Tactics vs. Strategy
Author
In the design process for an RTS, I've stumbled on a significant shortcoming of most 'strategy' games - most fall under the 'tactics' genre, because the player is spending more time micromanaging their troops than actually concentrating on grand strategy. After all, a modern-day four-star general usually would not spend his time on the front lines, directing his troops in person, but would operate from his war room. This, and not the role of the infantry captain in the foxhole with his sergeants and corporals, is the true meaning of a 'strategy' game.
That brings me to my main question: is it truly possible to develop a tactically and economically competent AI that would manage both the economy and the hands-on fighting of a normal battle while minimizing the amount of micromanaging involved? Tactics games are quite enjoyable, and they have their place, but strategy is a wholly different genre, and one that requires top-notch AI programming to work at its best.
It is pretty pointless, IMHO, to develop complicated low-level game play while building an even-more-complicated AI to abstract said game play away from the player. Why not focus your effort on making a compelling strategy game, and abstract away the tactical details?
Games like Kohan and Dawn of War show some good ways of giving you large tactical battles, while not forcing the player to micro-manage every individual soldier. Instead of complex AI, they simply group individual soldiers into squads or companies. This allows the player to spend less time controlling troops and more time thinking about the overall war or tweaking the economy.
- Mike
Games like Kohan and Dawn of War show some good ways of giving you large tactical battles, while not forcing the player to micro-manage every individual soldier. Instead of complex AI, they simply group individual soldiers into squads or companies. This allows the player to spend less time controlling troops and more time thinking about the overall war or tweaking the economy.
- Mike
And if you do, how can you do it in such a way that your game isn't criticized for being a sim city game (economics heavy) or a 'move the icon game' (abstracting the grand scale too much) or one that's not easily accessible (too wargamey) or one that 'plays itself' (AI does a little too much)?
*shrug*
*shrug*
I think it'd be pretty hard to run an RTS from a war room?
Tactical vs. Strategic is much more important in actual war games than it is in commercial gaming. If you want a game, make it. If you want to add strategic elements, add various bonuses depending on victorty conditions or resource allocations (ie: send 5 tank units to one front instead of the other). If you aren't designing for a defense wargame contract, you should just work on what you envision to be fun and challenging (profitable? I'm deferring to the forum title) rather than definitions. If you want a "strategic" RTS, in terms of a larger focus than the current battle, brainstorm as to how to effectively achieve that feeling and responsibility. Military definitions won't help you here.
Tactical vs. Strategic is much more important in actual war games than it is in commercial gaming. If you want a game, make it. If you want to add strategic elements, add various bonuses depending on victorty conditions or resource allocations (ie: send 5 tank units to one front instead of the other). If you aren't designing for a defense wargame contract, you should just work on what you envision to be fun and challenging (profitable? I'm deferring to the forum title) rather than definitions. If you want a "strategic" RTS, in terms of a larger focus than the current battle, brainstorm as to how to effectively achieve that feeling and responsibility. Military definitions won't help you here.
I believe Rise of Nations is a good strategy game.
-It can be played on massive maps, on which it takes 15 minutes to move infrantry from one side of the map to the other.
-Wars fought at sea, on land or in the air differ greatly. Whilst Sea battles are mostly paper-scissors-rock, land battles require tactical skill with around 12 (depends on the Age you are in) different combat unit (excluding Unique Units and Air and Naval units) like the General, Supply Wagon and Horse Archers. Air battles basicly are Bombers & Fighters, where Fighters kill Bombers, Bombers kill buildings, Anti-Aircraft guns kill both and ground troops are killed by both.
-Getting a powerful economy will take 40 minutes even if you don't fight, but a game "starts" after 5 minutes, when one player will have a reasonable army.
-No super units means that it is unlikely to have asudden end due to a Nuke (Nukes can be used, but the greatest damage they can do is destroy a city (which is usually rebuilt within 5 minutes).
