RTS innovations

[brianeyci]

Shep has a point that has been overlooked I think. That games having some of the features people want already exist. And that really kills SWPIGWANG's argument that adding certain features would not be beneficial at all.

A suggestion if people want the thread to turn away from SWPIGWANG bashing: to someone who hasn't followed RTS since high school, I was hoping for some discussion of the newer stuff that already existed in this thread (innovation?) but instead there's... Starcraft. I mean come on. What would make this thread really productive for people who haven't been following RTS would be some kind of summary of what's happened in the past say, oh ten years, and what exactly RTS has to offer now. Since apparently there's people like Vehrec who don't know what's happened, and even people like me who want to know what's out there and what's not (that's what I was hoping for in this thread, discussion of some kind of innovative games already out there.)

[SWPIGWANG]

I'll start a new thread then......just to start clean

I was feeling defensive because there appears to be an direct attack on my gaming group and myself at various points in the thread, making it hard to keep rational.

[brianeyci]

The first page just has people talking about what they want. Even if they are stupid. Only when you started defending them like a sacred cow did people start chiming in with the "direct attacks." The closest is Stark and besides being Stark it's not even a direct attack, I read his post many different ways. And so what if it was an attack, you could've quoted exactly what you thought was an attack and flamed that.

It just looks like one big persecution complex, sorry. If someone wants a combat initiative slider, arguing against a combat initiative slider at all is stupid because guess what... it's a game, if I'm entertained looking at a shit machine shit out turds, who are you to say how I spend my free time. That's another point that Mike made that's been missed that should've really ended it all.

[SWPIGWANG]

Even if they are stupid. Only when you started defending them like a sacred cow did people start chiming in with the "direct attacks."

Well, it was Dendrobius that started the entire war at 2nd page, I just got drawn into it whatever. I realized that is a mistake now, but rest of the thread does not care.

In anycase, from the first post, most of what I've done is defend the current set of game design and never really launched serious attack on usefulness of things like AI. However I did consistantly attack the opinion that AI will be this genre revolutionalizing feature that completely change how games are played for the better just by tossing it in.

Then I have people telling me that a good gameplay feature is one that the player can ignore and a good RTS is one where the player can take as long as they want to issue commands. (and implying that games that does not work to this principle are stupid and are played by stupid players) That I just can't leave alone.

so it escalates

it's a game, if I'm entertained looking at a shit machine shit out turds, who are you to say how I spend my free time. That's another point that Mike made that's been missed that should've really ended it all.

Isn't that like....post-modernist or something?

One could argue that entertainment is void of generalizable principles, but that would make this thread pointless.

Shep has a point that has been overlooked I think. That games having some of the features people want already exist. And that really kills SWPIGWANG's argument that adding certain features would not be beneficial at all.

It is not that those features don't work. It is that those features don't automatically remove things like micro at the highest level of play or vastly expand the complexity of the game in terms of viable actions in most cases because creative players always look for more things to micro to improve their advantage, no matter how small, and any increase in complexity makes game balance an terrible problem that companies can not solve in reasonable amounts of dev time.

[HSRTG]

Oh, you're back. Joy.

In anycase, from the first post, most of what I've done is defend the current set of game design and never really launched serious attack on usefulness of things like AI. However I did consistantly attack the opinion that AI will be this genre revolutionalizing feature that completely change how games are played for the better just by tossing it in.

What we want is an already good game, say Dawn of War, with the AI additions. No one ever contested that a game needs a good engine and balance.

Then I have people telling me that a good gameplay feature is one that the player can ignore

I would like to point out that resourcing is pretty much ignored in Dawn of War. Is resourcing then a bad feature in it? Hmm? Perhaps you should come up with something that actually resembles a bad feature that we've suggested, instead of making broad and baseless accusations.

and a good RTS is one where the player can take as long as they want to issue commands.

Oh ho ho, here we go. Let me check the thread, hmmm, I'm not seeing any comments to the effect of "I want as long as I feel like to issue commands to my units". Still resorting to blatant straw men, hmm?

(and implying that games that does not work to this principle are stupid and are played by stupid players) That I just can't leave alone.

I'm not disputing that at one time it was necessary to almost-ignore unit AI. In 1997 we did not have the CPU power to do anything with them. But today, there is no excuse. At this point it's retarded to have nonexistent unit AI. Pointless. It's a useless relic of the last cycle of RTS games.

It is not that those features don't work. It is that those features don't automatically remove things like micro at the highest level of play

None of us are at the highest level of play. Nor did we suggest removing the possibility of micro. We just want some bones thrown to us, some casual gamers.

or vastly expand the complexity of the game in terms of viable actions in most cases because creative players always look for more things to micro to improve their advantage

Those creative players will be able to look for things other than more micro if these AI suggestions were implemented. As it stands, there's nothing to do but micro.

and any increase in complexity makes game balance an terrible problem that companies can not solve in reasonable amounts of dev time.

*Looks at Supreme Commander, at Medieval Total War, at Madden 2007*

Damn, that's an increase in game complexity. Particularly for creating strategic AI in the first two cases. Looks like you're full of shit, no surprise there.

[RogueIce]

In anycase, from the first post, most of what I've done is defend the current set of game design...*snip*

Now see, I think that's your first mistake. You come into a thread titled "RTS innovations" to defend the current set of game design. Do you not see how that might be a problem?

And I am sure that the obsessive player types will always find some way to micro out a victory, even in a more AI-type game. Just like there are the players who always search out the bugs/glitches/exploits of a game. But I don't see that as a reason to not try to make it more accessible to the casual player, either.

Obviously one of the "expert professional" players will probably walk all over the casual gamer who boots it up for fun, regardless of how it's designed. But is that any reason to try and find a way to reduce the clickfests that seem to dominate the genre?

[Stark]

Wait, he just strawmanned everyone's position into

AI will be this genre revolutionalizing feature that completely change how games are played for the better just by tossing it in

Nobody EVER said that, and indeed much of the discussion before the RTS Thought Police showed up was on this very topic. And

good RTS is one where the player can take as long as they want to issue commands

That must be a PIGWANG perversion of 'clicking fast does not equal smarter'. Any surprise he's been tilting at windmills? Apparently 'innovation' has a PIGWANG meaning just like 'strategy'.

