Being good at "strategy"

Lol what? Right back at ya buddy

Lol what? Right back at ya buddy[/quote]
Help, stupid tablet, can't talk, argh.

Anyway, people constantly say stupid things about chess, it's tiring to read. Ledarsi's millionth didactic comment about chess from the perspective of someone that did intro computer science once was annoying too.

I think most people, even myself included for a while, were hung up on the difference between tactics and strategy. To help you out, let's look at a football game. Let's define our strategy: Win the Super Bowl.

The strategy is to score in the opponent's end zone in order to ultimately win the game. The tactic is how you get the ball down the field into the end zone. All those lines and circles the coaches draw up to position their players offensively or defensively are all tactical. Coaches and teams study football plays in order to answer tactical problems and to look for patterns in how the other opponent's defend and attack. Draft picks, deciding who is best suited to be a quarterback vs a wide receiver, and positioning players on the field to counter the opponent's strategy to get the ball in your end zone is all tactical. All of these small details that make up the means to accomplish the end (the strategy i.e. win the game/Super Bowl) is tactics. Applying this to PA would be like studying the ranges of all the different units in order to "stiff arm" and have an advantage in the battle, or combining units to be the most effective. How fast you build your eco & defenses. All tactics. In Chess: The strategy is to capture the king. The openings and moves you make that force your opponent to respond in a way that is favorable to your strategy is tactical.

There is a mid level tier in war: Operational. This is basically having separate but relevant objectives that ultimately lead to your goal of accomplishing the strategy. For example, you would need to win so many games in order to be in the playoffs, just to get to the Super Bowl. The operational level ties your tactics to your strategy. In PA it would look like this: First we take this planet, then the solar system, then the galaxy to ultimately win the entire war (assuming the war was on that scale).

Strategy is not so much just memorization as it is learning to recognize patterns and, if it isn't possible to recognize patterns, to adapt to the changing situation as rapidly as possible. In a contested environment, strategy isn't about memorizing and sticking to a plan. You wouldn't plan to make a million dollars like you plan to bake a cake or take out the trash. As Gen Moltke said "No plan of operations extends with certainty beyond the first encounter with the enemy's main strength." When you have opposing forces that get in the way of meeting your goals, sticking to a set plan is unwise. The plan itself should only be used to define the problem. The solution to that problem is always fluid and based on changing circumstances.
In the case of PA, the strategy, simply put, is to win the game by killing the commander (this is the plan). The solution is fluid because you don't always know what your opponent's plans to stop you/achieve his own goals are going to be. You respond in a tactical manner by getting as much information on your enemy as possible (intel), predict and develop answers to his courses of action, and then based on both yours and your opponent's situation, develop the means and applying them effectively to achieve your objective. The OODA loop - Observe, Orient, Decide & Act - is the decision cycle that players constantly go through to make decisions. I recommend studying and applying this theory (theorized by USAF Col John Boyd). The one who can do this the quickest often has the upper-hand in the conflict.

I believe that someone becomes good at a strategy game by studying the art & science behind it, and applying the theories and principles in game. Of course, it is one thing to know this stuff. It's quite another to master it. It takes practice. If you want to be a good strategist, I recommend reading on the ones who know it best: Sun Tzu, Clauswitz, Napoleon, etc.

If you want to be a good tactician, especially in PA, I recommend learning all the nit gritty details on the costs of all the units and their ranges, learning each one's strengths and weaknesses and using other units to compliment them (i like to call this the rock, paper, scissors method). Speed is also essential. If you can get your hands on some good IO devices for your computer and set up and use those hot keys, I would use these and any means possible to improve on speed. Develop good systems like dividing the planet into halves and trying to control as much territory as quickly as possible. Find the best methods to build and sustain eco. Remember that "invincibility lies in defense, the possibility of victory in the attack" - Sun Tzu, and finally, Intel.. Intel, Intel, Intel.

Practice. And, best of luck!

I found a notebook that I had written 6 Major Principles of Strategy. I can't remember what book I was reading that I got these notes from, it's kind of abstract, but I think these might help.

1. Win all without Fighting: Prioritize threats and determine strategic focus.
- Determine which enemy poses the greatest threat or provides the best opportunity to achieve the nation's objectives if successfully attacked; prioritize by threat - greatest to least
- Prioritize resources. What do you need to achieve your objectives?