-Games can take very long a short depending on how you play it (I've played a game of 9+ hours and one of 10 minutes)
Just my view ;)
-It can be played on massive maps, on which it takes 15 minutes to move infrantry from one side of the map to the other.
-Wars fought at sea, on land or in the air differ greatly. Whilst Sea battles are mostly paper-scissors-rock, land battles require tactical skill with around 12 (depends on the Age you are in) different combat unit (excluding Unique Units and Air and Naval units) like the General, Supply Wagon and Horse Archers. Air battles basicly are Bombers & Fighters, where Fighters kill Bombers, Bombers kill buildings, Anti-Aircraft guns kill both and ground troops are killed by both.
-Getting a powerful economy will take 40 minutes even if you don't fight, but a game "starts" after 5 minutes, when one player will have a reasonable army.
-No super units means that it is unlikely to have asudden end due to a Nuke (Nukes can be used, but the greatest damage they can do is destroy a city (which is usually rebuilt within 5 minutes).
-Games can take very long a short depending on how you play it (I've played a game of 9+ hours and one of 10 minutes)
Just my view ;)
Quote:Original post by doctorsixstring
It is pretty pointless, IMHO, to develop complicated low-level game play while building an even-more-complicated AI to abstract said game play away from the player. Why not focus your effort on making a compelling strategy game, and abstract away the tactical details?
To build on this, maybe what the AI should be doing isn't micro-managing the tactics and economy. Maybe, the AI should be making it look like it's micro-managing the tactics and economy. Think of it more like a complex visualization to your abstracted game play. The rules figure out which side of a confrontation wins and then the AI shows "how" that side won.
The rules see: 7 swordsmen advancing on 3 archers
The rules decide: 3 swordsmen and 0 archers survive.
The AI shows: The archers shoot the swordsmen once, but only one is taken down. The archers shoot one more time before the swordsmen have reached them and take down two more. The two archers are killed quickly, while the third dodges and manages to stab one swordsman before dying.
Generals don't build bases, or do much else for that matter. In an actual war, decisions are made all over the place by different people, and it's not fun to play as just one person.
By making RTS games fast-paced and tactical, you keep the average player from getting bored to tears.
By making RTS games fast-paced and tactical, you keep the average player from getting bored to tears.
Author
Let me clarify. I'm not talking about eliminating micromanagement, but making it less necessary for a player to perform. If they truly want to play the RTS in a manner that would bore the strategy fan with ADHD to tears, they would have the choice; if they wanted tactics, they would have the liberty to take control from the AI and do things themselves.
A second clarification: the entire idea is defined by one word - choice. If they really want to, for example, they can modify the tax rate of each settlement by the tenth; if they don't, the AI can take over, without increasing the tax rate to 500% to compensate for that arsenal of nukes the player wants to build. Another point is that some players simply cannot keep up with managing an economy single-handedly while applying professional and effective military tactics on the front lines. The AI can handle actual tactics (flanking, knowing when to retreat, calling for supplies, etc.), if the player chooses not to. There's nothing concrete about this.
A second clarification: the entire idea is defined by one word - choice. If they really want to, for example, they can modify the tax rate of each settlement by the tenth; if they don't, the AI can take over, without increasing the tax rate to 500% to compensate for that arsenal of nukes the player wants to build. Another point is that some players simply cannot keep up with managing an economy single-handedly while applying professional and effective military tactics on the front lines. The AI can handle actual tactics (flanking, knowing when to retreat, calling for supplies, etc.), if the player chooses not to. There's nothing concrete about this.
If you want to discourage micro-management, then make it less effective. Make it so that tactical decisions have a far more significant impact than just what soldier fires at what other soldier, or where your penny's go precisely.
Players, especially hard-core players, play to win. That's normal. And to win, you'll have to be efficient. If micro-managing your units gives you a better chance at winning, then you'll micro-manage your units. And in most games, that's how it works. So, technically you have a choice, but in reality, it becomes obvious that you have to micro-manage if you want a chance at winning.
