What is it to be good at a strategy game ?

It mostly makes sense, but your desire for concreteness seems pretty abstract

To clarify: are you wanting more chess-like strategy where being able to see and understand all potential scenarios X moves ahead is both desirable and reasonably possible?

Yes.

I admit I do not really know what it would mean for a game to focus more on calculation situations, but for sure it requires to have good information on your opponent and context, so it cannot be always the case (contrary to chess, where complete info is always available). So I guess my post is not very helpful, sorry.

But maybe what I said is more relevant to the way the AI should work - for instance, I have never understood why the AI in Heroes of might and magic was not able to play almost perfectly (luck notwithstanding) in tactical battles, given that all the info is available. Obviously it works with general rules instead of thinking moves ahead (attack the shooter first, even if it means you will be under the range of attack of a very powerful melee creature two rounds after). The more the game depends on calculation, the better the AI will be.

Perhaps it is better to think about this in a different way.  Instead of thinking how things are in a game, think of how things are in real life.  As a commander leading men into combat, you have access to two types of information.  Quantifyable, and not-quantifyable.  I can count how many tanks and infantrymen I am going to be able to put into the field with a reasonable degree of accuracy, and my scouts, satellites and aerial observation can also usually get an idea of what my enemy can put into the field.  Either way it can be counted and quantified, and how much you know is dependent upon how complete your information is.

The other side of things is information that cannot be easily quantified.  What is the morale of my troops and how does that compare to the enemy?  How well will my small unit leaders perform and how much will that influence the battle?  How will my and my opponents doctrine perform against one another?  Can I get inside the opposing commander's decision cycle, and how do I do that?  These are the questions that change war from a simply matter of arithmatic into an art form, and the best strategy games allow for a little bit of both sides.  If I want to play chess I will go out and play chess, and I think that is true for most gamers out there.  If I want to play chess with incomplete information, then there is a game called Stratego.  I want a game that isn't just a set series of calculations with greater or lesser degrees of information.  Strategy is more than just calculating the easily defined quantities of men and equipment and how they are employed.  It is about understanding yourselves and your enemy, and using that understanding to win.

One game that you might be looking for is Etherlords II. It is just the type of game when you calculate your and oponnents moves trying to emerge victorious; in some duels i even popped up a calculator to be sure about the result of this attack, lol The game also offers situations when you apply "general rules" the only difference is that those rules are made by you.

Your dividing of strategy thinking types is imho right, I believe a good strategy game has to offer something for both types.

My hopes for Elemental are "M2TW in fantasy setting". M2TW could be a really great LAN game with my friends, but it had 2 critical problems. First, it had no AI in MP. Second, it had no global mode in MP. Random skirmish battles lack the deepness given by the global mode (you fight here because you want that castle. And you want that castle because of his position for example...) A shame...

Then I discovered Elemental. Fantasy setting with magic ! Hip hip hooray. Global mode and tactic battles both in MP ! Hip hip hooray ! Now lets hope for AI and its clear. Im sold here.

The thing is, Elemental offers a lot more than just this...I especially like the idea of having a powerful essence-filled character on the beginning. My hopes here are that it will look cool and powerful. I always get a saddened laugh when "power" is promised to you and it is depicted by a simple fireball. If I have power, I want to feel it ! For example Sacrifice...I never forget the feeling when I conjured my first Volcano But look at Demigod for example. You are supossed to be a demigod, you look cool yeah...but your spells look like the ones ordinary mage in NWN or Warcraft III does. Eh.

I also like the idea of the ability to create user content. If it will include creating own spells, then its a hot candidate to finally create something Oblivion tried. In Oblivion you can create your own spells with your own effects, but the problem is...it looks identical to other spells. Fact is, every spell looks just the same. A simple fireball 1pt damage VS advanced fireball 100000pt damage on 25ft ? No change. You just fling out a tiny fireball. Thats ridiculous.

That is one reason why I lean on the side not random anything.   Good calculation can create something you can become "good" at.

on the other hand, random lets casual players still play in the same game without being totally creamed.