HSRTG, you've got more patience than me. PIGGY showed up in a thread about *discussing* *innovation* and tried to sit on it and make the bad words stop.... then complains about being flamed. Maybe if he stopped repeating himself ('any increase in complexity makes game balance an terrible problem that companies can not solve in reasonable amounts of dev time' lol unsupported claim) and stating his conclusions as truth that wouldn't happen... but then he'd be having a *discussion*, which appears to terrify him. Oh and Dendrobius and PIGWANG not only prove everything I've ever said about RTS players, but they're also staggeringly arrogant. Do they *honestly*, *honestly* think RTS designs are made for 'the highest levels of play' - what, 10,000 people? - and not the tens of millions of casual-to-regular gamers? NAH.

EDIT - wow, I didn't notice his attack of insecurity on the last page. He even hid his problem (that clicking faster /= strategy) behind 'we are more intelligent because we don't like clickfests (even though we ALL PLAY THEM, no everyone in this thread hates the precious RTS)'... but still said 'arrogance concerning strategy'! Clicking faster /= strategy. Did that hurt your feelings little baby? Get this: it fucking isn't. Man up, dumbass.

[Darth Wong]

I've dropped the claim that a game that is different will necessarily be bad. I did demonstrate that it had not been successful (player base-wise) up to this point.

You have, however, made the absurd claim that it is impossible to determine what is and isn't real strategy, even though these are military games, and hence their strategy would ideally be based at least vaguely on military strategy, not ant-colony strategy.

The reason I post in this thread is two fold.
1. Arrogance concerning what is strategy. (those Realtime click-festers, I am more intelligent because I hate those games)

It's "arrogant" to think that real-time strategy games should resemble actual military strategy rather than a fucking ant colony? You're an idiot.

2. What I believe to be an mistaken belief that simply adding features would result in desired gameplay dynamics.

And that's a strawman; no one ever said that if you add a certain feature, you will magically get a great game regardless of the skill of implementation. And you are only creating this strawman because you refuse to admit that your beloved RTS genre has become bloated on its own complacency. Utterly idiotic unit behaviour is simply accepted by die-hard fans of the genre, just as painfully idiotic character behaviour is simply accepted by Star Trek fans. In both cases, after a while the fan is so accustomed to the idiotic behaviour that he actually expects it, and thinks it's natural. So instead of having a unit which is smart enough to do something when someone just outside his range starts shooting at him, we are regaled by your tales of how it requires "skill" to pay attention to all of these idiot units and manually micro-manage them.

But really this thread has been too poisoned now it is pretty much off topic in the flame-war territory. It could be just me reading this wrong, but no one is correcting me on that.

Everyone has been correcting you, fucktard. You're just ignoring points.

[SirNitram]

1. Arrogance concerning what is strategy. (those Realtime click-festers, I am more intelligent because I hate those games)

Strawman me again. I'm sure knocking around fake arguments makes your life alot easier than addressing the criticisms brought up, over and over again.

[Edward Yee]

I won't contest your ideas re: a fighting retreat, Fire Fly, but perhaps implementation of a turret system in combination with the facing and retreat options (the latter being as simple as using the "RoE" toggle) might make such work? (Although clearly with the advantage to units with turrets so as to bring their best weapons to bear on their attackers... perhaps a unit with multiple weapons could have differentiation between which of them is turret-mounted?)

[SWPIGWANG]

Those creative players will be able to look for things other than more micro if these AI suggestions were implemented. As it stands, there's nothing to do but micro.

A player always has the choice of not micro-ing something, only that it gives them a disadvantage on a tactical level.

*Looks at Supreme Commander, at Medieval Total War, at Madden 2007*

Damn, that's an increase in game complexity. Particularly for creating strategic AI in the first two cases. Looks like you're full of shit, no surprise there.

I'm not talking about AI complexity, I'm talking about strat complexity. The more viable strats you want to build into the game, the harder it is to balance. It would be full of irony if a game has all the ai build into it, and everyone just masses <insert unit> for the win.

As it stands now, Supreme Commander have largely symmetrical units, which limits most of their balancing concerns to experimentals. Medieval Total War is not balanced for ladder-style MP. (but the player base don't want ladder anyways)

Even games like Chess or Go has inbalances stemming from who starts first. In Go, the later starter actually gets "free land" when counting at the end game, and that bouse have changed over the hundreds of years of the game's existence. (making it a game that took hundreds of years to balance with no proof that it is balanced even now)

Do they *honestly*, *honestly* think RTS designs are made for 'the highest levels of play' - what, 10,000 people? - and not the tens of millions of casual-to-regular gamers? NAH.

What is this group of casual gamers? The casual gamers have been buying up those clickfests far more than the ones with innovative AI for the last decade or so. For the power gamers, anything that can improve their game is another thing to learn.

even though these are military games, and hence their strategy would ideally be based at least vaguely on military strategy, not ant-colony strategy.

It's "arrogant" to think that real-time strategy games should resemble actual military strategy rather than a fucking ant colony? You're an idiot.

No, most of what we've discussed are RTS games, not military simulations. Military is just the background, it is not the mechanics and was never designed to BE the mechanics. It is like D&D have a feudal background but is never intended to resemble feudal age combat. Trying for realistic sword damage in fantasy games is just a non-starter. (to people that actually play those games as opposed to loath it)

The reason with lack of military strategy is because the entire interaction is simply laughable with respect to the real world on every level. Armies that gets raised in 5 minutes, gold on the ground, 30 second buildings and tanks that take 30 shells to kill while guns have range of 10 meters. Unrealistic mechanics produce unrealistic strategy no matter how everything else is build around it. Guns with range of 10 meters produce tactics surrounding formation and crowd control. Fast build times produce non-unit-conserving or even produce-as-required strategies. Terriblely condensed chokepoints produce traffic jams. (and traffic jams happen in real life too if give the kind of traffic)

To those with the power gamer mentality (like myself), those are not infantry and tanks, but 50 hit point units with 10damage per second at 6 tile range. They simply do not make sense in any other way and we wouldn't have played RTS if we couldn't accept whatever weirdness that comes from it. In fact, those weirdness and counterintutivitiveness is a source of amusement subject to exploration.

