Being good at "strategy"

After we "invent most of the things" then Uber comes out with a new update or we install a new mod.

Based on Uber and the Community's plans for constantly updating and improving PA, there will likely never be a time where we have figured out all of the "new" strategies or options. By the time we think we have something figured out, there will be something new that completely changes the game.

That's one of the things that excites me most about the future of PA.

This is great that the topic has been raised and PA tries to solve it. The side effect is unfortunately early game unfair imbalance which i've described above. This is not a good fix.

^This is nice post from other topic. Adding more strategic options makes the "inventive and creative and adaptive" phase longer. It doesn't solve the problem of memorization, it just delays it and forces you to remember more.
That's why I leave most of RTS games after 1-3 years. Basically you have 2 phases of RTS game development. Creative and inventive people push boundaries of strategies, invent new build orders, new adaptations, new army movements and they build up huge amounts of knowledge required to play the game, they solve early game, mid game and then eventually the lategame too. Then you play the game of memory and not too much what you call "strategy". There is always limited number of options, the more competitive game is, the faster they are discovered. As you see even chess have their limits.

Careful, I am not saying that the player has to have memorized the perfect solution to the _exact_ situation he is facing right now. If that is possible like it is for world class players in chess it's of course the best thing to do, but I don't think we will be able to do it in PA. However we will be able to know similar situations. That's something that a General is thought as well: Look at your basis of known situations and their known solutions (known thanks to previous trial and error), compare it with the current situation and solve the current situation that way. This means that knowing more previous scenation/solution combinations yields a huge advantage.

Well, so we do it more like League of Legends when we constantly allow people to innovate through adding more stuff to the game so metagame never gets stale. I guess this is actually interesting solution.

World class players in Chess think there is no strategy, precisely because they have memorised every solution.

This is the problem with any static game, someone will always develop an optimal strategy/strategies over time. Which is why I think random map generation is a good idea, because it at least tries to alleviate this problem by introducing more dynamic change into the game. True, procedurally generated maps are not truly random and there is always a hard limit to how many unique maps there are, but still. This will only lead to the question of what is random. And what is memory. And what is thinking. Is there really choice in any action we take in life?

Yes, but there is actually quite a lot of analysis and thought process packed into the process of applying past experience to a novel situation. You have to identify what makes the current situation similar to your other experiences, and in what ways it is different. You have to determine which aspects and details of the current situation are important, and which ones are irrelevant. Even after you have decided on a strategy that was effective as a solution to a different situation, you have to apply that strategy to a new set of facts, which means your actual decisions and actions will probably be different than before.

Obviously being more experienced is hugely beneficial, and that is just as true for generals as it is for doing anything else well.

I see your point here. Now a question I cannot answer myself is: How does one practice that kind of thing?
The only thing I can come up with is getting to know more previous solutions and generally practice how the game is played. Which all boils down to memorization.

I think now we're using confusing terminology. When chess players talk about there being no strategy what they mean is that there is a perfect response for every single position and that you can't generalize. Strategy in this sense is one's personal model of the game that will serve as your guide for determining the right course of action in a position. Unfortunately, this can leave you easily exploited by players that try and steer away from generalities and strive for the right move for the position. That's why currently chess players try to adopt "universal styles" (i.e. no style, purely playing the position).

Keep in mind that chess having no strategy is a very broad generalization, and second of all, there is often no perfect response to any given position because that is to deny that humans are playing the game, not super computers, and human players have different comfort levels depending on the position.

However, often in this thread people use strategy to talk about decision making. And chess certainly does require a very high level of decision making, but if some people use the word strategy as a catch-all word to describe decision making in games and others use strategy in the sense that you used it, then we are at cross purposes and won't understand each other.

So I didn't read any of the answers, yet.

To me strategy is the ability to plan ahead, and adapt to the actions of your enemy.

To be honest, I would say that I use the terms strategy and tactics in the same way that I use macro and micro.
Strategy is the overall scheme you have behind your actions, tactics are the way you execute it.

That basically means that everything you do is part of your strategy (besides doing mistakes, they're not).
May it be building a power plant, or an army of tanks. Not the execution, but the reason why you do it is strategy to me.

