latter you don't have one computer sitting there playing another computer, you have two or more people playing against each other with a computer in between them.  Who can really say at this point weather or not people don't have the ability to generate fields of bad luck as a result of the impact of experiencing a string of bad luck draws which started out as an independent phenomenon?  If you factor into this the changes in the attitudes and behavioral responses of the other players towards the player who is having the bad luck, and other ways in which they might be affecting probability fields, then you have a situation which is quite different indeed from the laboratory, and beyond the confines of one simple logical principle.

I am therefore leery of the advice offered by those authorities who often become incensed when it is suggested that you should tighten up on your play or simply leave a room when you are taking a beating by continuous and inordinately bad luck.  If you take the fact that the reason they do become incensed is because they feel that those who say that you should tighten up or otherwise change your playing methods, are incorrectly contradicting sound mathematical principles, then at that the very least, choosing the other option, which is just to get up and leave, can in no way hurt your game in the long run.  Since there is no reason to think that you are going to suddenly make a huge windfall after experiencing a long run of bad luck at a particular table, then what can it hurt to just get up and leave after something like that, and make that your standard response?  If you are afraid that you are going to hurt your game by becoming too tight and missing opportunities, then why not just leave?

So my own advice in terms of dealing with these kinds of bad luck situations is not to alter your playing style to adapt to it, but instead to choose the only option which we can point out with certitude is the one that, while perhaps not being helpful(maybe, maybe not), is the only one which will definitely not be hurtful, and that is just to leave before you go broke.  Take a shower, read a book, shake it off, and come back to play later on.

The other definitive and really important thing I can say about these types of situations is that many players, due the emotional impact of having lost a lot of money in a short space of time, start to play really bad after a long run of bad luck.  They are so worn down, frustrated, and angry that their attitude becomes "why not throw it all away?"  This is a big, big area to conquer when it comes to the subject of self-mastery.  You can not allow yourself to fall into this attitude.  If you started out in a room with 300$ and lost 240$ it is better to leave with 60$ than it is to leave with nothing.  In the course of your poker career you have to learn to fight for every little bit, and to conserve those resources which still remain at your disposal.  Later that same night, or the next day, you might take 20$ of that 60$ you have left and use it to win a tournament prize of 500$.  You don't know.  But you will kill any chance you have of recovery if you allow your negative emotions to get the best of you and fall into inferior and desperate styles of play due to bad luck runs.