RTS innovations

[Darth Wong]

You know what, your endless wanking over the glorious benefits of excessive micromanagement is so goddamned tiresome that if you continue, I'm going to strip it out of this thread. Nobody here but you is enamoured of hardcore micromanagement. Nobody here but you thinks that the RTS genre is just perfect the way it is. Deal with that fact. If you think we're ALL wrong, fine. Go wank off with your "APM" buddies.

[SWPIGWANG]

Subleaders and radio operators etc could be incorporated for modern armies, but the idea of moving backwards from 'general psychically connected to every man and woman in the whole army getting instant 100% accurate information and delivering orders instantly' standard is at least interesting.

I agree with this train of thought. If you want to reduce micro and the likes, there is no better way in my opinion.

Micromanagement won't work if they can't do it Trying to beat the player at micro by replacing it with a computer is similar to trying to beat the player in intelligence at every level except the top. (tough)

Instead of resourcing, there are many different ways to get 'units' or 'money', and some have been seen in games already. One race could be forced to 'find' stuff in caches or 'activate' their deadly robots in some way

It is as if a thousand whiners have all cried out at once.......

This is really problematic if you are trying to balance it for mutiplayer. You need to keep near constant access for all the races or else the game breaks completely. For example, if on one map, the enemy can build an anti-air turret on high ground preventing race A getting air reinforcements, it would be totally unbalanced. Alternatively, if on map B, the natives can be cut off from race C by bunker push, than it totally imbalances the game too.

It can possiblely be done, but only with massive game development time and rigid map design rules.

I meant something more along the lines of "Starships travelling at c-fractional speeds blasting at each other from across the gulf of space" type of combat. The type of combat that would be more likely to see than the nBSG knife fights or other such things. With inertia and such.

Ships in homeworld stop moving when their engines turn off, move incredibly slowly without ever really accelerating, and have pretty dinky ranges. I certainly know how to visualize a simulation of space combat that'd fit into hard sci-fi environs, but I'm finding it hard to visualize the fun aspect. There's certainly no spaceship porn in a situation like that.

It can be fun for the harder sci-fi fans and those that have imagination. (niche market)

The real issue is how to implement an interface that can do this. Orbital trajectories and the likes are counter intutive and the learning curve would be immense. It will need an interface that automates the most difficult parts and still leave enough for "traditional strategy."

-------------------------------------------------------------
I'll add in my .2cent on what I'd like to see out of new RTS:

------------------------------
I want to see "balanced game management issues over time." In traditional 4x RTS, units can grow near exponentially, decreased only by combat where players can avoid. The usual solution to the disparity between early and late game unit count is by having "technology" that makes expensive late game units replacing more numerous ones. Nonetheless, most games change character over play time, usually starting unit-micro intensive and end with domination of economic control. However, no matter what is done, few games maintain the entire richness in units and combat over time.

What I'd like to see is a "flat development" game where players starts with enough units to start fighting rightaway. Combine that with a territory-control economic system (as in DoW/CoH) as opposed to investment based economic system (TA/SupremeCom) one can make the game more "flat", development wise. With flat development, one can focus on rich gameplay within that one time frame as oppose to worrying about balances that grow out of economic or tech development. There is no longer the need for super peon wait at the start of the uber unit spam at the end.
-----------------------
Since we are talking about AI alot, I'll talk about what kind of AI I'd like to see as opposed to "meh automate everything or else it fails because it ain't as intelligent as me or the micro-whore!"

I'd like to see attempted is unit control based on an completely different principle. Instead of just controlling unit behaviour on units directly, one tags "relationships" to other objects other than the unit in being.

For example, one can tag a mine field area with "enemy presence, avoid" tag and any unit on "transit movement" would avoid crossing it with its path finding AI. Alternatively, one can mark areas as "safe" for pathing AI to use maximum speed passage, and other areas as "contested" where the AI would reform formations of command groups and move in combat formation.

One can also tag units and behaviour to both area and unit types and the likes. For example, one can set up an "target air transport first" command for a piece of high ground and any anti-aircraft unit in the area would automatically behave as such. In another example, one can set up an "moving area" tagged around an aircraft carrier and have "defeat anti-naval units" behaviour for units in the region. Finally, one can tag the enemy's, say supreme commander, as the "target and encircle at all cost" and units around it would respond accordingly.