What is difference between strategy and decision making in your opinion then?

Well, strategy can be a confusing word, but I guess there is personal style vs decision making? Decision making would be about what is the right move for the situation while personal style involves how you arrive at your decisions.

To give a starcraft 2 example, since I recall you said that you played that game on a high level, decision making is an attribute of the game and would involve there being situations that require some sort of decision such as whether you should build drones or units, or whether you should engage or not engage. Sometimes it's unclear, but often there is a correct answer, and your decision making abilities are judged by how often you get it right, so to say. Personal style however is an attribute of the player, it's your model of the game and how you arrive at your decisions. So if you are more comfortable defensively because you know more defensive timings and that's what you practiced most and because you want to take the game into the late game to exploit your mechanical edge on the opponent, then that is part of your personal style.

In chess it would be: do you prefer to have an open line for your rook or a good spot for your knight? Do you mind your king being vulnerable if you can win a pawn? The correct answer depends on the position, but depending on the player they will tend towards the answer that more closely fits their own outlook on the game. That's one of the advantages computers have, actually, since they can be less biased while human players always have some preconceptions.

A "strategy" is more than one decision. A strategy is an extensive set of plans that the player uses to make individual decisions. Any plan that would allow an actor to react to any situation in a game is a "strategy," even if it is terrible.

To illustrate using a stupid game, in Rock-Paper-Scissors, "Always play Rock" is a strategy. I can articulate that strategy to a person, or to a computer, and that actor is capable of playing the game. As a strategy, it is terrible, but it technically meets the requirements of being a strategy.

In this example, a particular "decision" would be a single round of RPS. And, following the always-rock strategy, the actor would always play rock. Which would very likely cause an intelligent opponent to recognize a pattern, and change their strategy to the equally bad but situationally useful "always-paper" strategy.

Most good strategies contain many layers of contingencies and conditionals, in the form of "if-X-then-do-Y" elements, as well as general principles and guidelines that, together, create a strategy that tends to win games. Conditionals like "if the opponent seems to be playing a lot of rock, then play paper" to again use the very simple, silly game of RPS to demonstrate the concept.

Actual RPS is too simple to be a game of strategy, with the optimal strategy of a random distribution being rather obvious. But in complex games, strategies can be very complicated and vague since they must handle a wide variety of different situations. For example, in Go, proverbs and general principles like influence are used to guide individual decisions in novel situations because a brute force analysis is infeasible for a human.


The reason why this is such a confusing subject is because in a strategy game, players frequently change their strategies on the fly. So there is some confusion about "strategy" as the player's selection of strategies in real time, and "strategy" as the plans that are being created, improvised, and discarded in real time.

Can you demonstrate this?

Copied from other thread:

Which is fine.

Lot of people seem to mix the two together. So I thought it could be good to clarify that.

Their ability to think into the future and plan accordingly.

Exercise your creative and critical thinking skills.

The ability to make plans. That is strategy.

Not follow plans, make them. Knowing previous plans can(and does) help in the creation of new ones, but the fundamental ability to make a plan is what strategy is about.

Procedurally generated things are great for testing those planning skills.

Right it is the tactician that deals with the here and now.

You can make plans on the fly. Those are generally called short term plans.

Thought he's right, in that tactics generally matter in the short term, while strategy rather revolves around long term plans. There can be short term strategic decisions, though.

Basically, if I say "take that bunker", then it is definitely a strategic decision. But at the same time it could be considered a tactic as well. Even though there are strategic decisions, usually long-term, that can't be considered tactics.

Edit: Basically tactics are usually part of the strategy.

I don't think there are non-strategic decisions, since everything you do in a game is part of some larger approach. But something can be a mainly tactical decision, tactics are about securing objectives defined by your strategy.

I think this is like asking how does one become more intelligent. Intelligence is not the same thing as remembering lots of stuff, even though memory and intelligence are obviously intimately linked. Strategic intelligence is in my opinion like what ledarsi described.. a creative act of synthesis and the application of that synthesis to a new situation. And how do you become better at that? Same way you become better at riding a bike or anything really: by trying to do it. I'm not sure there is any better answer than that, without delving deep into metaphysics.