Tuesday, March 24, 2009

Spy is Sapping my Sentry

Today I'm going to talk about competition again. This one's going to be on the fine art of information gathering.

In any game that is not a game of complete information, we gain a competitive advantage when we know more about the situation than our opponent does. The more information we are able to gather about a situation, the more certain we can be about what our opponent is doing. If we know what our opponent is doing, we can counter him.

There are games where this is not applicable - specifically, games of complete information. Since that's a big term that not everyone understands, I think I should explain it further. If we can clearly see everything that is going on in a match at once, the game is a complete information game. For instance, Chess is a complete information game. Street Fighter is too. CounterStrike, Magic: The Gathering, and StarCraft are not complete information games. If we can see the whole battlefield, or at least all of the battlefield that matters (eg. fighting games; we can't see all of the map at once, but only the part with the characters actually matters) then the game is a complete information game. Although technically Smash Brothers can be a game with hidden information if the two characters are close enough that there can be hidden things offscreen like items, it is close enough to a complete information game that we can lump it in there too - especially since if there is something hidden, neither us nor our opponents actually know.

Magic for instance is not a game of complete information. Each player has cards in their hand as well as their entire deck that is unknown to the other player. In addition, players can play cards face-down in some cases, and do other things that hide information from the opponent. The value of hidden information is so good that in Magic there are cards that let you look at the opponent's hand or deck but don't do anything else. Clearly in Magic there is a big value in knowing the opponent's hand or what he will draw next.

The value of information varies from game to game. The value of knowing people's hands in Poker for instance is gigantic. It's so good that there is no way within the game to see what people have until you've made all of your decisions. In Magic, the value of your opponent's hand varies a little bit depending on the game situation, but it is never as valuable as knowing your opponent's hand in Poker. In StarCraft, the value of knowing your opponent's tech and expansion status is generally huge, even compared to the opponent's hand in Magic, although not quite as valuable as the opponent's hand in Poker.

In any game where our play is determined by what the opponent has up his sleeve, it's best to get as much info about them as possible. That's where spies of various types come in. In TF2, spies are useful if they communicate with their team. They can tell the allied team all sorts of useful things, like where enemy medics are and how many there are, the locations of enemy sentry guns and dispensers, and things like that. I think this is sort of undervalued by the TF2 community, but knowing that there is an enemy medic and pyro team ready to charge your position is probably pretty useful information. Even if your spy never gets kills or whatever, getting that information is very huge.

One of the things that I think is sort of undiscovered by players in lots of games is the value of information. I hear a lot about things like covert tactics being frowned upon in EVE in favor of more conventional warfare, and it makes me sort of sad. People treat covert spies such as using newbie alts to scout areas as inconsequential, where nothing could be further from the truth. If you scout an area in EVE and are destroyed by hostiles, you have a huge amount of information about the composition of the enemy force. Anyone who participated in killing you will be displayed in the 'kill mail' you receive when your ship is destroyed. You can then relay that information to friendly players who then know the exact number of enemies and how they are set up. This can be invaluable information, but it seems to be totally underrated by the EVE community.

One community where information has really been greatly understood and coveted is the StarCraft community. Scouting is a huge part of the game. Scouts typically occur as early as one minute into the game and scouting is a continuous thing, taking tons of mental effort from players who might be otherwise building armies or attacking. And yet, the community all agrees that scouting is a massive part of the game. I've seen numerous battles where a player completely turned the game from an absolute loss into a victory just because they scouted the enemy early enough to prepare a defense and avoid losing, and eventually gained a material advantage because of it.

I know I mention the game a lot, but StarCraft is one of the most well-understood and developed games in the world. People know a lot about how to play StarCraft well and have turned it into a science that is now worth tens of thousands of real dollars at any major SC event (in Korea, at least). Because the game is so well-developed we can learn tons of things from it that can then be adapted into other games.

When we have no way of gaining real information though, every subtle clue helps. We can't always send an invisible Observer into the opponent's base or fly a newbie ship into the enemy fleet just to find out what all of their guys are flying. Some games don't let us do that, or in the case of Magic, being able to see what is in the enemy's hand usually comes at a pretty big opportunity cost compared to drawing a card that can do some game-winning thing. This is why we need to practice reading tells and predicting enemy tactics.

Tells are the Poker way of saying that a player might give away what is in his hand by how he acts. It's a trick that I can't really teach you how to do, but in most cases it is possible to tell whether a person is lying or not if they are visible (not hiding their face in some way) and talking at all.

What I can do is tell you how to predict people, at least a little bit. I'll go into this in great detail when I talk about mindgames at some point in the distant future (I have a lot of other topics to cover before I can really explain them) but for now, I can give you the basics.

In a given situation, we can group people into playstyles and such. If people have a reputation for behaving a certain way, we can use that. Otherwise, we can make smart guesses based on how good we think they are at the game, and what a typical person would do in their situation. Once we have a guess as to who they are, we can guess as to what they'd do. If we think they are not very smart, we might guess that they will do 'noob' things like build lots of turrets in their base early, or 'call' bets every time or something. If we have multiple chances to see what they will do, we might be able to reinforce this a little more. For instance, if we play a balanced style of play at first, we might see that the opponent always plays defensive early, or the opponent does lots of high risk, high reward moves, or the opponent only raises when he has an ace in his hand or something. We can use this to our advantage and play to counter their play.

Yes, I realize that this is common sense, but the goal is to read the opponent and understand that humans are creatures of habit. This is because we chunk things that we know, and we tend to follow these chunked patterns unless we learn something else. Therefore, if we know what patterns the opponent knows, we can counter him.

Ugh, I know it sounds kind of complicated. I'm not really sure that makes sense, or even if it does, I'm not sure if it's not common sense. Maybe it's not? I don't know.

I'll talk more about viable options and tempo and advantage later on though, and maybe that will help. I don't really have enough time or space for that now, but I'll try to get back to it. I'm kind of irritated talking about GMing anyway, because I feel like too many people will hate me for what I feel is the best way to sheperd players, ha.

No comments:

Post a Comment