Another thing is that there should be programmable behaviour on the interface with reguarding informing the player! There should be a "popup sub-window and warning to track X if unit X is under attack or if Y unit is spotted in Z region" interface setting somewhere.

That said, I don't think this will reduce the amount of click-festy-ness of the game as there is often far too much stuff to "program" the "AI" for each sistuation and the player would be busy, especially when faster players can plan even more complicated attacks that takes even more effort to diffuse.

It is probably more interesting micro than pulling units out of the gunline at 10% health or blocking the path of the enemy melee units though.

[Nieztchean Uber-Amoeba]

Interestingly enough, Empire at War has a rather realistic way of conducting ground warfare: You have a limited number of ground forces in your fleet, built in the strategic phase at the Galactic level, and once you're planetside you start out with a piddling number of units, but if they can capture and hold strategic points (Dawn of War-style) you get to call down reinforcements of your choice from your limited pool. Discourages turtling, since the dfender can keep pumping out his own soldiers from factories and barracks, and makes every soldier valuable since if you throw away the guys you get at the start, you're fucked.

[SWPIGWANG]

You know what, your endless wanking over the glorious benefits of excessive micromanagement is so goddamned tiresome that if you continue, I'm going to strip it out of this thread. Nobody here but you is enamoured of hardcore micromanagement. Nobody here but you thinks that the RTS genre is just perfect the way it is. Deal with that fact.

I'm providing an explanation why that micromanagement have never been removed despite an decade of complaints and why the often proposed solution is inefficient that it could not and have not been implemented to achieve the required effect up to this point (the future, perhaps) and is probably misguided.

Micromanagement is an outgrowth of detailed and responsive unit control. It happens in TBS, or even RPG games as well when the system allows it.
---------------
If you want to discuss this topic, show me how one could possiblely program an AI that removes the need for the player to play fast in an nonlinear, discrete-step engagement. A rough sketch on how it respond to the bewildering set of sistuations would be nice.

(saying unit inititive is not enough, as units with inititive is not automatically more intelligent, only less predictable and perhaps even more sucidal)

[GuppyShark]

The reason why TBS or table top gaming involves more originality is because start conditions is heavily randomized, greatly limiting pre-game planning at the cost of letting the worst player win when lucky.

Care to explain this? There's nothing random about the starting conditions of a tabletop wargame. I pick my army, you pick your army, we deploy and engage.

[Stark]

If you are talking about standard RTS where air units have hundreds of hit points and take dozens of hits to take down, it make perfect sense.

Wrong. If you have 50 turrets and there are are ten inbound bombers with 5hp, you DON'T want to fire all 50 shots at one bomber, EVER. You want five shots per bomber maximum. This is my point: not focus vs spread (which is useless in hp games, as you say) but WASTED SHOTS. In many situations it's not 'some' overkill, it's MASSIVE overkill, and this is exploited by players all the time. It's stupid, and I repeat - if enough shots are already travelling towards a unit to kill it, no more shots should be fired at it. Every SAM in your base should NOT fire at one target.

However, if you want autonomous (as opposed to player set behaviour) AI to do this, you'll be meeting with a group of cursing gamers at the horribleness of the game AI when their SAM refuse to fire at those bombers that have just dropped bombs when there is no better target in range, or those fixed anti-aircraft guns that decided to spread fire at cheap intercepter as oppose to the nuclear bombing strategic bomber, especially when the bomber has been targeted once by the player already, but flew out of range of the turrets in its first run.

So, a fucking retarded and utterly broken implementation would be bad. No shit. Oh wait, are you using 'worst case scenario' to demonstrate why an idea has absolutely no merit AGAIN?

1) the SAMs are only going to 'not fire' if the units are already fired on and toast, not AT RANDOM or something. Why wouldn't they fire at bombers after release? Explain how the suggestion would result in idle defences with enemies in range.

2) How do you go from 'spend just enough fire to kill as many units as possible per volley instead of firing at closest target' to 'shoot interceptors instead of bombers'? In case you hadn't noticed, screening good units with crap units works in pretty much every RTS ever. At least with sensible weapon allocation, you'd get the maximum number of units killed per volley... unless you're microing to target bombers, in which case this is irrelevant. Indeed, once you've got such target-sensitive targetting, adding 'oh and kill bombers first' would be trivial. Explain how this is a weakness.