Most RTS games are really "real-time 4x game", with incorporation of elements of "eXplore, eXpand, eXploit and eXterminate." Games of only combat is considered by some as real time tactics games. Now those realtime 4x games can never be realistic as they have to compress the time scale for production to nothing. When that first step is done, the rest usually follows. Those games makes as much sense as an ant-colony game (ant-colony have resource collection and production) as a military one. (which does not have those in remotely close to the fine frame we are talking about in RTS)

----

Of course, that is not the issue here since we are talking about future games. Would it be possible to build a game that uses military tactics? Absolutely, but the requirement is realism on the level of player control, not smart unit AI.

Lets look at the original Shogun total war. It was far more successful in capturing the dynamics of combat than a complex AI, weird unit interaction game like dark reign.

The greatest innovation of shogun is not smart units. At the level of technology back than, smart units relative to what we know today are not possible. It isn't even the unit facing, formation, terrain or unit type-counters. Those are in other games.

No the innovation is by forcing the commander to command entire blocks of hundreds of man, and limiting the player to only two handful of "units." The critical part about this is that it is not that each man is more intelligent, it is that the commander can't control them.

Without this limitation, you can bet on that alot of high level play would revolve commanding small chunks of troops everywhere, micro-ing like mad to gain minor tactical advantages. With the addition of perfect AI to remove the player from such insane micro, the battle tactics would be even more distroted as the age old command and control limitations of historical battles would cease. You'd see things like formations that "passing through each other" in battle as perfect pathfinding AI manages the crowds with absolute precision or something nuts like that. (if it doesn't, some nutty micro-manager is bound to try by hand)

I should note that historical evidence suggest that battle of agincourt was decided by crowd mechanics. (in this case, the rear ranks pushing the front to death)
----

I also want to turn to another type of gaming that has realistic military strategy but none of the underlying details: Good old Hex-tiled wargaming. Those games are played by history fans which have high expectations for realism. Those games have even less processing power behind it, limited to only D20 dice and what can be computed in an average person's head with help of a piece of paper.

But it worked fine. The reason is that all the small things are abstracted, an on the level those people are playing one, the input-output response of the game is correct. The chances of a SS Panzer assualt of breaking a Soviet Union conscript division is 75% at 40% readiness blah blah.

----

So what am I saying here? The critical part of emulating military strategy is to limit the player to what it should be capable of doing, and abstract the rest to produce the correct output from the player input. Whether you make it by monitering the stats of a million troops or a simple computation on a sheet of paper, both would work.

A real time-hex game would produce the required set of strategy.

Sure you could produce those numbers by processing a million varibles and show it on the screen, but that is just eye candy. Thats fine, but it is still just eye candy.

Utterly idiotic unit behaviour is simply accepted by die-hard fans of the genre

I accept idiotic unit behaviour because I know that units can never be as intelligent as the commander. It simply does not have the information required for intelligence and the game is about the commander giving units the proper behaviour.

For many units, there simply isn't a universal "good" behaviour that works for all or even most sistutations, especially when the enemy is trying to exploit weakness in behaviour. To determine what is good behaviour, the unit need a wealth of information of all the other friendly units and a great deal of what the enemy is doing in the fog of war. It has to know a traffic jam on the other side of the map or a battle behind a hill to be "intelligent."

However the unit is not suppose to know what is unseen and unknown. (espeically that of the enemy) Therefore, units are stupid, differing only by degree.

I can accept that units that sit there while a battle just outside their "response radius" is raging out of control. I know for the trade of that "stupidity", it prevents the unit from being drawn away into range of enemy artillery in another case. It is up to me as the commander to decide whether this is a battle to fight, or a trap to draw my troops out of position. Now it would be nice if I could command those troops quickly once I've made my decision, but it is my decision to make.

-----
I'm reminded of an discussion about Combat Mission (a serious wargame) where units are actually MORE(!) "intelligent" than their counterparts in real life and the great cheesiness that results from it. The discussion is about the "recon by sucide trick" that power gamers use in the game. How this works is that one sends a low cost vehicle against the enemy front line to force them to open up and expose their position, while using cheap binocular equipped infantry to spot. Right afterwards they allocate precision artillery on top of the enemy line on the next turn.

That is unrealistic because the sucide vehicle can not really transmit useful enemy position info back when it is too busy getting blown up, and that cheap infantry without radio can not inform artillery observers which is hidden far away about the precise cooridnates about the enemy in a short time.

However, in the game, thanks to the power of the omniscient player, the very real communication limits have been bypassed and the player can prepare artillery in under a minute (turn length of combat mission).

The counter is almost as unrealistic. What it involves is a complex fire plan where every weapon covers a small arc thus exposing only a small set of weapons against those recon raids. While this is doable in defense sistuations, in an advancing and dynanmic battle, it would overload the communication networks of the time. Thus, units once again is more intelligent than it could possiblely be in real life thanks to help of omniscient player.
-----
So what does this say? Units are "stupid." People on the battlefield are stupid too, often not knowing what is happening on another flank or behind the hill. France lost the war in WW2 partly due to stupid artillery troops retreating against manstein's insecure beach-head across the rhine due to a rumor. The US lost the first battle of Salvo Island because half the defending force was unaware that the other half is underattack. (a very typical RTS event...it did result in court marital however)

-----
The purpose of the commander is to collect information from the troops and give commands. Without the commander, units are indeed "stupid."

If there is a fault to RTS stupid units, it is that it replaces the dozens or so sub-commanders in a battle of that scale and replace it with a single commander. Instead of feeding a dozen commanders information over the period of hours, RTS players are feeding hundreds of units information over seconds. The entire communication network have been replace by one person.

To me, it is only a matter of how many as opposed to how stupid. Without the chain of command, real life units can't even produce effective friend or foe identification. (and friendly fire is never intelligent unit behaviour)

[HSRTG]

A player always has the choice of not micro-ing something, only that it gives them a disadvantage on a tactical level.

With better unit AI it wouldn't be fucking suicide not to micro, however. As it stands, you have to be a moron not to.