2. Avoid strength, Attack weakness: Develop attacks against the enemy's weakness
- Once priority is established and resources allocated, focus on and determine the chosen
enemy's strength's and weaknesses, as well as understanding one's own.
- Prioritization of the enemy's weaknesses can then be done by elevating in importance those
weaknesses that, if attacked successfully, would severely degrade and unbalance that
adversary's ability to fight.
- Next, the enemy's three to five most critical weaknesses should be studied and two to three
potential attacks for each should be developed

3. Deception and Foreknowledge: Wargame and plan for surprise.
- Each potential attack must be wargamed using the knowledge of the enemy to play out
the moves and countermoves that could occur
- It is important to forecast how the enemy might leverage its strengths in counterattack
- As part of exercises, it is critical to think through how surprise can be achieved to disguise
attacks with deceptive moves against the enemy

4. Shape the Enemy: Integrate best attacks to defeat the enemy
- Select the enemy's one or two key weaknesses to exploit.
- Results of wargaming will provide insight and shall assist in which set of attacks to utilize,
and how they can best be integrated for maximum impact on the enemy. This becomes the
overall strategy.

5. Speed and Preparation: Ready attacks and release them.
- To support the strategy, preparations (orders, logistical support, troop deployments) are
required for successfully executing the integrated set of attacks.
- Attacks must be executed with speed, shock and firepower

6. Leadership: Reinforce success, starve failure.
- Once attacks have been launched, the leadership must support the strategy with prompt
action, determine quickly which attacks are succeeding and which are not.
- Starve failure. Ruthlessly reinforce success.

Maybe the question is best asked by example. Which of these games are strategy games? Are some more strategic than others?

• Starcraft II
• Battle for Wesnoth
• Quake 3
• Civilization (whichever one you like)
• Chess
• Go
• Tic Tac Toe
• Poker

The varied solvability of Chess, Go and Tic Tac Toe leads to an interesting question. Each of these games is solvable in principal because it is simply a matter of expanding the gamestate tree for every move and chosing the best. Trying to answer it could shed some light on what you think a strategy game is.

• Would access to unbounded computing power make Go no longer a strategy game?

Those boardgames are deterministic so what if you relaxed that condition. Is a game like Backgammon immune to brute forcing? I don't think it is immune because it is easy enough to carry through the probabilities when executing the gamestate tree and aim for the best probability of a win.

• Which types of games are immune to brute forcing?

Clearly Rock Paper Scissors is immune to this, at least against someone playing randomly.

If you had unlimited computing power then you could calculate all possible current gamestates based on the partial gamestate that is visible to you. This is possible because your opponent is constrained by rules and these rules are knowable. For example if you scout your opponents commander 2 minutes into the game and you have not seen a bot factory within a certain radius of that commander you know that your opponent cannot have bots.

In principal this could be generalized. You could simulate every action which your opponent can take with their units which you know they had but can no longer see. These units could cause more units to be created and those could be simulated too. So you can know exactly what your opponent could or could not have. Furthermore many of the states are inferior to others so a perfectly playing opponent would not choose to reach them. Each of these states can also be simulated forwards. You don't know which choice they took (for example they could have started spamming bombers after you scouted) but you know the states they could be in and how to best counter each. At this point your game is effectively a very complicated game of RPS.

Following this chain of reasoning there is no objective strategy game. Whether we consider a game to have strategy 'in it' depends on how the complexity of the calculation in that game relates to our metal capabilities. It is not just about calculation though, some forms of calculation annoy us because they are 'tedious' or 'arbitrary'. Other types of calculation are more intuitive for us and done less consciously, they tend to be the types of calculation which we like underpinning our strategy games.

I will now define strategy in terms of these ideas. Strategy games are games that require the player to make choices in which calculation would be very useful but is infeasible to perform. Players have to use heuristics for their choices in these games so to improve at a strategy game you should train your heuristics. If no usable heuristics exist then the game is felt to be arbitrary and people probably won't like it. People also do not like the game as a strategy game when the heuristics are too accurate.

I think this way of looking at strategy matches some commonly held opinions. For example a game which is mostly down to chance is not considered strategic. With my idea this is because the choices in the game have little impact on the outcome so calculation would not be useful. The same can be said of a game which only has very similar choices. Tic Tac Toe is not considered a strategy game because the calculation is trivial to perform. FPS games may not be considered strategy games because your skill at implementing the choice that you make is far more important than your decision.

People often say that strategy games require you to think quickly and change your plans but that a strategy is an overarching plan. This is confusing. I would say that a strategy is some precalculated heuristic which you have developed prior to a game which affects your actions during the game.

I don't know how useful this wall of text is. I rarely know exactly where they are going before I start writing them and I am not trying particularly hard to convince anyone. I may have convinced myself of a load of bollocks in the process and hopefully someone will point that out or expand upon the ideas. In any case it was interesting to write.