So if you provide AI-stand ins, make them solid and trustworthy, but most of all, make them effective. If players see their AI do stupid things, they'll loose trust in them and most will then rather do those jobs themselves. Unless time is more important than efficiency, in which case time is a more valuable resource than economic or combat efficiency. Then, maybe even rough AI agents can be usefull. As long as it actually helps the player, it will be used.
Bytheway, I'd call micromanaging units 'operational' rather than 'tactical'. Telling a soldier at who to fire deals with how he operates. Telling several squads of soldiers how to time their movement, when to ambush, from where to attack, is what I call tactics. Strategy plays on a higher level: which battles to win and which to loose in order to win the war. Or at least, that's how I interpret these terms. I could be wrong on how and what they exactly stand for. :)
Players, especially hard-core players, play to win. That's normal. And to win, you'll have to be efficient. If micro-managing your units gives you a better chance at winning, then you'll micro-manage your units. And in most games, that's how it works. So, technically you have a choice, but in reality, it becomes obvious that you have to micro-manage if you want a chance at winning.
So if you provide AI-stand ins, make them solid and trustworthy, but most of all, make them effective. If players see their AI do stupid things, they'll loose trust in them and most will then rather do those jobs themselves. Unless time is more important than efficiency, in which case time is a more valuable resource than economic or combat efficiency. Then, maybe even rough AI agents can be usefull. As long as it actually helps the player, it will be used.
Bytheway, I'd call micromanaging units 'operational' rather than 'tactical'. Telling a soldier at who to fire deals with how he operates. Telling several squads of soldiers how to time their movement, when to ambush, from where to attack, is what I call tactics. Strategy plays on a higher level: which battles to win and which to loose in order to win the war. Or at least, that's how I interpret these terms. I could be wrong on how and what they exactly stand for. :)
the first game I remember with a usefull AI standin was Earth 2140. In it you had a limited number of "generals" which could be assigned control of squads and set to either Aggressive, Scout, or Defensive commands.
The game in no way required you to use the AI generals at all, and at any time you could manually control the units instead, with all the traditional groups. But it sure was handy to be able to grab a few units, assign them to commander 1 and tell that commander to scout.
All the AI generals we're was a limits set of slots that could run simply dependeble choices.
Personally I think the most important thing about AI standins is not that they be competitive against direct human control, espcially not for a whole range of situations ... but instead that they do something well enough that they can releive the human from having to think about it if the human is willing to settle for small efficiency / control penalty. Also improtant is that they not have any obsurbdly broken flaws that they constantly fall into (say the scout AI always goes clockwise, or the defensive AI doesn't aid friendlies under attack 1/2 screen away, etc.
Some turn based games have AI stand-ins. For isntance civilization has governors (which I don't use - but that's just me), Stars! had computer controllers that could play for players who miss their turns, and the AI would not do any aggression, just keep scaling the economy while the player was away, and keep the existing defensive millitary in place.
The game in no way required you to use the AI generals at all, and at any time you could manually control the units instead, with all the traditional groups. But it sure was handy to be able to grab a few units, assign them to commander 1 and tell that commander to scout.
All the AI generals we're was a limits set of slots that could run simply dependeble choices.
Personally I think the most important thing about AI standins is not that they be competitive against direct human control, espcially not for a whole range of situations ... but instead that they do something well enough that they can releive the human from having to think about it if the human is willing to settle for small efficiency / control penalty. Also improtant is that they not have any obsurbdly broken flaws that they constantly fall into (say the scout AI always goes clockwise, or the defensive AI doesn't aid friendlies under attack 1/2 screen away, etc.
Some turn based games have AI stand-ins. For isntance civilization has governors (which I don't use - but that's just me), Stars! had computer controllers that could play for players who miss their turns, and the AI would not do any aggression, just keep scaling the economy while the player was away, and keep the existing defensive millitary in place.
This topic is closed to new replies.
Advertisement
Popular Topics
Advertisement