I'm not talking about AI complexity, I'm talking about strat complexity. The more viable strats you want to build into the game, the harder it is to balance. It would be full of irony if a game has all the ai build into it, and everyone just masses <insert unit> for the win.

Where, oh where, have I ever stated that pure unit AI would be the saving grace of games? I have always been saying that it requires a backbone of both a good engine, and good balancing for the factions.

As it stands now, Supreme Commander have largely symmetrical units, which limits most of their balancing concerns to experimentals. Medieval Total War is not balanced for ladder-style MP. (but the player base don't want ladder anyways)

No no no no, you stated that any increase in game complexity would make it impossible to develop in reasonable amounts of time. Don't move the goalposts fucker. I also note that you haven't commented on Madden '07.

Even games like Chess or Go has inbalances stemming from who starts first. *Snip Go balance*

RTS =/= to TBS. Hasn't ever been. I agree, it's hard to balance factions, but adding sane unit AI shouldn't affect balance at all. It's still Infantry > Tank > Arty > Infantry.

What is this group of casual gamers? The casual gamers have been buying up those clickfests far more than the ones with innovative AI for the last decade or so. For the power gamers, anything that can improve their game is another thing to learn.

You're looking at 'em. Folks like Darth Wong, me, Stark, basically everyone who doesn't play games for $$$ is a casual gamer. Kind've like that softball team for the office drones, the one that meets on the local baseball diamond?

The reason with lack of military strategy is because the entire interaction is simply laughable with respect to the real world on every level. Armies that gets raised in 5 minutes, gold on the ground, 30 second buildings and tanks that take 30 shells to kill while guns have range of 10 meters.

Protip: Making points against your opponents is a good thing. This argument you're making about "Hay guyz its not realistic lolz" has nothing to do with unit AI.

Fast build times produce non-unit-conserving or even produce-as-required strategies. Terriblely condensed chokepoints produce traffic jams. (and traffic jams happen in real life too if give the kind of traffic)

First you bitch about how these aren't realistic at all, now you're saying that bad pathfinding AI resulting in traffic jams is a good thing 'cause it's realistic. Make up your mind.

To those with the power gamer mentality (like myself), those are not infantry and tanks, but 50 hit point units with 10damage per second at 6 tile range. They simply do not make sense in any other way and we wouldn't have played RTS if we couldn't accept whatever weirdness that comes from it. In fact, those weirdness and counterintutivitiveness is a source of amusement subject to exploration.

That's what they represent to you. To us they represent reasonably intelligent people, and we expect the units to act like it.

Most RTS games are really "real-time 4x game", with incorporation of elements of "eXplore, eXpand, eXploit and eXterminate." Games of only combat is considered by some as real time tactics games. Now those realtime 4x games can never be realistic as they have to compress the time scale for production to nothing. When that first step is done, the rest usually follows.

Translation: "Lets not improve anything, the status quo is sensible, with nonexistent unit autonomy"

Those games makes as much sense as an ant-colony game (ant-colony have resource collection and production) as a military one. (which does not have those in remotely close to the fine frame we are talking about in RTS)

But we don't need to have ant-like units. Can't you get this through your thick skull?

Of course, that is not the issue here since we are talking about future games. Would it be possible to build a game that uses military tactics? Absolutely, but the requirement is realism on the level of player control, not smart unit AI.

So you say. Reality disagrees. If I tell some tanks and machine gunners to go take a bridge, I expect them to utilize their weapons in a sensible manner, and to use cover sensibly as well. These are supposed to represent reasonably intelligent people. Why not have them act like it?

Lets look at the original Shogun total war. It was far more successful in capturing the dynamics of combat than a complex AI, weird unit interaction game like dark reign.

It also had no micro at the person level. We didn't tell swordsman #157 to step to the left into cover.

Without this limitation, you can bet on that alot of high level play would revolve commanding small chunks of troops everywhere, micro-ing like mad to gain minor tactical advantages. With the addition of perfect AI to remove the player from such insane micro, the battle tactics would be even more distroted as the age old command and control limitations of historical battles would cease.

Straw man. Just because the AI is reasonably intelligent, I don't expect it to run my battle totally. I tell an AI subcommander that he has two phalanx, one cavalry, one catapult, he needs to hold the left flank. See, real life generals have an ability called "Delegation". It allows them to focus on the overall battle.

You'd see things like formations that "passing through each other" in battle as perfect pathfinding AI manages the crowds with absolute precision or something nuts like that. (if it doesn't, some nutty micro-manager is bound to try by hand)

Straw man. I never advocated perfect pathfinding for each unit in the Total War series. I advocated better unit AI, and better pathfinding for things like Supreme Commander.

I should note that historical evidence suggest that battle of agincourt was decided by crowd mechanics. (in this case, the rear ranks pushing the front to death)

How nice. Every RTS battle ever played then mirrors Agincourt perfectly then.

I also want to turn to another type of gaming that has realistic military strategy but none of the underlying details: Good old Hex-tiled wargaming. Those games are played by history fans which have high expectations for realism. Those games have even less processing power behind it, limited to only D20 dice and what can be computed in an average person's head with help of a piece of paper.

Ah, going back to strawmanning about how we wanted RTS to be TBS. Keep knocking it over, I'm sure someone will think you're making any points.

So what am I saying here? The critical part of emulating military strategy is to limit the player to what it should be capable of doing, and abstract the rest to produce the correct output from the player input. Whether you make it by monitering the stats of a million troops or a simple computation on a sheet of paper, both would work.

*Sigh* Back to stating that only player interaction defines a feature, only in more words? You see, I don't want to have to fight the interface of a game. I'm fairly sure that games today get bad reviews if they do force you fight the interface.

I accept idiotic unit behaviour because I know that units can never be as intelligent as the commander. It simply does not have the information required for intelligence and the game is about the commander giving units the proper behaviour.

No, but they can be smarter where it comes to retarded tactical choices. Fw: Machine gun vs Tank+Infantry

For many units, there simply isn't a universal "good" behaviour that works for all or even most sistutations, especially when the enemy is trying to exploit weakness in behaviour.

In Starcraft alone 99% of micro is preordained to having units fire on something that they can hurt. Automating that would be a universal good.

