The Let's Play Archive

Civilization 2

by Melth

Part 14: Mechanics: Some techniques for good strategic thinking in general

cheetah7071 posted:

Out of curiosity, do you have a rough idea of how long it takes you to achieve this level of mastery in a strategy game? Do you have to pay your hundreds of hours of dues, or do you make a concious effort to think at this level of depth from the beginning to cut some of the learning curve out?

It's a little hard to answer that. At the risk of sounding boastful, I think the short answer is no, I'm very good at most strategy games immediately and can beat experienced players in multiplayer or win on the hardest difficulty singleplayer. Like anyone, I do get much better with practice, but I start well. It's a big surprise to me when I can't beat someone at their favorite games on my first try as long as I've been allowed to read the instructions.

In the case of Civ 2 I started playing when I was a kid with no instructions. So game 1 I didn't even know how to move diagonally and whatnot for a while, but I think I scraped together a win since it was only Prince difficulty or something. Once I actually knew what I was doing, it wasn't really hard to jump up to Emperor or so. Making the jump to successful Deity took me till game 3, and I had to read some articles on the actual mechanics of the game because the in-game instructions were totally unhelpful about the details.

You're basically correct that being good at games on the first try is all about making an effort to think deeply from the start. I'd add that strategic thinking is a skill one can get better at and a habit one can get into. I pretty much try to think strategically about everything all the time and that helps a lot. It also helps to have a strict policy of always playing on the hardest available difficulty and to try to win with self-imposed handicaps in multiplayer (Keep them secret so you don't hurt anyone's pride). You get better when you challenge yourself. I think I'd also add a few general tricks, principles, and helpful ways of thinking I employ in no particular order:

1) Opportunity cost. There are some very interesting economics concepts which are incredibly powerful for approaching strategy games as well as real-life decision making. The absolute most important is "opportunity cost". If you are not thinking about opportunity costs, you are thinking wrong. I've talked about this idea before, but to sum it up again the opportunity cost of doing something or using a resource for a particular purpose is the cost to you of not doing the best possible alternative instead. Don't ask yourself if, say, paying to speed-build a Temple now is worth the price in gold, ask if it's better than anything else you could possibly spend that gold on or save it for. Don't ask yourself if you'll benefit from taking an opponent's pawn, ask yourself if you could benefit more by taking another piece. Don't ask yourself if you want to have the Great Library enough to spend a bunch of turns/Caravans building it, ask yourself if there's another wonder you'd rather have more. Don't ask if Guy can beat the enemy, ask if someone else can beat the enemy more effectively.

I see loads of people putting forward clearly terrible strategies and tips and such in all kinds of situations because they aren't considering that there are much better options available. They're forgetting the opportunity cost.

2) Think from the opponent's perspective. The simplest version of this is noticing the quirks of an AI and exploiting them. For example, the Civ 2 AI just LOVES to move its units, and will move them even when staying put is better. GBA Fire Emblem enemies have no concept of miss chance and can thus be manipulated into making totally pointless and suicidal attacks. The next-simplest version is trying to figure out what your opponent's moves will be several turns in advance. Where things get complicated is when you need to consider what the opponent actually knows and what they can guess and try to figure out what they'll do based on that. And where things get really complicated is trying to identify the irrationalities, preferences, and goals of human opponents that will make them behave differently than you would in their position.

3) Play the opponent, not the game. This is related to the above. One thing that always amazed me in game theory classes was how my classmates would foolishly assume their opponents would actually play rationally. You can lose by being too "smart". A lot of people who are expert at a game fall into this trap by thinking their opponent will actually play in the "best" fashion. This is really a big deal in some collectible card games where "great" players spend vast amounts of time and money building decks that perfectly counter the current meta and then get flattened by some guy who made a deck that's actually just good.

This is particularly important in games with more than 2 people. In a 2-person game an opponent playing sub-optimal strategies will typically just lose to you. In a 3 or more person game, not being ready for one player playing "badly" can allow the 3rd person to flatten you. For example, in FFA Warcraft 3 games it's generally optimal to sit back and build up and try not to aggro any opponents. If you expect your opponents to also do that and aren't actually ready for one to stupidly attack you, you can lose the game even if you defeat that person.

4) Own your defeats, concede your victories. Even if you lost by luck or the like, it's important to think about how you could have played better. Sure the proximate cause of your defeat might have been that a 90% thing didn't go your way, but could the odds have been 99% in your favor instead if you were better? If you lost to a human, ask them how they beat you. I learned a fair amount about the competitive scene for Warcraft 3 that way. Granted, a lot of people won't give you a helpful answer, but it's worth a shot. Similarly, if you win, acknowledge how good luck and so forth played into it and think about how you could have increased your probability of winning.

The least helpful attitude you can possibly have is to blame all your losses on bad luck or wild accusations of cheating or whatever.

Of course, it's perfectly legitimate to criticize a game for having too much luck or whatever in it, but it's not helpful to you to just blame that for losses and not think any further about them.

5) Reason. This is something I talked a lot about this update. Being good at analyzing intelligence is at least as important as being intelligent in most strategy games where information is scarce. It's important to go beyond just the facts you learn directly from scouting or other information gathering. Think about implications of those facts and make some inferences. And go beyond that to consider things which aren't quite certain. What do the facts PROBABLY mean? Do some Bayesian reasoning.

6) Know ALL the rules.

7) Whenever you look at some option in the game and think "This is useless. Why would I ever do this?", stop and think hard about that question. Why would you ever do that? Under what circumstances could it actually be good? Can the weaknesses of the option be turned into strengths in some way?

This kind of thinking is where a lot of my cleverer and more surprising tricks and tactics come from. It's finding ways to make the useless into the incredible that most interests me.

8) If you can't explain it, you don't understand it/the Canas approach. One of the best ways to learn about a game and how it works is to try to explain it or talk about your favorite strategies and why they work to someone else.

9) Read other people's strategies and such, but think critically about them. That means you guys reading this right now! Don't assume I actually know what I'm talking about. When you look at someone's replay or a let's play they did or an essay on strategy they wrote (like this one I read early in my Civ 2 career:, think carefully about what they say are good ideas or about what they do. Figure out mistakes they made or room for improvements. And make sure you don't mistake confidence for correctness. That "Fire" guide is factually and tactically wrong on pretty much all points.

Dead Reckoning posted:

I'm still unclear on how you won this with no casualties. Since the archers are guaranteed one square of movement per turn, it seems like you wouldn't have enough time between archers to heal your horsemen enough to avoid having to take the last one on at 50/50 or worse. Maybe I'm too used to Civ 4.

Well I have 2 turns before the next Archer arrives. Then 3 before the one after that. 3 turns in a city is often enough to be fully healed. I say often because it depends on all kind of factors like unit type. And since my Horseman is now veteran, he can just smash the Archers when he attacks. Even if healing WAS a problem, I have a second veteran Horseman standing right there. This fight was easy. It would have been nearly risk-free if I hadn't forgotten that one of the two Horsemen wasn't a veteran to start with.

Tiggum posted:


Got one of those already.

Delenda est Carthago.