Theory On Strategy

Mostly by Lindir The Green

Feel free to add stuff/rephrase stuff/properly format stuff; if I don't like it I will just revert it :)

Basics of Strategy

In real life, strategy is trying to figure out the "rules" of the universe and then using them to accomplish something. But in a game, the rules are (usually) already known. So strategy in a game is using those rules to accomplish an objective.

In reality, strategy mostly requires observation.

In a game, strategy mostly requires logic and creativity.

In a multiplayer game, the new element of trying to outguess your opponent is added. The simplest multiplayer game is Rock/Paper/Scissors, which requires no logic or creativity. It is pure multiplayer strategy. The only differences between the rock, the paper, and the scissors are that they come in different orders in your mind. So if you are playing an opponent that you have a low oppinion of, you usually use paper. But if you are playing an opponent that you have a slightly higher opinion of, you usually use scissors. And so on.

The only way to predict what your opponent is going to use is to try to guess what your opponent thinks you are going to use, based on how smart you think that your opponent thinks that you are.

Almost every multiplayer strategy game involves figuring out what your opponent is going to do, and then reacting accordingly.

The multiplayer game that requires no multiplayer strategy is Tic-Tac-Toe. If you logically think about the game for 10 minutes, you will become unbeatable.

Most multiplayer card and board games are a cross between "normal" strategy and "multiplayer" strategy. But if you completely understand the game (like some chess masters, and me at Euchre) the game for you becomes only "muliplayer" strategic.

But some multiplayer board games also require diplomacy, which is just using social skills to your advantage.

Magic the Gathering requires insane amounts of "normal" strategy, moderate amounts of diplomacy, and almost no "multiplayer" strategy.

Roleplaying games are different though. They mostly require creativity, with some "normal" strategy if the players are competitive. But, since there are many different games and many different DMs, there are many exceptions to the rule.

Strategy in Video Gmes

In video games, the rules are much much more complicated, and in non-turned based games, the elements of reflexes and hand-eye coordination are added to the brew. This complexity of rules is, of course, secondary only to the complexity in some roleplaying games.

I think that even the most hard c0r3 action games have more strategic depth than any board game, because of their extreme complexity. I would estimate Jazz Jackrabbit (my first computer game) to be at least 20x more complex than the nerdiest military combat board game.

Conter Strike is probably around 500x more complex.

The reason most people don't think that action games are strategic is because the strategy is "diluted" by action, a common myth. Action simply places strategy in a time-limited frame.

Most video games, even if all the rules are known, require lots of observation to figure out how they all interact.

But once you figure out (or are told by a walkthrough), the game stop requiring any "normal" strategy, and become pure hand eye coordination and reflxes. But that level of understanding would only be possible by a computer program.

Multiplayer video games are often really just extremely complex games of r/p/s, but with significant differences between the different choices. The realism of the games makes it easier to figure out how the r/p/s arrangements work; it is easier to remember that early anti-air defense beats an early air raid thant it is to remember that if a construction unit builds a certain type of building, that building will make the fast units from a certain type of building less effective.

The RTS

The Real Time Strategy game is IMHO the greatest genre ever, because a (good) RTS is so intellectually stimulating. Someone playing a good RTS is trying to come up with good unit combos, trying to figure out what the enemy is doing, trying to figure out how to best counter what the enemy is doing, trying to keep track of all units and how the overall battle is going, and looking out for emergency situations that require immediate attention. The winner has the best combination of:

• "Normal" Strategy
• "Multiplayer" Strategy
• Diplomacy
• H/E coordination
• Reflexes
• Multitasking skillz

But, like all games with static rules, once the rules are theoretically perfectly analysed, all "Normal" strategy disapears.

However, since the rules are so complex, and there are almost infinite combinations of units, maps, and starting resources, it is impossible for a human to perfectly analyse an RTS. But, fortunately, Humans are very good at abstractly coming up with generalizations, that often carry over from the real world. For example, for a human, it is common sense that tanks are good against standard infantry, buildings, and on plains. We shouldn't even have to analyse the game to figure that out.