To determine what is good behaviour, the unit need a wealth of information of all the other friendly units and a great deal of what the enemy is doing in the fog of war. It has to know a traffic jam on the other side of the map or a battle behind a hill to be "intelligent."

It doesn't have to know anything about what's behind the FoW, just to fire on shit that it can hurt, perhaps pursue when it's told to, and to find other ways around the goddamn traffic jam.

However the unit is not suppose to know what is unseen and unknown. (espeically that of the enemy) Therefore, units are stupid, differing only by degree.

In pitched battle, having to tell my fucking Goliaths to shoot the fucking Mutalisks should be redundant. They should know to fire on shit they can hurt.

I can accept that units that sit there while a battle just outside their "response radius" is raging out of control. I know for the trade of that "stupidity", it prevents the unit from being drawn away into range of enemy artillery in another case. It is up to me as the commander to decide whether this is a battle to fight, or a trap to draw my troops out of position.

Or if you had a Combat Initiative Slider, the AI could probe the enemy to see about a trap, and then pull back if it starts being hit by to much. But that's heresy. Blasphemy!.

Now it would be nice if I could command those troops quickly once I've made my decision, but it is my decision to make.

See, it shouldn't have to be your decision, you should have the option of deciding about each and every raid, or setting your units to a particular mode of defense.

*snip cheese tactic*

Okay. Whatever. What does this say about unit AI?

So what does this say? Units are "stupid." People on the battlefield are stupid too, often not knowing what is happening on another flank or behind the hill. France lost the war in WW2 partly due to stupid artillery troops retreating against manstein's insecure beach-head across the rhine due to a rumor. The US lost the first battle of Salvo Island because half the defending force was unaware that the other half is underattack. (a very typical RTS event...it did result in court marital however)

They can be dumb, but they shouldn't be retarded. FW: Machine guns/tanks/infantry.

If there is a fault to RTS stupid units, it is that it replaces the dozens or so sub-commanders in a battle of that scale and replace it with a single commander. Instead of feeding a dozen commanders information over the period of hours, RTS players are feeding hundreds of units information over seconds. The entire communication network have been replace by one person.

So let's add some decent AI to it to reduce the person's workload. No wait, heresy etc. Sorry. See above note about delegation. It's up there somewhere.



Independent Idea: Giving sub commanders some preset missions about defending a flank with some given units, or watching for an attack, etc. The AI would take some designated units and use them to fulfill the objective you've given it.

[Darth Wong]

Holy fuck, am I ever sick of that idiot PIGWANG ignoring the fundamental point (that micromanagement is no fun) and pretending that adding extra command features to an RTS would somehow result in less command flexibility. It would result in more command flexibility, but it would be different from what he's used to, and he's so goddamned stupid and close-minded that he can't see that.

I never thought I would actually see someone defend the stupidity of RTS units as a "feature". What a fucking tool. "Oh no, you can't make units too smart or even create an option for a combat initiative slider because then the gameplay would be different from what I'm used to!"

And I'm sick of his wanking about "dedicated" gamers too. Guess what, PIGWANG: normal, well-adjusted human beings play games to have fun, not to win at any cost. And part of having fun is immersion into the game, which requires that your tanks and infantry actually act like fucking tanks and infantry, not goddamned soldier ants.

[Vympel]

I didn't even fully realize there was such a thing as the "hard core RTS player" until reading this thread- the moment I saw "APM" (actions per minute) as some sort of measure of how formidable a player is, the existence of an entire subset of gaming culture that treats playing a game as some sort of professional sporting event, complete with obscure statistics ... Jeez. Sad.

[SirNitram]

What's most disturbing about this 'Dedicated gamer' and 'APM' and other such bullshit is that it's disconnected from those who are high on the ranks. Yes, the high ranks are all clickfests with pre-ordained build orders and absurd micro(Voice of experience here), but I'd never heard 'APM' or how 'Dedicated' the Eldar guy who whooped me because, at higher tiers, Eldar sodomize Orkz hardcore.

As far as I can tell, this mentality is the sole realm of either the top-one-hundred players(Unlikely), or of deranged little cultists like PIGWANGSUCKER.

[Xisiqomelir]

I didn't even fully realize there was such a thing as the "hard core RTS player" until reading this thread- the moment I saw "APM" (actions per minute) as some sort of measure of how formidable a player is, the existence of an entire subset of gaming culture that treats playing a game as some sort of professional sporting event, complete with obscure statistics ... Jeez. Sad.

Pro sports generally turn fans into statisticians, no?

[Darth Wong]

I didn't even fully realize there was such a thing as the "hard core RTS player" until reading this thread- the moment I saw "APM" (actions per minute) as some sort of measure of how formidable a player is, the existence of an entire subset of gaming culture that treats playing a game as some sort of professional sporting event, complete with obscure statistics ... Jeez. Sad.

Pro sports generally turn fans into statisticians, no?

The obsessive ones, yes. But at least sports generally tends to involve athleticism, fitness, and face-to-face interaction with other human beings, hence it is somewhat more defensible than computer game playing. That doesn't excuse rabid sports fans; they are just as bad as loser geeks in any other genre.

[Vyraeth]

or vastly expand the complexity of the game in terms of viable actions in most cases because creative players always look for more things to micro to improve their advantage

You seem to think that the point of Real Time Strategy games is to encourage players to come up with the best build orders and develop the best micromanaging skills. That's simply not the case.

Yes, I know all about the so called "strategy" for build orders, and what not, I used to play Age of Kings (AoE II) quite a bit, and the higher level players had the best build order and could do it with the most hotkeys, etc... but I don't think that's the spirit of the game.

The point of the Real Time Strategy genre is to let things unfold in real time. To see actual tanks duking it out, instead of waiting for them to trade shots in a turn based setting. To see units get produced, to forge groups, to have to adjust strategies in mid battle etc.

The micromanaging and build order bullshit is nothing more then a side effect of simple/crude implementations of this idea.

And in case you think I'm just giving you my opinion, let me give you the opinions of people who actually develop RTS games for a living and thus are shaping what it means to play an RTS.

Take a look at some developer Q&A's with Big Huge Games about their game Rise of Nations (quite old). One of their biggest features, on top of all the Civ-like things was citizen automation. That is, making it so that your citizens act intelligent and complete tasks/buildings as needed.

