As we all know, Rock-Paper-Scissors is a game where each option defeats one other option and loses to one other option. This has the important effect that regardless of what you choose, you don't have an advantage before the game starts (you will win if you chose -eg- rock and the other player chose scissors, and lose to paper if you have chosen rock).
That said, due to a linguistic ambiguity (how the game's terms translate to Greek), in some areas (including, luckily, my old neighborhood) the game got a fourth option. It was "glue", because "paper" got translated as "sheet of paper", and the greek term for "sheet" is homophonic to the greek term for glue: κόλα and κόλλα respectively.
So this variation of the game has four options, but as can easily be seen it leads to imbalance, because you cannot have equal number of win/lose if the remaining options (you choose 1 of 4, 3 remain) aren't perfectly divisible by two. So you end up with one of the options winning against more enemy options.
To be more illustrative:
Now, as far as I recall, since the children weren't aware of either having thus made up a new option, nor that the dynamic changed, the option Glue defeated the objects one would realistically expect it to. It defeated two (paper, for it folded it, and scissors, because it glued together the dangerous edges) and lost to the rock (it got crushed). But do notice that Rock also becomes more powerful than the other old options, cause it has two options to defeat. In fact Rock becomes the most powerful option, since it also defeats Glue, the other powerful one.
While one could choose to play as Rock (or Glue) and still lose, the possibility of a win would be higher. But then something important happens: people start to notice the difference in dynamic, which causes them to be more weary of choosing Rock, and even wearier of choosing Glue. So by the end Glue is almost never chosen, while Rock reverts to being a regular option which risks defeat by paper, that in time becomes more popular. Scissors, on the other end, in time became the least popular option to choose, for reasons we can gather from the above summation... (while it defeats the popular Paper, it will lose to everything else and starts as less popular than the other 'lose to everything else' option: Paper).
I think it is interesting, cause it goes to show that what nominally is the best option, may become unpopular if enough calculation takes place, which calculation takes into account not which option had the biggest probability of winning but which option defeats that option.
Perhaps of more practical interest is that such dynamic elements can be used (eg in a formal logic system) to push problematic (here symbolized by "overpowered") elements out of the way, while still making use of the dynamic of the default (pre-altered) system.