Guppyshark, he probably thinks it's 'unfair' if both sides don't have identical terrian. It's his whole 'unbalanced for multiplayer' thing - even though tournament level miniatures play STILL has either random or preset non-symmetrical terrain.

[SirNitram]

If you want to discuss this topic, show me how one could possiblely program an AI that removes the need for the player to play fast in an nonlinear, discrete-step engagement. A rough sketch on how it respond to the bewildering set of sistuations would be nice.

(saying unit inititive is not enough, as units with inititive is not automatically more intelligent, only less predictable and perhaps even more sucidal)

Dear Christ, you're pathetically stupid. You can do this with Red Alert 2 coding.

Reset snipers to only target 'Hero' armour, so they'll knock out heros, leaders, and the like. Reset rocket troops to only fire on vehicles or structures. Reset normal troops to only fire on infantry. Trivially easy; put a 0% or 1% in their effectiveness against other armour types.

Now just have a toggle where they'll either target the units with the leader max health, or most, depending on your strategy.

[Vehrec]

As for goliath targeting, it is critical for them to target the highest threat first (basic micro) and that depends on the tactical sistuation. Spreading out fire make things worst.

Not neccesarily. A Group of 12 gliaths could focus fire on wraiths, but then they would be overkilling each wraith by about 400% But if they split into four sub-groups automatically when targeting wraiths en-mass then they would take out the wraiths much quicker than they would with either simple focus fire and sustaining less damage than they would if they had simply bombarded the wraiths willy-nily with their missiles. does this take more processing power than a simple dumb 'stand there and only fire back if fired upon' AI? Oh there is no doubt about that! But saying that it's beyond us is just silly. You might have to make some graphical sacrifices to free up some time, but damn if it wouldn't be worth every polygon!

[SWPIGWANG]

Like I said, when air units have relatively low hp to firepower, than spread fire would be needed.

If you want to solve the problem of overkill, there is better ways than AI.

To do it with what would be worthy to be called "Artificial Intelligence", you need to:
1. Identify the enemy unit hitpoints and formation
2. Identify all other air defenses near by
3. Somehow split air defenses after using the data above

As you've aluded to, The better way is to make the weapons be mechanically "instaneous" thus do not waste any firepower as the target is tagged as dead before it is over killed. (as a aircraft is unlikely to be in range of too many units before the computer can compute a kill as oppose to the slow missile animation) Graphically one can still draw slow missiles however. That is a game mechanics fix that is far easier than try to coordinate and entire set of air defenses. This effect can be done by a editor to unit stats in almost any RTS without too much hassel as developing inter-unit communication intelligence.

I mean, for gods sake, even C&C1 has "hit scan" weaponary in the form of minigunners. There is no need for an more advanced engine for this sort of stuff. No need for any new "super innovation" but just tweeks to the engine that any bored modder can do.

If you are talking about a mechanics change, than I have no problem with that. If you are talking about developing inter-unit communication for something that can be solved by something simple than I would not agree. If you want to remove micro, do it at the mechanics level not at the "sub-player input level." Doing engine->AI->Player stack is double the work for the same effect.

Care to explain this? There's nothing random about the starting conditions of a tabletop wargame. I pick my army, you pick your army, we deploy and engage.

Guppyshark, he probably thinks it's 'unfair' if both sides don't have identical terrian. It's his whole 'unbalanced for multiplayer' thing - even though tournament level miniatures play STILL has either random or preset non-symmetrical terrain.

In a tournament miniature game, there is ALOT of randomness.
1. Armies are build before hand, thus one has to guess the opposing army composition as oppose to doing it on the fly.
2. Terrain is set up just before the battle. Missions are determined by Dice rolls.
3. The first turn (in games that have it), is often determined by dice rolls too.
4. There is alot of randomness in combat computation. Because there is few hit points on the table top, most weapons deal damage directly and good and bad rolls can have critical effects on battle.

Tournament level miniature play is an very different environment than RTS tournament play. Meta-gaming is the "strategy" of the day where ever competitor try to out guess the opposing army composition and build an counter army. Once into the tournament, luck is an extremely important factor in both what opponents one draws and what tactical sistuations araise.