So you know, wouldn't have to micromanage them.

And while I haven't played any modern RTS's yet (I personally don't like the genre because of the excessive micromanaging), I would be utterly shocked if this trend hasn't continued.

What most every single person (that I've seen) is advocating in this thread, is the most common sense leap forward.

You don't build a game so that it's a clickfest, you build a game so it's enjoyable. I have a feeling that your resistance to the "evolution" (so to speak) of RTS games is a result of you spending a huge amount of time becoming an expert at clickfest style of games and wanting to preserve your advantage.

Get over it, smart unit AI would be a dream, and it would definitely make the RTS genre far more appealing.

[Oni Koneko Damien]

I'd like to point back about...oh...six or so pages.

I want to make it perfectly clear that on page two or three, I pointed out that there are idiots who had trained themselves to play uber-well around the limitations of AI in older RTS games.

I also want to point out that I stated that it was these idiots who were so damn opposed to giving better AI, more flexibility, etc, because it made the skills they had spent so long honing in that restricted environment suddenly useless.

Now, I want to point to the five or so pages of spinwang or whatever, proving that point I made to be completely true.

Vindication, it tastes soooo damn sweet.

[MKSheppard]

I tried playing new Stock CoH 1.5 and it was boring as shit. I don't want to spend 15 minutes just building up so I can build tanks, I just want to start sending Shermans to the other side...which is why I like my ShepMod for COH a lot more..

[ray245]

So your mod has tanks at the start of the game? Can you past around the mod, I love those kind.

Can't we just delete base building in RTS and focus on producing units itself? So that each side can produce at around the same rate, unless you want to go for better but more time-consuming tanks instead of standard ones.

So that we depends more on tactics instead of pure numbers, how we are able to make full use of our force we currently have instead of throwing them all away because we can re-built them without hurting things like morale?

The need to ensure getting all of your attacking force wiped out isn't a good idea...

[Hotfoot]

So your mod has tanks at the start of the game? Can you past around the mod, I love those kind.

Can't we just delete base building in RTS and focus on producing units itself? So that each side can produce at around the same rate, unless you want to go for better but more time-consuming tanks instead of standard ones.

So that we depends more on tactics instead of pure numbers, how we are able to make full use of our force we currently have instead of throwing them all away because we can re-built them without hurting things like morale?

The need to ensure getting all of your attacking force wiped out isn't a good idea...

That was pretty much Ground Control 2 for you. You captured points, some of which could be used as drop zones for bringing in more units. The only real "base building" was upgrading your dropship to carry more troops or to be more effective in combat when you brought it on the field, and a few turret units you could load into troop transports.

I'd be content if the only real base building involved erecting defensive structures. This is one of the reasons CoH works so well in my opinion. The only base building is minimal: There's never a reason to build two bunkers or really two of any "true" base building, since the cost of one could easily buy a unit, and you almost never need the speed an extra production building would give you since you don't have the resources to work with it.

I do admit I like the method escalating a conflict from light skirmishes to heavy tank battles, as it represents (to me, at least), the nature by which conflicts can occur. Two patrols run into each other, call for reinforcements, have to hold their ground until support can arrive, and then they lay into each other. However, Shep does have a good point: sometimes you just want to dive right in and start whaling on each other with tanks right from the get-go. It would be nice if RTS games easily allowed for both without the use of player-made mods.

[SWPIGWANG]

The point of the Real Time Strategy genre is to let things unfold in real time. To see actual tanks duking it out, instead of waiting for them to trade shots in a turn based setting. To see units get produced, to forge groups, to have to adjust strategies in mid battle etc.

And I'm sick of his wanking about "dedicated" gamers too. Guess what, PIGWANG: normal, well-adjusted human beings play games to have fun, not to win at any cost. And part of having fun is immersion into the game, which requires that your tanks and infantry actually act like fucking tanks and infantry, not goddamned soldier ants.

I suppose that is what the other people in the thread wants out of their RTS. They want to see things happen as opposed to just (abstract) strategy and abstract victories.

For me, the fun in the game is learning how to win it. Once that is done it becomes pointless, or at least pointless after running though the campiagn once.

Beside being a RTS player, I am an Custom Maps/Mods player and I am familiar with massive abstractions with reguards to units and interations. When you are used to games where an individual infantry unit is suppose to be an stand in for an entire infantry division on the east front, things like immersion is like lol. Immersion to me is like eye candy. (and eye candy exists for immersion)

(that micromanagement is no fun) and pretending that adding extra command features to an RTS would somehow result in less command flexibility. It would result in more command flexibility, but it would be different from what he's used to, and he's so goddamned stupid and close-minded that he can't see that.

I'm not arguing that it would result in less flexibility if implemented correctly, but it would not automatically result in more flexibility either.

Flexibity here meaning as "the number of useful choices the player have to make in a fixed period of time" as competitive players do not care for features that do not effect their win rates. I realize you people don't care about those things however and I won't push it.

So let's add some decent AI to it to reduce the person's workload. No wait, heresy etc. Sorry. See above note about delegation. It's up there somewhere.

Independent Idea: Giving sub commanders some preset missions about defending a flank with some given units, or watching for an attack, etc. The AI would take some designated units and use them to fulfill the objective you've given it.

Well, to the gamer that "power games", only the results of unit interactions matter. Whether they are the results of stupidity or intelligence matters little. I don't care whether 5 tanks fighting 10 infantry man is fought in an whatever manner, I care whether I'll have 3 tanks left in the end.

Things like delegation itself is really an "micromanagement" issue not unlike transforming tanks or casting spells especially if it is critical to success (eg. suicidal not to do) and time urgent. In the original Homeworld, the most critical skill for the player is the Formation-Unit Behaviour AI juggle. It is rather silly that the best player in the game is those that can be time the "sphere-claw formation, aggressive-defensive behaviour AI, Z-afterburner and K-kamikaze" "micromanagement." The bad part is that such juggling was more issue of rhythm than any long strategic thought.