But life isn't a system where you can go in, a nation state can learn the system and then beat everyone else indefinately.  It isn't about luck, but about the difference between crass calculation and the art of war.  A player can be very good at a system that takes these things into account (we call these players Gudarian, Manstein, Zhukov, Patton, to name a few WWII examples), and an organization can be very good at implementing systems to maximize its advantages and take advantage of its opponents weaknesses (like the US Army today), however things never truly work out the way they would seem to be leaning.  In the American Civil War during the battle of Antietam the relative size and training of the two forces should have resulted in a massive Confederate defeat (as contrary to popular belief your average Union soldier was generally worth about as much as your average Southerner, and McClellan did have Lee's marching orders), however Lee fought McClellan to a draw.  On numerous occasions throughout history victory or defeat has hinged on a seemingly illogical rout or failure to follow orders or indecision on the part of the commanding general... that is war, and those who take advantage of that aspect are the ones who go down in history.  Games that play out like a game of chess always strike me as artificial and contrived.  If you want predictability, just use autocalc with an odds calculator.  I want a game that gives the essence of warfare between humans.  It isn't necessarily random, but it can be unpredictable.

Stratagey games are about being able to adapt to situatitions that you have not forseen to me that is the entire point. I want to be able to pull some crazy stuff of in a scirmish or even rage at being defeated dispite my upper hand. In multiplayer i would like to be able to co-ordinate my attacks effectivly. I also want this to be similer to a fantasy total war. (though i would like some mythical creatures like elves and minataurs to be availiable in small amounts - as specialty battalions). I enjoyed total war skirimishes against my friends and had an even better time when someone was successfully able to launch a crazy counter attack.

Etherlords fits your needs. Because even when you precisely calculate the damage and launch an attack, the enemy can defend. And that must not always mean by creatures; there are many "surprise" spells that turn the tide. Launched attack does not mean you won. He casts "fog" and all attacks are not possible; your creatures simply return back. An unexepected resolution.

And that is something I would like to see in Elemental aswell, since it will be turn-based like Etherlords.

If this were going to be a pure wargame, I might end up with the minimal-randomness folks. But one of the most interesting things I've seen so far at this early stage is that Elemental is intended to be a serious RPG-TBS fusion. Depending on how that works out in the details, the game might well support players (single and multi) who have a very subjective, story-centered idea of what would make a given game "good."

If Stardock's past history is any indicator, I expect they're going to work hard to satisfy a *really* wide range of player types. Brad loves to code game AIs, so there will undoubtedly be a great many "hard" calculation aspects to the system. But he also apparently has some affection (past history?) for table-based RPGs. With luck, calculation lovers will have plenty to sink their teeth into, and folks who prefer playing by some combo of gaming instincts and staying in character will be able to be equally serious players "without being totally creamed."

These (1) and (2) techniques that you speak of, both of them are applicable to all strategy games, and to master any strategy game, one needs to master both of these techniques, and master combining them as well.

An example of (1) from chess would be to know that the queen is a powerful defender and that the pawns are good attackers.  Or that two bishops can work together to make a knight into a very strong attacker along a certain line.

An example of (2) from GalCiv2, would be that I know that founding an Influence Base at a certain point will force the AI into declaring war on me if he wants to win, therefore I should be ready to defend it before building it.

People just tend to focus on (1), because it always stays the same, allowing one to rely on ones memory, rather than thinking.  (You only have to think to figure it out, after that you practise it in and rely on memory.)

The power of (1) is only unbeatable when the game is unbalanced, meaning that (1) forces you to take all the right actions and the enemy cannot do anything about that.

In good games, like GalCiv2, StarCraft and chess, every action has a cost and therefore, if the enemy can detect or guess your action they can choose to counter it.

For example, say building only cavalry are the 'best' according to (1).  The other player can then guess that you will do so and build much pikemen to counter you.  Then cavalry in this instance are not the best choice, but you can anticipate this and plan accordingly.  To really beat your opponent you need to spy on him, but then again, he can build a bunch of cavalry where you expect them to be while his second army is elsewhere and follows a different plan.

The same rule determines whether an AI is good or not.  The AI needs to be able to calculate the cost of different actions (i.e. use (1)) and guess what the player is doing by examining his actions and counter it (i.e. use (2)).

For more information, check out Dave Sirlin's book (for free on his website), "Playing to Win".

StormSeed, I'd just like to say "good post" Not just because I agree with you, but you put forth good examples of how (1) and (2) are always there (though perhaps not always in a balance that make for a good game).

Thanks.

I see your point about the balance between the two, and it's giving me something to think about.