science of board games

I have not been playing many games lately, unfortunately. Thursday sometimes wrangles me into a few hands of Gin/Rummy occasionally, though. At first I wrote the game off as silly and totally random, but as I have been playing it more, I have noticed some more subtle strategies in it.

For example, one intuition I initially had was to throw away worthless cards from my hand. This isn't always a good idea: the cards may be worthless to you because another player has the rest of that set. However, if you throw away, say, half of your pair, then it is significantly less likely that your opponent has the other pair needed to make that a threesome (careful that your opponent doesn't have a small straight to stick that on, though).

Another tidbit: if it is still early in the game, and your opponent just lays down (in a straight) a card you need to complete your three-of-a-kind, then you might as well start throwing those two cards away: there's only one card in the entire deck left that can complete your hand, and you may not ever even get to it (this will change depending on how many cards are left, and maybe knowledge of how long the "average" game takes).

I am very tempted to start up a simulation of Gin/Rummy with some simple strategies just to determine things, like, what is the average length of a game. Also, maybe to determine some real-world confirmations of the above strategies. Maybe if I get bored.

Don't get me wrong: I still think the game is extremely random, but I can still enjoy it.