Comments on: Gambling and Strategy

By: SadMonk

SadMonk — Thu, 29 Jul 2010 22:47:53 +0000

Ah okay… gathering information in RTS in this form is some kind of meta-game… browse forums for buildup-information, read the (f’in) manual to get information about your unit stats and make up some plan how to build up your units with maximum efficiency in minimum time.

But okay, thinking about browser games, you’re probably right. The important factor in this is that the outcome never is deterministic, even if the game mechanisms themselves are. Each player for himself is additional game content to the others as they usually can’t predict how the other player is going to react on other actions. If a player is a lone wolf he may be oblivious to your attacks. If he’s in a strong ally, he may call his friends for retaliation. He may start insulting you, start guerilla warfare or try to befriend you hoping for a strong protector. This is something that would be quite hard to perform by some AI…

So to get back to your “calculated gamble” theory… in Browser games maybe there is some information you can gather, regarding the opponents resources and possibilities but it might be hard to forsee his personal reaction to your moves. So this often ends in “just try it and let’s see what happens”… at least for me

By: Martin Nerurkar

Martin Nerurkar — Thu, 29 Jul 2010 17:34:20 +0000

Another good point. And yes, I was thinking about browser games primarly.

That said, yeah in a real time game you’ve got less time to “gather information” but you still need to calculate risk/reward. You need to decide if you’ll be attacking him now or later, whether to invest in your production or your army. There’s less information about the opponent involved though.

By: SadMonk

SadMonk — Wed, 28 Jul 2010 18:37:53 +0000

So when we talk about weighing risks and rewards, getting a lot of information etc, I suppose we aren’t talking about any real-time strategy games, right?
Gathering information usually takes quite some time, so this would be something you would rather do in some turn-based game. In RTS games you will usually have to make quick decisions, use more tactics than following some grand strategical scheme or some opponent might just be faster and outmaneuver you.
In RTS this is exactly the thing that thrills you… so that might be more of a motivation than weighing rewards and risks.
But if you play something turn-based I think the interesting thing would be the planning involved in making your next move (and beyond that)… so yes.. weighing pros ans cons of your next actions.

By: Martin Nerurkar

Martin Nerurkar — Sun, 25 Jul 2010 20:52:17 +0000

Well alright, you do have a good point there:

There might be cases where you have perfect information about a risk/reward situation. At least when seen from a tactical perspective. In these cases there’s often a strategic layer that puts the tactical option in a greater context. Then it’s less about the individual choice and more about the clever allocation of resources. This extends to your chess metaphor:

As for chess, there’s no imperfect information about the individual moves: You know beforehand which move is possible and what it’s direct results will be. However there’s imperfect information as to the plans of your opponent. Maybe his move is a feint and he’s setting you up. Here you gamble that your move is the right one to counter your opponent’s strategy but you can’t ever know for sure. That’s the central risk/reward part here.

By: Mayec

Mayec — Sun, 25 Jul 2010 19:30:42 +0000

“However in the end there should still be a certain sliver of insecurity before you make the gamble.”

Usually, but not always. A very few strategy games still present you with absolutely all the information you need to take your decision. Their appeal lies in complexity rather than uncertainty. That is, the high amount of possible combinations that a certain action will create. I’m talking about (You probably guessed it already…), best known example, chess.