Micro in an RTS

Micromanagement is probably the single most fought over subject in an RTS. Starcraft players like a lot of it, TA players like less of it. But in the end, it all boils down a few simple concepts.

The more things you can pay attention to, the more skill you have at multitasking. And so the more micromanagement in a game, the higher the gap between the skillful multitaskers and the non-skillful multitaskers

There are two different types of micromanagement: "tactical" and "pointless".

"Tactical" micro is the small scale r/p/s games of maneuver. Deciding which unit should attack which and where each unit should move, for optimal enemy destruction. It involves split second assesments of the situation, and out guessing your opponent. IMO, "tactical" micro in a game = good.

"Pointless" micro is the stuff that requires little thinking to pull off. Clicking 100 checkboxes requires skill to do quickly, but doesn't require any "strategy." Telling every unit to attack a single unit for increased efficiency requires valuble brain time, which skilled multitaskers have more of, but doesn't really add strategic depth to the game. Telling all of your hawks to go forwards and backwards and forward and backwards requires a lot of thinking, but really doesn't add anything to the strategy of a game. Starcraft is filled with "Pointless" micro. IMO, an RTS should try to minimise "pointless" micro.

Note: It seems like SupComm will severely reduce both types of micro, which is IMO not very good.

Balance in an RTS

In an RTS, there are an almost infinite number of possible situations, and possible strategies. Balancing the RTS is the act of trying to get all of those strategies to work together and create the best game.

Every strategy has several characteristics:

• the strategies that it performs well against.
• the strategies that it performs poorly against.
• the difficulty of pulling it off.

For anybody to ever use that strategy, there must be some strategy that it performs well against. And the strategy that it performs well against must also be feasable, which means that that strategy must perform well against a different strategy. But to prevent a strategy from being overpowering, it must perform poorly against some other strategy.

As long as there are several strategies that fit the above in an RTS, the RTS should hold together.

But to make an even better RTS, you must pay attention to the difficulty of each strategy. The difficulty of the strategy is almost the same as the amount of micromanagement required to use it.

Some RTSs have all of the best strategies be the most difficult, and some are the other way around. But as long as a particular strategy is good against another particular feasable strategy, it will probably be utilized at some point, even if it is extremely difficult to pull off.

The learning curve of an RTS is determined by:

• the GUI.
• how effective the most difficult strategies are.
• the strategic complexity. That is, how many feasable strategies are there?

For somebody who completely understands a RTS, the more strategic complexity there is in that game, the more fun it is. But for somebody who does not completely understand how every single strategy fits together (like most of us), more strategic complexity just means less comprehension of what's happening, which usually means less fun. Unless of course you enjoy being confused.

That is why realism, or pseudorealism, is important in a RTS: To give players an intuitive understanding of how the RTS fits together. It is much easier to understand that some anti-air is useful to carry with the assault force as defense agains air raids than it is to understand that unit A should be grouped with unit B, C, D, G, J , and L when they are doing actions 1, 2, 3, 3, 2, 2, 4 respectively to prevent against trouble caused by the enemy utilizing units H, I, and N.

Here are examples strategies and how they fit together in my favorite RTS ever: Total Annihilation.

Porcing

• Good against: Attacking with units which happen to lose to the defenses.
• Bad against: Expansion.
• Difficulty: Very low.

In TA, porcing is not really a viable strategy. Sorry to shatter reality as you know it.

Porcing isn't a very good strategy in any current RTS, because if it were then nobody would do anything else. Unless there were a single strategy that worked really well against porcing but really horribly against anything else. But I can't think of anything like that, that would still retain some pseudorealism.

Going advanced at a certain time in a certain game

• Good against: Being defensive at between this certain time and the certain time when the player has made the jump to advanced.
• Bad against: Being agressive at between this certain time and the certain time when the player has made the jump to advanced.

Yes, I am using the word strategy extremely loosly, it really sorta means "decisions" or something. But I'm gonna keep calling it strategy.

more later... I have to analyse a few more strategies