I thought I'd start this thread to air a few ideas on how to improve the RTS genre, which IMHO needs a fair bit of improvement. Besides, I thought people might like a break from discussing RPG's.. 1. Why do people tank rush? Most RTS's fill the pattern of "get some defences up, advance till you can build the best unit, then build squillions of them and wade into the enemy". Absolutely no strategy here at all, and I have never played an RTS where this kind of play isnt dominant. Why does this happen? a)Micromanagement is tedious and difficult Micromanaging units is pointless, so I dont think we should even try. Units should by default operate in groups, a la Ground Control / Shogun b)Coordinating cunning strategies is too difficult I always get the feeling that implementing a cunning plan is usually a lot of hassle, and the benefits simply arent worth it - in the end you might as well just bunch up a gigantic clod of troops and send them all in in one go, and swamp the enemy with numbers. c)Troops have little or no value Most players dont give a damn about their units - they just send them in and if they die, who cares? They just build some more. Unless of course, they have run out of resources, in which case they are doomed anyway. I recently played a (custom) starcraft map where I had 8 seige tanks at the start of the game, but I was unable to build any more, since they had been banned on that level. Suddenly those 8 seige tanks were incredibly valuable - just losing one was a tremendous blow. Therefore there must be some way of making the player want to conserve his units. d)Dominant/Dominated units A lot of RTS's go on about the huge variety of units in the game. Some boast about units that you can design yourself, giving you billions of possible combinations. But if all but two of these combinations are completely crap and pointless, what is the point in the other 999,999,998? There isnt any point, and you might as well do away with them. So once the player discovers which units are worth building, he just builds as many of those as he can. If you forced the player to use a variety of units, then he would be forced to think a bit more about his attacks. I think its worth looking at how table top games balance this sort of thing - take warhammer as an example. Your army list will have certain restrictions, eg "you may only have one cannon per unit of guardsmen" or "You may only have one unit of knights in your army". Although slightly false, it does prevent the player from abusing the more powerful units. e) Lack of intelligence By lack of intelligence, I dont mean the player is stupid, I mean he doesnt have enough info about the enemy and the terrain he is fighting over to make sensible tactical decisions, (unless that is, he buggers around building 'scout units' and explores the map, which to be honest, is tedious, and shouldnt really be necessary. Certainly in a modern/futuristic setting, it is absurd to think that the players forces dont have access to decent maps of the terrain. If people knew what the enemy was doing, they would be able to formulate strategies to exploit weaknesses. A tactical enemy would be aware that there might be weakspots in his defence, and try to work around/hide them, whereas a stupid enemy would ignore his weaknesses and leave himself vulnerable. And lose. Heres my idea... Instead of the standard "mine ore, build barracks, build men" style of game, how about something a bit different. Lets say you have a village with a given population. A certain proportion of these people will sign up, and become military units. However, not all people are cut out for all roles, some people may not make very good knights, but might be good at archery, others are just talentless cannon fodder and end up in the guard. Whether the player likes it or not, he ends up with these units which he might not have bothered building otherwise. The percentages of people signing on could by decided by a series of dynamic factors, for example, if your knights get all the glory, then lots of people will want to be knights. But if knights take heavy casualties, people will be less keen. Hence your army will evolve according to your style of play. Of course you can still tank rush, but it will prove a lot harder to win in the long run, since it is such a wasteful tactic. Also, I think that infantry can be made more useful by giving them the ability to occupy buildings. This not only provides them with considerable protection, but it also denies the building to the enemy. I think that controlling buildings is probably more valuable than blowing the shit out of them anyway. Finally, I propose doing away completely with the fog of war concept, and giving units sensible sight ranges based on their vertical elevation (making high ground suitably valuable) In a modern/futuristic game, perhaps you could also have a satellite that updates every so often (as it orbits the planet) revealing the enemy's current positions (well, those that could be seen from the air anyway) 2. Crap Air units Aerial units are a tough thing... They are very fast, very powerful, and very difficult to balance. I'd be inclined to avoid them completely (ie do a medieval/fantasy setting to completely sidestep the problem, but while I am here I might as well have a go at sorting them out) Basically air units tend to be a bit crap - they hover around unrealistically, they are either stupidly limited or ridiculously powerful. Yet while getting air superiority is a very important step in any modern/futuristic combat, it is not the end of the battle. Air units should be powerful, but they should not be so good that you can win a game with air units alone... Air units come in four basic flavours: fighters, bombers, multirole, and assault. You can probably add a fifth - support - but Ill ignore that for now. Fighters are the F15's, the interceptors. They are very effective at taking out other air units, but they arent really designed to take out ground units or buildings - they can, but they generally carry a small payload and therefore are very limited in the amount of damage they can do. Bombers (B52's etc) on the other hand, are virtually useless against other air units, and too slow to be of much use against specific ground units. Instead, these things attack a fixed area.... anything in that area basically gets the crap blown out of it. But these guys will always lose in a fight against interceptors, and are likely to take heavy losses from any anti aircraft weaponry they might be flying against. Multirole fighters (the F111's) are kind of halfway between fighters and bombers. They can dogfight (but not so well) and they can bomb (but not so well) - useful because of their versatility. Finally, assault craft are the A10's and apache gunships. These are primarily used to attack enemy units, and are designed to be bloody tough. I see the air units being used in a completely different manner to ground units. Every aircraft type has a range (based on fuel - bombers have a long range, multiroles have a medium range, fighters and assault craft have a short range). Maybe an icon appears at the bottom of the screen, telling you what units are in range given your current location. You can then select the unit by clicking the icon, and select a target, and the air units will immediately scramble to attack. A few moments later, the air strike occurs, and then the units return to base ready to reload (may take some time). Interceptors would automatically accompany bombers provided they are in range, and would automatically scramble to attack any other air units entering their range. Hmm. That'll do for now I think... Edited by - Sandman on April 28, 2001 7:29:56 PM