In an extreme case, let take Warhammer 40K table top match up. In an Raider Dark Eldar vs Infantry-heavy weapon imperial Guard, winning the 6D roll for first turn means a massive advantage that may well seal the game with a few more lucky breaks, as both sides are extremely vulnerable to each other's attacks and can not consistantly survive one turn without being ineffective. We are talking about losing 40% of your army on turn one before you can make even one move! (to be fair, this problem is specific to a limited number of TT games)

Its telling that when table top players ask about who wins touraments, they ask what army is used as opposed to what "strategy." The army is the strategy (for example, your army only have 150 foot guardsman with long range heavy weapons, there is nothing to do other than shoot until the enemy dies) and dice the path to victory.

Sure, some intellect is needed since being stupid never wins games. However the best player in a screwed up match up still lose against average ones. The advantages the better player is probablistic as opposed to deterministic. Usually, over the round of 6 games, the better players would have enough good and bad luck to cancel out the effect, so terrible players don't win touraments that much. (however they still can, unlike RTS) However the there is simply a smeared out group of good players as opposed to a clear ladder of ability.

...blah blah

Now just have a toggle where they'll either target the units with the leader max health, or most, depending on your strategy.

....I have nothing against something like a toggle or even some unit AI (I've even proposed what kind of AI I'd like to see).

Basically, I have nothing against things where the player controls like toggles and the likes. All it really does is it averages out inputs overtime (you need to set the toggle) as opposed to have it happen all at once in the tiny time frame of battle.

I just don't see how the issue of "clicking faster = more power" can be solved by it easily. Maybe it is a strawman position that no one is supporting, but considering the hatred against micro shown in this thread, I doubt it. To control a toggle, one needs two clicks. To tell a unit to target a specific unit, one needs two clicks too. Someone with 300APM can simply have far more fine grained AI script control than someone at 50APM and thus once again gets a advantage, as both values are averaged not just peak.

To stop the clickfest you need to stop "everything" where fast commands can be used to gain an advantage. AoK for example have formation and behaviour toggles. While it stop frantic "select-and-move" control, players with high APM use their time in micro-ing horse archers, dodging seige engine attacks and manage the overwhelming economy.
----
That is why I think TBS is just a much better way if one wants to give the player time to think. With hybrid RTS-TBS engines, where the game is mechanistically real time but inputs happen after a set period of time. (eg. commands at end of one minute, game pauses) It can be more immersive than a RTS as the player has far more time to spot and manage things, allowing far more realistic games without overwhelming the player completely. It is scalable as in it can fight small and large battles equally well. (as opposed to tiny battles of boredom and massive battles of sheer chaos)

I mean, the complaint is mostly about the lack of time. Why not just add time.

[Vendetta]

If you are talking about a mechanics change, than I have no problem with that. If you are talking about developing inter-unit communication for something that can be solved by something simple than I would not agree. If you want to remove micro, do it at the mechanics level not at the "sub-player input level." Doing engine->AI->Player stack is double the work for the same effect.

Except, of course, this is not the "same effect" is it?

One of them looks like a competent military with modern levels of inter-unit communication, and the other doesn't.

As we have pointed out before, one of these things is an immersive experience for the player, which is what everyone in the entire world except you wants from a videogame. But then you're a weird fucker.

Tournament level miniature play is an very different environment than RTS tournament play. Meta-gaming is the "strategy" of the day where ever competitor try to out guess the opposing army composition and build an counter army. Once into the tournament, luck is an extremely important factor in both what opponents one draws and what tactical sistuations araise.

Actually, most tournament armies aren't built like this at all. Because when you go into a tournament you simply don't know which forces you are going to be competing against and with what objectives, so you have to field a force which deals relatively well with any situation.

To control a toggle, one needs two clicks. To tell a unit to target a specific unit, one needs two clicks too.

However, once the toggle is set once it takes effect without any further clicking. Unit targetting requires further clicking every time a new target is chosen, which with any reasonably competently designed unit scripting system (something I'm guessing you wouldn't recognise if it bit you on the arse), would happen far more often than would standing orders need to be changed.

[SirNitram]

...blah blah

Now just have a toggle where they'll either target the units with the leader max health, or most, depending on your strategy.

<Snip distortions and red herrings>

I mean, the complaint is mostly about the lack of time. Why not just add time.

I answered your question, troll. You said this:

If you want to discuss this topic, show me how one could possiblely program an AI that removes the need for the player to play fast in an nonlinear, discrete-step engagement. A rough sketch on how it respond to the bewildering set of sistuations would be nice.

