First of all, there is likely NOT a "popular consensus" as to what the tipping point is -- and this is not surprising, since there is not a point in time or number of hands where you can suddenly say "I'm there" -- you gradually play more and more, and in the process, go further and further from the short run and closer and closer to the long run, whatever those terms mean, with no "line in the sand" where one crosses over, ie, no tipping point.
I think what people are talking about when they use the term "long run", is whether they are "pretty likely" (which requires definition, see below) to get the expected results or "close to it", which also requires definition.
Therefore, I think you are free to create the parameters for a definition and then you can come up with a number of hands within those parameters.
The first thing to define is what constitutes "pretty close" to the expected outcome. For me, I would want to define that as being, for example, within plus or minus 0.1% from the expected result for my machine and game, that is, I would want a pretty good likelihood (still to be defined) of getting the theoretically expected outcome within 0.1% (this requires that people understand that there are ALWAYS results that are NOT close to the expected outcome, no matter how much you play, but they just constitute a progressively smaller portion of all outcomes as you play more and more). I'm talking 0.1% as a figure for my return, not as a 0.1% variation from the expected return, but with a return that's close to 100%, there's not that much difference in the two -- although it could make a difference in how many hands you end up calling "long run".
So, if I'm playing Jacks or Better 9/6, my expected return is 99.54%, and my definition of "pretty close to the expected return" is met if I'm between 99.44% and 99.64% (note that 99.54% plus or minus 0.1% of that, or plus or minus 0.09954%, is almost the same as my 0.1% definition). If I'm playing a positive expectation game with an expected return of 100.1%, I would want to play long enough to be between 100.0% and 100.2%. Note that there's an underlying assumption that I'm equally likely to make a little more than the expected return as I am to make a little less than the expected return, which in fact is probably not the case.
Secondly, once you've defined the goal (what does "pretty close" mean) for your results, you MUST define how likely you want it to be that you achieve that goal. If you want a 99% probability of getting that goal, then there will be a fixed number of hands for that game that will produce a 99% probability of being in that range you've defined as "pretty close" to expected outcome, and that number of hands is the "long run". If you want a 95% probability of getting that goal, it will not require as many hands to be a "long run".
So "how many hands" can't be answered absolutely, only within the parameters of a definition of the outcome you want to achieve, a definition of how likely you want it to be for that definition to be achieved, and of course, the expected return of the game, and its variance, must be entered into the calculations.
As to how to proceed from there to do the analysis and find out exactly what the figures are, I have no idea how to do the math.
Again, most people use the term "long term" to describe a period of time (or number of hands, being a better way to look at it, of course) which, if they play that long (or that many hands), they are "pretty likely" to come "pretty close" to the expected outcome of their machine.
Those two terms in quotes require a numerical definition if you want a number of hands to attach to the term "long run".
--BG
===================
> 11a. Short term vs Long term
> Posted by: "Frank" frank@progressivevp.com
> frankkneeland
> Date: Sat Jul 2, 2011 3:19 pm ((PDT))
>
> For an analysis I'm currently doing I need a solid
> definition of Short term vs Long term play. Since I'm
> unfamiliar with these concepts I'm having trouble locking
> down the meaning either on paper or in my mind.
>
> Short Term = How many hands?
>
> Long Term = How many hands?
>
> What is the popular consensus on the tipping point?