While I think we all agree that PA should be a macro focused game, without too much micro, I think the 'no micro' group think on the forum has gotten out of hand. People are now taking any example of micro as a bad thing the devs should remove because "PA is a macro only game". A real macro only game is not a real time strategy - we need to differentiate between good micro that adds depth and bad micro that just strains the player with unneeded APMs.

To become good at a strategy game, you need:

1. A feeling for mathematics: PA consists of many mathematics. The economy is math pure, and if a player does not understand the principle of exponential growth, he will not play well. (Although there are players, that understand the principles of math, without knowing it. So they just do it intuitively.
Also what can I do to counter something else, is a matter of mathermatics mostly. Will i counter that pelter creep by t1 bombers, or will i use a nuke? Both will work, but which one is the better choice? Maths will tell you

2. Experience: Anyone will confirm, that without experience you will fail. Some will take 10 games to understand the game better, some will take 100 games, some will take 1000. eventually you will get a little better, but some people (to which i refer as "noobs"), are not new at the game, but played dozens of games, and still play like it was their first game. They are learning resistant, maybe they dont care about improving. They will never profit from experience, cos they dont have any.

3. Making the right decisions at the right time: This is also a mixture from the 2 above. PA even has this more than SupCom. A good example was a 2v2v2 i played lately. My teammate and me totally lost on the first planet, but we had enough eco to start a fast expansion to the second planet. And afterwards we saw, "hey it was the right decision" we managed to pull out an eco far superior to the other players on a 2nd planet. Why did we knew it was the time to go off planet? Experience told it and also maths, because we knew, if we go now to the 2nd planet, and we build up an economy, we might outproduce them!

4. Taking risks: Like the case i described in 3, we had to take a risk. Its not always clear if a certain strategic move will work. Sometimes you just have to risk it. And you will find out if it works. Then you come back to point 2 and now from experience if it works or not.

There are five things that define a strategic position:

1. The overall stated mission or goal
2. The ground where the battles take place
3. Climate that describes trends and patterns that change from moment to moment (each ground has its own climate)
4. Leadership character or your level of individual abilities at decision-making
5. Systems, skills and methods (also known as tactics)

There are some issues with the mathematical explanations of strategy in games.

First of all, it's very selective, I constantly see chess, tic-tac-toe and rock-paper-scissors as the main examples used, while a million other games that are equally solvable are overlooked as if the same mathematical principles don't apply to them. Planetary Annihilation for one, a game that is also theoretically solvable. (although of course this is completely irrelevant in real life)

Secondly, the assumptions that you have to make to produce the mathematical arguments don't hold in real life and therefore have not proven themselves to be relevant to the discussion. In real life you don't have infinite computing power, ability to execute or ability to generate perfectly random values. You are also not concerned with playing perfectly, but you are concerned with beating a human opponent. Human psychology is more important than mathematics, and understanding the statistical nature of applying strategies to real-life situations and finding one that best fits your desired parameters is an important aspect in itself.

As an example, a perfect game of chess could look like this:
1. Nf3 Nf6
2. Ng1 Ng8
3. Nf3 Nf6
4. Ng1 1/2 - 1/2

That's because in all likelihood in any of those situations the game is theoretically drawn if played perfectly, therefore any of these moves, ridiculous as they might seem, are quite adequate. However, if you want to actually beat your opponent it becomes a question of how to create situations that give complex problems for him to solve, so that the right move will not be so easy. An oracle that tells you the perfect move might actually be useless in many real-life situations in case you want to actually win a game of chess. And it doesn't exist, since unless you want to build a computer the size of Jupiter you will lack the computing power.

Rock-paper-scissors is another example. It's a silly game, but nevertheless: if you have access to a random number generator you will be able to play even, but if you want to win you will have to take advantage of weak play by your opponent. If your opponent always plays rock, rock, rock, .. you might start to think it could be advantageous for you to start playing only paper. However, your opponent could anticipate this and switch to scissors in time. You could also remain steadfast and stick with your random strategy, yet in real-life you will not have a perfect random number generator available. Again, a mathematical solution to this game taken from one's intro CS textbook is useless because it doesn't help you if you want to actually win a game and it makes assumptions about magical oracles that tell you perfectly random numbers that don't exist in real life.

When designing a game, it's still useful in some situations to take a skeptical look at the math behind the game and ponder about its applicability, but we need other criteria than purely asking questions about the existence of perfect strategy.

A weakness of chess, for instance, might be that at the limits of human capability the strategies used start to converge. There is a lack of personal style because the draw of the universal style of trying to play the correct move for the position is too great. Although in reality this is only a generalization and in every chess series you see stylistic differences between players. In the recent chess world championship match it was Anand that tried to seize the advantage in the opening, while it was Carlsen that tried to kill the dynamics and find himself in the quiet endgame positions where he is most comfortable.