Some great ideas there man.

I like your idea of having a set group of people who can become certain things and prefer careers based on your style of play.

But just in case that doesnt lend itself to someones game how about this.

You mentioned warhammer, and how it has rules about units with other units, now I dont actualy play war hammer but some of my friends do so ive picked up on some of the rules

They will play games with X Point armies (normally 1000)

now I propose that instead of setting the population like you would in AOE or something, you set the point value. And then assign different points to each unit. For example if you make the best unit in the game 100 then in a 1000 point game one army could only have 10 of them.

Done right this could solve the problem.

I like the way you think Sandman.

I''ve been thinking about this as well. My ideas concerning this were:

1. No base building. You go in with a limted amount of supplies. If those run out, game over. You get supplies by bringing villages to your side.

2. No unit building. Limited number of men. You get men by recruiting from villages. I was thinking, however, that the villagers would sign up, and then you could allocate them to different tasks. The more people die under you, the less willing they are to join. So then you''d have to start drafting. Which would lead to a poorer quality soldier for your side, desertion, defection, running to Canada, etc.

3. You can only build large units(siege towers, ballistas, etc.) if you have enough men to crew it. People aren''t automatically ruslted up.

I like the point value idea J-Muff. I was thinking of something like that except in terms of people. The way I was thinking, the more "elite" troop formations would take 1 person away for that unit, plus however many it would take to train him. So, if you have a really elite ifantry guy, there''s another 12 people behind him that won''t be fighting.

Actually, a lot of these points were addressed in Myth and Myth 2, which are my favorite RTS'' of all time. In a good campeign, each unit is important (especially experienced archers). You start with a limited number of people, and if you lose some, well, they''re dead, and you''re down one more person. In multiplayer, you can really see that tactics are key. Everyone has the same number of points to choose troops, and so you choose a good makeup of troops and then use strategy to meet your goals. In some of the multiplayer games, the point was not to kill your enemy, but to do other things, like control land, for example, and often, the one with the bigger army in the end was not the one that wins the game.

It''s really a shame Myth wasn''t even more of a success. Anyway, I just thought that i''d interject that thought into the conversation.

Back to debugging my game.

-ben.c

"1. Why do people tank rush?
Most RTS''s fill the pattern of "get some defences up, advance till you can build the best unit, then build squillions of them and wade into the enemy". Absolutely no strategy here at all, and I have never played an RTS where this kind of play isnt dominant. Why does this happen?"

it is because the tech tree isn''t nice and smooth flowing. The ideal tech tree encourages a certain rate of investment, the correct rate is based on the game''s pacing. The correct way to make a tech tree is where each tech level counters the level below it. That encourages step by step teching, always trying to stay one step ahead. If you jump two or three steps ahead you get tend to get overwhelmed and die.

For example I am on the Strifeshadow beta. At first the game was all rushing all the time. Was it that infantry was too strong? No. Was it that the counter was too weak? No. It was that the tech trees were ordered incorrectly. For example three of the early Accursed units are skelletons, spectral archers, and dire changlings. skellies beat archers, archers beat changelings, and changelings beat skellies. The tech tree went skellies, archers, changelings. So if you were going skelletons and you wanted to tech you had to tech up to changelings to gain any benefit. Meanwhile your opponent is massing guys and he can destroy you. Even if he can''t destroy you he will be able to quickly tech up to archers to make your teching worthless. Thus neither player would ever tech. One of the other races had basically the same problem. I recommended that they change the tech trees, they did, and now it isn''t a rush game anymore.