You implied it was impossible or at least infeasible; it's as simple as existing code and an If-Then Statement. Now you're trying to wriggle away again, because you know you're full of shit, you just can't dare admit that.

[Elessar]

If you want to solve the problem of overkill, there is better ways than AI.

To do it with what would be worthy to be called "Artificial Intelligence", you need to:
1. Identify the enemy unit hitpoints and formation
2. Identify all other air defenses near by
3. Somehow split air defenses after using the data above

Do you always split hairs like this? Artifical intelligence is a catch-all term for a smarter way of doing things. It is a trivial problem for a turret to 'discover' all friendly turrets within range (Hint: they do it for hostiles already). Identifying the type of enemy is also trivial, seeing how the fog of war does not vary the moment you are in range. Therefore computation of expected firepower necessary to eliminate said enemy is done with awe-inspiring power of division, add fudge factor, allow that many turrets in range to go weapons free.

Turret refire can be modelled as an inventory problem. Can we wait to restock using the same supplier or do we have another take its place?

There is your 'AI'. No it doesn't achieve sentience. It just acts a bit smarter.

As you've aluded to, The better way is to make the weapons be mechanically "instaneous" thus do not waste any firepower as the target is tagged as dead before it is over killed. (as a aircraft is unlikely to be in range of too many units before the computer can compute a kill as oppose to the slow missile animation) Graphically one can still draw slow missiles however. That is a game mechanics fix that is far easier than try to coordinate and entire set of air defenses. This effect can be done by a editor to unit stats in almost any RTS without too much hassel as developing inter-unit communication intelligence.

I mean, for gods sake, even C&C1 has "hit scan" weaponary in the form of minigunners. There is no need for an more advanced engine for this sort of stuff. No need for any new "super innovation" but just tweeks to the engine that any bored modder can do.

This brainstorm of how the game is modelled internally has no place in the actual design of the algorithm.

If you are talking about a mechanics change, than I have no problem with that. If you are talking about developing inter-unit communication for something that can be solved by something simple than I would not agree. If you want to remove micro, do it at the mechanics level not at the "sub-player input level." Doing engine->AI->Player stack is double the work for the same effect.

Inter-unit communication? Fancy words for extending hostile alertness to friendly units as well. No developer would write routines to do inter-unit messaging when the game doesn't even bother modelling distant unit reports to the player. The end result is the same for the player -- the units act smarter -- and that's what matters.

[GuppyShark]

Care to explain this? There's nothing random about the starting conditions of a tabletop wargame. I pick my army, you pick your army, we deploy and engage.

In a tournament miniature game, there is ALOT of randomness.
1. Armies are build before hand, thus one has to guess the opposing army composition as oppose to doing it on the fly.
2. Terrain is set up just before the battle. Missions are determined by Dice rolls.
3. The first turn (in games that have it), is often determined by dice rolls too.
4. There is alot of randomness in combat computation. Because there is few hit points on the table top, most weapons deal damage directly and good and bad rolls can have critical effects on battle.

Tournament level miniature play is an very different environment than RTS tournament play. Meta-gaming is the "strategy" of the day where ever competitor try to out guess the opposing army composition and build an counter army. Once into the tournament, luck is an extremely important factor in both what opponents one draws and what tactical sistuations araise.

Sure, if you want to talk about shit tournament level miniature games. For fuck's sake, you can win every game in a 40k tournament and still come in the middle of the pack, because the winning the games is only part of your score.

In the WARMACHINE format you prepare two lists beforehand and can pick which you take after seeing your opponent's lists, terrain is symmetrical, scenario is preset, the first turn is rarely important (there's only one scenario that it advantages in the current format and they're likely to remove it for this reason for the next iteration) and you throw enough dice that a few lucky rolls aren't going to let a scrub win the tourney.

Randomness isn't inherent in the tabletop gaming genre any more than retarded unit AI is inherent to RTS.

I also love that you put strategy in scare quotes when talking about tabletop gamers taking an appropriate force for the expected enemy. Why is that if you only have a decent guess of encountering armour it's not 'strategy' to take anti-armour weapons but if you know for certain that the enemy is taking them it is?

EDIT: In GW's defence they've never marketed 40k as a serious tournament game, IIRC. To actually treat the ever-expanding and -growing game of 40k as a serious competitive game would be counterproductive as it would be extremely hard to balance. The most competitive games become static - see Counterstrike, Chess, and Poker.

