Though I am tempted to be flippant, this question is fascinating because it becomes a real issue in almost every scientific experiment and sexual orientation survey. That is because the vast majority of men raised in Western cultures (upwards of 95%) have some male-male experience in their lives. (In fact, though I don't have the survey data with me at the moment I think I recall that a slight majority have at least one male-male experience
after their teenage years are over.)
It really starts fuckin' with the data when you apply the kind of definitions that Rugbyscott is using, because you end up with almost no "straight" control population with which to compare "gay" men.
Instead of facing what this really means, which is that almost everybody is bisexual to some degree, we persist in trying to draw hard and fast lines between straight and gay.
However, treating to OP's question with some dignity,
most surveys and most experimental protocols use a five year cut-off period and treat you as "str8" if you have had no homosexual contact in that period regardless of what you may have done, no matter how many times, and no matter for how long before you changed behaviors. (Not defending this position, which I find astonishing, but when you read the propaganda that says the gay population is 4% or less of the population, you are seeing this definition in place.) In the world of sexuality research, at least, your teenage experiments
don't count.
The biggest exception of which I know has to do with blood donation. In the US and Canada if you have had male-male sex (any activity involving the actual or potential exchange of body fluids) since 1978 you are ineligible to give blood. In that case, at least in theory since in practice everyone lies, your adolescent experiments
do count. Fortunately this is not a problem since the blood supply is so overstocked that no more donations are needed...