Perhaps it is a bit more interesting micro than fighting unit targeting and path-finding. As for reduction of workload? Workload is defined by the pace of the game, not by the stupidity of the units. If you played any RTS at 10x the speed, you need to be 10x faster to play it.
---------------------------------

With better unit AI it wouldn't be fucking suicide not to micro, however. As it stands, you have to be a moron not to......
In pitched battle, having to tell my fucking Goliaths to shoot the fucking Mutalisks should be redundant. They should know to fire on shit they can hurt.

The problem isn't so much that units are stupid than that humans are smarter.

Lets say your Goliath piroritize air combat units. Than your opponent can use hallunicate to cast a few hallunicated Scouts/Corsairs (or even Carriers, if the AI cant tell an empty/fake carrier from a full one) and attack with a wave of zealots, owning your Goliath. Hallunicated Zealots are ineffective as it gets in the way of real zealots and have less health. Doing this attack requires the opponent to micro (cast spells, cooridnate attacks) and thus once again give micro manager once again an edge. A human player controlling the Goliath would not target Corsairs and Scouts over Zealots, as both have terrible anti-ground damage and can be left to kill after the high threat zealots are killed. (however they would have to kill those first if the Goliath is covering for a full mana science vessal which can then cast irridate on the zealots, killing them)

Or lets say the opponent is zerg and knows goliath piroritize mutalisks. So he runs a few mutalisk around the edge range of your Goliath to make them fire on that while moving a stack of overlords to drop ultralisks in your base. Once again micro have fooled the "now more intelligent unit." Of course, a human player can determine whether the overlord is loaded or not from complex analysis of the situation. For example, if your Goliath is close to the Zerg player's base, you can expect the enemy units to walk to fight you instead of being carried by an overlord, where the overlord is only for cloak detection. On the otherhand, if you are on high ground with a protected chokepoint, overlord you see is probably is loaded. Similiarly in battles against mutalisks and Zerglings, both units are almost equally vulnerable to the Goliath. What to target depends on the other parts of the force composition perhaps at a remote area. (eg. Whether you've build AA missile turrets at base, or a firebot-filled bunker at the choke)

For the Goliath to know what target to hit against a micro-ing, intelligent opponent, it needs to know what the enemy has, what forces you have, what the terrain is like, what is the tactical strategy and so on. That is absolutely non-trivial AI worthy of a player to control. (it would be nice to have a hotkey to hit certain kinds of target however) Sometimes such micro can even confound players as one can easily see how inexperienced players can make mistakes at the above sistuation.

Incidentally, Starcraft does have simple unit targeting piroritation. Military units are targeted over non-combatants and buildings automatically, and units that have fired on your forces targeted over units that have not fired. This targeting rule have been exploited ruthlessly by micro-managers, like the "medic wall" micro against melee units. In this micro, a medic forms a tightwall in front of ranged combat units, while enemy melee units try to get past the medics to get the fighting units. In the absence of micro on both sides, this is very good behaviour as marines are more vulnerable than medics and cutting them down stops return fire. To counter the medic wall, the unit would need to compute the pathfinding possibility to the firing units in conjunction with all the other attacking units (usually there is a small hole to the rear, but it can't fit a dozen attacking units) and the relative number of enemy firing units. (if there is only one or two marines and a small hole to the rear, than killing the marine is more effective as it stops all return fire)

Even for the machinegunners vs Tank issue, there might be a time where machinegunners should target the tank over the troops beside it. (assuming the machinegunner can damage the tank at all) For example, if you know you have an anti-infantry minefield around your base making you safe to infantry assaults, and the enemy builds a light tank to assault you. If you kill the tank, the assault halts even if you lose all infantry, while if you shoot the supporting infantry the enemy can just zoom his tank into your base and win the game.

In anycase, almost all games what have implmented hard armor classes (starcraft does not) have piroritized firing. That have not removed micro from those games in general, but it has changed it into a different form of micro. Most commonly: armor class overloading, like having a tank oriented army with micro tanks to specifically target missile troops as opposed to any troops, or an infantry army focused at elimated machinegunners even for the few tanks in the army. Aside from targeting, there is formation based armor class-overloading by having an entire flank loaded with one armor class to gain advantage against mixed units....etc etc

Short of beating the turing test, trying to beat every trick of micro that a player can come up with in a "micro-permissive" game is a losing proposition.

If one wants to emulate modern combat: There is a commander for every 9 man and every fifty of them will have a full time commander with full set of intel that a RTS player have. The difference between miniscure unit tactical control differences seperates the performance of the Iraqi army from the US, and the ww2 French Army from the Germans. Trying to emulate an dozen intelligent human commanders and countless intelligent humans with one computer to the point that the commander on the higher level can find no flaws (to take personal command and gain advantage) is just....hard..... When even real people train to fight can't "micro-command enough" on the modern battlefield, one wonders whether it makes sense to try to match it with a computer.

Probably better to design games where tiny differences in unit behaviour make little difference as oppose to all the difference, if you want to stop micro. Games that are won on the basis of "because he had more units grrr" is a testament to that.

First you bitch about how these aren't realistic at all, now you're saying that bad pathfinding AI resulting in traffic jams is a good thing 'cause it's realistic. Make up your mind.

RTS games, especially of the 4x kind, is universally unrealistic.

Given the mechanics of those games, however, things like traffic jams is bound to happen no matter the kind of intelligence you put in. We are talking about unit density where it covers 50% of the land surface area with muti-direction direction traffic. There is no pathfinding algorithm, human or otherwise, that can solve this sort of problem universally for all sistuations. There is alot of mixed requirements, sometimes the players want speed (where a stream of units is optimal), sometimes the players want formation, sometimes the players want agility. Sometimes the player wants to piroritize a stream of traffic over another, while other times movement can be halted for combat.....etc. The interface can help, but player input is always necessary for smooth flow.

This is a lot less of a issue in real modern combat as weapons have ranges of muitple kilometers and units are seperates by dozens of meters inbetween them while being tiny. Instead covering 50% of the surface area of the battlefield, it covers under a 1% and there is alot of space of everyone to go everywhere.

If you want to fix traffic jams, give the units the space it needs, not try to make a better ai to navigate a dynamic maze.

No no no no, you stated that any increase in game complexity would make it impossible to develop in reasonable amounts of time. Don't move the goalposts fucker. I also note that you haven't commented on Madden '07.