[Adrian Laguna]

Here, I skimmed through the front page and found this snippet for you. Please don't make me sludge through their godawful website any more.

2. PDS "Experimental" - NGCS, V7.6 and V9.0. These three separate projects attempt to base HW2 gameplay upon semi-Newtonian combat maneuvering and a significantly larger range of engagement. If you want to see lasers and strategic missiles engage from "beyond visual range" while ships zip along "dogfighting" at up to 5km/s, get V7.6. V9.0 Alpha is more of a compromise that serves as the basis for our latest project,

There's a couple of videos (found here) that let you see NGCS in action. From the looks of it, you have ships shooting lasers and spitting waves of nuclear missiles across some fairly large distances. Apparently they made a beta but has since been abandoned. They say it's been merged with something else, so they might still be toying with the concept.

The problem with the website is that it's designed for people who are already very familiar with the mod. The same applies for the board culture in particular, I've seen a few threads asking the same questions I have in my mind and the responses are rather unhelpful. It's really hard to understand what all the different versions they have do unless you've been following it from the beginning.

[RIPP_n_WIPE]

One thing I really wish, in addition to the limited "Commander sight" (and I know this isn't going to happen for a long long time) is instead of "bonus'" give to units vs other units but to have their differences balanced out closer to the way they do in real life. Pole arms don't magically harm cavalry more than regular troops. They hurt them more when charging because they can hit cavalry before cavalry can hit them and because a 1500lbs horse moving at fifteen miles an hour is gonna get skewered on a diverging point that's less than .5mm^2. Cavalry doesn't magically do more damage to archers. They hurt them more because they have light armor and can't defend against cavalry and so on. That of course would probably require complex calcs for velocity, armor, penetration and such that probably won't be seen in an RTS for at least a few years.

[Covenant]

Ripp, you can see some of these things coming into play in the Total War series of games, but precious rarely elsewhere. They're a good example of how these sort of things can be done, and it's certainly not hard to do. Generally, it makes perfect sense for pikemen to have a bonus to damage versus cavalry because generally cavalry are the only ones moving that speed. You can also make it so they have a bonus to anyone moving at a 'charge', and have that include both running infantry and horsemen.

And so on, of course. What we need are less unit bonuses for things that don't make too much specific sense (cavlary versus archers) and specific bonuses that do (bonus to ramming spears into charging horses when set, bonuses to the rears of tanks, etc).

Adrian, that link seems toast.

[RIPP_n_WIPE]

Slightly off topic, of mods for HW2 I will saw I really like the Warlords Mod. So much starwarsy goodness .

[Covenant]

Slightly off topic, of mods for HW2 I will saw I really like the Warlords Mod. So much starwarsy goodness .

The first version I saw was extremely poor. I stuck some of those ion cannon platforms in my base while I tried to pour over the obtuse sets of ship designs, and there was a giant pile of permanently ionized ISDs out at my perimeter. I was completely inable to take them over, which made me rather sad.

Homeworld is one of those games that I think would benefit greatly from some pre-bought "Tabletop Style" unit layouts. With some minor replacements mid-game, that is. Perhaps it's a calculated risk--you can purchase a carrier for points and attempt to construct an armada, or buy two battlecruisers for the same cost.

I'd be interested in seeing the new versions. I wonder if anyone around here still plays HW2.

[brianeyci]

I think it would be nice to have "deployment" instead of "recruiting" as the means of getting troops and vehicles onto the battlefield. In other words, assume you have a fixed number X of units offscreen which can be deployed to the battle. This makes a lot more sense than magically constructing tanks and infantry out of thin air anyway. Building more landing pads or naval unloading facilities or some other logistical structure allows you to bring these available troops to the battle faster or slower. But this way, there's a limit to the number of troops you can make in total. To a certain extent, you have to work with what you have, and success in the campaign game affects how many units are available for any given battle.

Many RTS games have a unit limit, but that only limits how many units can be onscreen at any given time. None of them, to my knowledge, limit the number of units you can recruit for the entire battle. And this is absurd, since it means that you effectively have unlimited troops for any battle as long as you can hold your base together. This also gives you a rather strong incentive to treat the lives of your troops as valuable.