And on another note, how would you even go about finding the perfect response in a given chess position? It would certainly require time, yet the game is on a clock and you only have a limited quantity of it. How do you allocate your time? What do you look first for in a position? These are all aspects of strategy.

Another thing,

For Starcraft, the high apm requirements actually add to the strategy because they introduce time as a factor. You only have limited attention span yet you still need to make constant decisions about how to allocate this. It's far from mindless and very much involves strategy. Micro adds to strategy: if you can only control one army at a time, which one do you choose? Difficulty in execution adds to strategy as for every decision you will have to take another parameter into account: how well can I and can my opponent control these units and how will that impact the battle? It adds unpredictability also, that rewards on the fly thinking. Many times in Starcraft history have strategies become viable only for the select few that are sufficiently skilled in unit control. I think that's very exciting.

Sometimes the people on this forum feel to me like people that listen to a Rachmaninov piano concerto and complain that all the notes get in the way of the music and that a virtuoso performance is just a result of rote memorization. I very much respect virtuoso performances, even if they don't guarantee that the music is of high quality, and I also think that virtuoso technique opens up new avenues of composition. If you don't like this sort of music that's fine by me, but don't pretend like it has no value and isn't worthy of respect, because then I start to suspect bitterness due to jealousy.

I'm more than happy for PA to be a slower paced game with less mechanical requirements, but you have to keep in mind that if you remove the mechanical requirements that a lot of the strategy in the game will disappear into thin air and if you don't build in valid replacements the game won't be very fun to play.

(It's also not that easy to just remove mechanical requirements, one runs into a lot of practical issues)

Starcraft can be similar to poker, where players are trying less to read the battlefield and more each other, whilst making bets based on that info to get more of an advantage.

While I agree that choosing between what to control does add strategy, I also think that having no theoretical limit to the speed at which you can do so subtracts from it by putting emphasis on physical performance.

My issue with Starcraft style apm requirements is that anybody below the highest level of play will always be able to get a significant advantage through apm without adding to their strategic insights.

It's one of the reasons I consider Airmech to be one of the more strategic RTS's, despite being action oriented. Moving a unit requires the same amount of time no matter who does it, and while you can shave a bit of that time off with fast reflexes, it's nowhere near the multiples of Starcraft controls.

Airmech is a prime example of a game that removed mechanical requirements for the betterment of strategy.

SPAM BOTS!
Lol

You know, I was fairly good at Starcraft 2, and my APM was 80 at average (highest around 180). Doesn't sound that much to me. People look at progamers in Starcraft and think that to be good you'd have to click spam like them. But did you ever look at what's happening on their screen? Most of it is spamming mouseclicks on nothing. It's to keep their attention up.

What that means is that you don't have to have high APM to be good at Starcraft.
Strategy is still the major factor in that game.

Having high APM obviously helps with the execution of your plans, but after a while you learn to decrease the number of clicks you have to spend on each task, and get used to switching between macro and micro frequently.

Starcraft 2 is a game that requires very high attention. That's why the games are designed to be so short.

I actually think the problem with the communities aversion to APM is actually much simpler than that.

APM comes from understanding, and feeling comfortable with what you are doing. Just like every profession in the world, the more comfortable you are with a subject, the quicker you become at doing it. APM will never disappear, and ther could very well still be someone with 300 APM on average playing this game. But you know what? They will still moan that the game is too much about APM.

The less significant your actions are, the less of them you need to make. If an action facilitates a 1 minute action, then one click is required during that minute. If an action facilitates a 3 second movement you need to make another in 3 seconds before you lose efficiency.

Starcraft has a higher APM requirement because is nothing but short term actions. Every little thing you do needs to be reevaluated very, very quickly. Cooldowns, movement, and building all happens very quickly and you need to keep tabs on all of it.

PA has a higher APM because you need to manage a bunch of different things. The sheer number of units and resources you are working with forces you to make lots of descisions and perform a number of actions on a regular basis to stay efficient.

If you make your actions impact more units, or last longer, you reduce the APM requirement.

That's just not true, sorry

Economy and upgrades are long term actions, as is every structure that you build, because you have to plan ahead.

I'm not talking about the implications, I'm talking about how long it takes to perform the action.

Even the most expensive units in starcraft take a few minutes to build, and resources are always consumed instantly.

How is PA different, other than in having a streamlined economy?

Units in Starcraft take way longer to build than in PA. It's just that every single unit matters relatively more than in PA.