Madden '07 does not need to have "teams balanced enough for a ladder" I suspect, however I can't comment on it without having played it.

It is not that complexity makes it impossible to develop in time. Complexity makes it impossible to balance in time. Even really stupid and simple game designs of a dozen units can fall victim easily to bad balance, I shudder at the thought how long it'd take to balance a strategy game that captures the real complexity of modern warfare as some would want. Note that reality is not balanced.

[SAMAS]

With better unit AI it wouldn't be fucking suicide not to micro, however. As it stands, you have to be a moron not to......
In pitched battle, having to tell my fucking Goliaths to shoot the fucking Mutalisks should be redundant. They should know to fire on shit they can hurt.

The problem isn't so much that units are stupid than that humans are smarter.

Lets say your Goliath piroritize air combat units. Than your opponent can use hallunicate to cast a few hallunicated Scouts/Corsairs (or even Carriers, if the AI cant tell an empty/fake carrier from a full one) and attack with a wave of zealots, owning your Goliath. Hallunicated Zealots are ineffective as it gets in the way of real zealots and have less health. Doing this attack requires the opponent to micro (cast spells, cooridnate attacks) and thus once again give micro manager once again an edge. A human player controlling the Goliath would not target Corsairs and Scouts over Zealots, as both have terrible anti-ground damage and can be left to kill after the high threat zealots are killed. (however they would have to kill those first if the Goliath is covering for a full mana science vessel which can then cast irridate on the zealots, killing them)

Or lets say the opponent is zerg and knows goliath piroritize mutalisks. So he runs a few mutalisk around the edge range of your Goliath to make them fire on that while moving a stack of overlords to drop ultralisks in your base. Once again micro have fooled the "now more intelligent unit." Of course, a human player can determine whether the overlord is loaded or not from complex analysis of the situation. For example, if your Goliath is close to the Zerg player's base, you can expect the enemy units to walk to fight you instead of being carried by an overlord, where the overlord is only for cloak detection. On the otherhand, if you are on high ground with a protected chokepoint, overlord you see is probably is loaded. Similiarly in battles against mutalisks and Zerglings, both units are almost equally vulnerable to the Goliath. What to target depends on the other parts of the force composition perhaps at a remote area. (eg. Whether you've build AA missile turrets at base, or a firebot-filled bunker at the choke)

I'm failing to see where any of this is a Bad Thing.

You seem to be disproving some of your previous points, about intelligent unit AI reducing micromanagement and strategy.

This is the exact kind of strategy and dirty tricks we want.

Here's let's take your examples one step further by adding another RTS concept, and applying it to the matter at hand: Veterancy.

Let's say that a unit that gains veterancy not only becomes stronger/tougher or gains new weapons or abilities, it's AI also goes one step higher. So a Veteran Goliath will also target Transports over other air units, or will ignore Hallucinatied enemies if it's not taking damage.

In fact, for a game like Starcraft, this would be ideal, as it allows players to come up with such manevers, and have them actually be useful.

For the Goliath to know what target to hit against a micro-ing, intelligent opponent, it needs to know what the enemy has, what forces you have, what the terrain is like, what is the tactical strategy and so on. That is absolutely non-trivial AI worthy of a player to control. (it would be nice to have a hotkey to hit certain kinds of target however) Sometimes such micro can even confound players as one can easily see how inexperienced players can make mistakes at the above sistuation.

Actually, it's quite trivial. Remember, an RTS puts you in the seat of a commander giving out orders. You shouldn't have to tell Trooper #418 to not bother shooting the tank if there's a guy with a machinegun standing next to it, or for Mecha Pilot #37 to use its missiles on the enemy aircraft first, as the tanks have no defense against it. Training is supposed to take care of all that. You shouldn't have to do more for that kind of stuff than setting the aforementioned initiative slider, or clicking certain pre-set directives, like you find in Dark Reign, Total Annihilation, and Dawn of War.

Micromanagement is the equivalent of a commander personally leading his troops into the field. Naturally, it should give you better results then just telling your boys to hold this hill or take out that outpost and moving on. It shouldn't have to be a requirement for basic play, however.

Reducing necessary micro allows the player to concentrate more on overall strategy, and makes intense micro for decisive blows. For example, I send a large force of basic troops at one of your expansion bases, forcing the player to perhaps micro the defense of that base to survive until help arrives.

But the attack is a feint. I'm personally micro-ing a small group of Veteran/Elite units to make a strike at your main base. Be micromanaging the defenses at your expansion, you may be able to fight off my larger attack, while my micro-ing of my little strike force will allow me to inflict some serious damage to your base.

So now you have a choice. You can leave your expansion to its fate and defend your main base, send a large number of troops to defend the expansion and take me head-on, as it were, by micro-ing the remaining defensive units, Micro the expansion defense and leave the main bulk of your force to hold me off, or whatever other dirty trick you can come up with.

Even for the machinegunners vs Tank issue, there might be a time where machinegunners should target the tank over the troops beside it. (assuming the machinegunner can damage the tank at all) For example, if you know you have an anti-infantry minefield around your base making you safe to infantry assaults, and the enemy builds a light tank to assault you. If you kill the tank, the assault halts even if you lose all infantry, while if you shoot the supporting infantry the enemy can just zoom his tank into your base and win the game.

That only works if you have only machinegunners, and a lot of them. In which case, a tank rush will tear you apart in the first place(on the next assault. My feint proved useful, didn't it? )

[/quote]

[Vendetta]

Lets say your Goliath piroritize air combat units. Than your opponent can use hallunicate to cast a few hallunicated Scouts/Corsairs (or even Carriers, if the AI cant tell an empty/fake carrier from a full one) and attack with a wave of zealots, owning your Goliath.

Unless, of course, only some of the Goliaths or analogues in a given battlegroup are set to prioritise air targets, at which point your entire argument falls to smoking bits.

Because, of course, this would be an option that the human player has set, in accordance with the strategy he has devised, and the AI would then follow that strategy without micromanagement, giving the impression that it is in fact an intelligent troop under player control.

After all, you say immersion doesn't matter to you, but as this thread demonstrates, you're in a minority of one on that issue, and can therefore be dismissed as irrelevant.