Sorry DW, that's too much like a wargame and turn-based strategy like Panzer General and X General lookalikes or whatever is passing for turn-based strategy these days. We can't have that, because RTS is awesome and clickfests are sacred. Amen.

On the serious side, I didn't play it but that new army Future Combat wanking game had that kind of stuff from the screenshots. From the screenshots you could pick and choose which units to deploy, and you didn't really build a base. I wonder if it was a good game at all. I'm so out of touch with RTS nowadays, the last one I played for any length was in high school and I've only seen screenshots or people playing them or played them for half an hour at a time. So from an outsider's point of view, clickfests blow and micro blows. I used to do it, but then again I used to like cartoons too.

EDIT: Just skimming through the thread, is it real he thinks everybody should have identical starting points in RTS? Everything he's saying seems colored by Starcraft. I remember playing Starcraft in high school and it was a really big deal, everybody having the same starting resources and same terrain. Playing Lost Temple over and over again or playing Big Game Hunters where everybody had the same starting stuff. Of course this was a fundamental flaw in the game design, a person should still be competitive even if he starts in a slightly shittier position. Why is identical placement anything to brag about, if anything it means a flawed game.

[Edward Yee]

Reading the comparisons to TBS made me realize why I've been playing TBS for over a year now -- when it's strategic TB/tactical RT, I aim to operate in such a way as to avoid having to actually handle the combat, delegating to the AI whenever I can. And when the battle itself is turn-based and battle/construction are both on the same map, i.e. in Advance Wars... in that example, at least, unlike StarCraft, terrain matters very much. Thermopylae and supply lines for the fucking WIN.

[ray245]

I hear that there's a game call World in conflict...(can't remember where I heard that from), and from what I heard about the game...it seem TACTICS are required rather than building a base and spending time building army.

Anyway...let's do a vote count.

WHO HATES base building here?

[Elaro]

Well, I don't HATE it per se, but I do think it could be less of a chore. Maybe have a way to draw plans for a base beforehand, and tell the AI: "Do this"?

That'd be fun.

[CaptHawkeye]

I'm getting a little tired of the RTS tendancy to make troops so easily replaceable. Even games which incorporate unit experience still give little reason to keep men alive. Most "conquest" style RTs games come down to just who can capture the most cities in the shortest amount of time and who can spam the unit creation button. War has always been defined by the goal of one army to corner and destroy another. Conquering cities and destroying industry helps to bring an enemy to his knees, but it's all primarily to handicap the enemy in the end or just starve him. Wars are not won by taking random villages and farming fields scattered throughout the countryside.

[Darth Wong]

I'm getting a little tired of the RTS tendancy to make troops so easily replaceable.

That's exactly what I was thinking, with the "limit on total deployable troops" change. When I play Total War games, I have to think about how to use my limited number of troops most efficiently. I can't throw away their lives the way I would in an RTS game.

[Hotfoot]

I'm getting a little tired of the RTS tendancy to make troops so easily replaceable.

That's exactly what I was thinking, with the "limit on total deployable troops" change. When I play Total War games, I have to think about how to use my limited number of troops most efficiently. I can't throw away their lives the way I would in an RTS game.

I agree in that this would be interesting, but I also think that reinforcements add an interesting dynamic to the game and allow for a bit of flexibility and back and forth in battles.

A scheme where you recieve reinforcements from a limited pool might work. You would still have limited forces, but the fight wouldn't be over right away necessarily. You could also use this to create online scenarios that are more than just "You have an equal force, you meet in battle. Roar." and would allow you to create situations like Total War seiges, where the attacking force can be much greater than the defending force. Moreover, you could have a scenario where the defenders must hold the line/castle/base/whatever for however long until reinforcements arrive. Additionally the morale benefits from a weary group receiving said support from the reinforcements would (or should) be significant.

On another note, a game where rescuing wounded soldiers has an effect would be nice, something like giving a morale boost to the surrounding units while defending. Really though, a solid morale system just makes any RTS better. But I would like a system where the singleplayer campaign was persistant and the units had some more character. Something of a cross Ground Control and Hostile Waters so that you're more interested in the units beyond stats and numbers. The chatter from Company of Heroes would also be good, since its useful to figuring out what the hell is going on. Hearing "We need AT! German Armor!" from the infantry is FAR more useful than "units under attack" from a generic voice. Also, hearing things like "Kowalski's down! We need medevac!" every so often adds a lot to immersion.