Just getting the tech tree ordered right isn''t enough either. You have to actually make the units counter each other correctly. Sometimes you try to make a unit counter another, and in certain sized battles it does, but in real situations sorry no luck. For example in Starcraft Zealots are supposed to counter Dragoons, and in small numbers they do. However since Dragoons are ranged units they get an advantage in larger fights. Six rows of dragoons can fire at once, only one row of Zealots can slash at a time. Even worse you can achieve critical mass with ranged units, that is where units cannot even get more than a trivial amount of damage before dying. Complicated terrain also favors ranged units, heck pretty much ever game dynamic favors ranged units in some way.

That leads to another problem: versatility

Units should not be versatile. It either leads to superior units that dominate the game or inferior units that can''t be made stronger without turning them into superior units. If a unit is pretty good in every situation it will dominate the game. During the normal course of a game players will switch strategies a couple of times, each time taking on significant risk. Every switch is a gamble, will this tactic pay off or not? A versatile unit does not come with this risk, it is a sound investment. Too sound. Units shouldn''t be so narrow that they never get used but they should have clearly defined roles. No more hydras please.

"a)Micromanagement is tedious and difficult
Micromanaging units is pointless, so I dont think we should even try. Units should by default operate in groups, a la Ground Control / Shogun"

Not that hard, perhaps you are referring to macromangement? Microing a couple spellcasters isn''t difficult at all. Trying to macro them... yeah that''s not easy. There are techniques to make stuff like this easier and occasionally a new interface enhancement comes out so it isn''t that bad. The problem is unwieldly armies. What is so much better about controlling three huge armies versus three men? That''s why I am in favor of smaller unit counts, it leads to better control without sacrificing diversity.

"b)Coordinating cunning strategies is too difficult
I always get the feeling that implementing a cunning plan is usually a lot of hassle, and the benefits simply arent worth it - in the end you might as well just bunch up a gigantic clod of troops and send them all in in one go, and swamp the enemy with numbers. "

yes this is a problem, designers are working on it though. There is a trend towards and more controllable smaller armies coming.

"c)Troops have little or no value
Most players dont give a damn about their units - they just send them in and if they die, who cares? They just build some more."

yeah, there are a lot of fixes for this though. Making troops take longer to build for example, or making bases closer. So if some guy tries to overwhelm you and you beat him efficiently you can run over to his base and smash him some before he is able to build a new army. Unit experience instead of global upgrades also solves this. Now the guy with stronger units isn''t the one who has been pumping money, it is the guy who has been fighting smarter.

"d)Dominant/Dominated units"

just go with fewer units, that lets you focus on balance. If I am checking out a new RTS and I see them boast about how they have dozens of units per race I just hit the back button. I was playing a bit of Myth II a while back (got into it late, too bad) and there were so few units in the core set, like 6 I think, and everyone used the same set. It was such a good game though


e) Lack of intelligence
"(unless that is, he buggers around building ''scout units'' and explores the map, which to be honest, is tedious, and shouldnt really be necessary."

if you aren''t willing to scout you don''t really want to play RTS games. So much depth would be lost. Seriously scouting is a big big part of these games, the war of information can be just as important as the rest of the game. My most memorable games all involved intensive scouting

Thanks for the feedback guys... Ill answer some of your points.....

quote:
1. No base building. You go in with a limted amount of supplies. If those run out, game over. You get supplies by bringing villages to your side.

I agree with this wholeheartedly. I want to control armies, if I wanted to build stuff I''d play sim city. I think this could make for a much faster paced game - you can get stuck in right from the start instead of pissing around building factories.

quote:
2. No unit building. Limited number of men. You get men by recruiting from villages. I was thinking, however, that the villagers would sign up, and then you could allocate them to different tasks. The more people die under you, the less willing they are to join. So then you''d have to start drafting. Which would lead to a poorer quality soldier for your side, desertion, defection, running to Canada, etc.

Thats very similar what I had in mind, but it gives the player more control. The question is, how do you force him to build a balanced army, rather than massing the best troop type? I supposes you could combine this with my original idea, so that of all the people you have available to recruit, only a certain number will qualify for any given troop types. You can maximise your favourite units, but you still have to use them well or the maximum number you can recruit goes down next time around...

quote:
3. You can only build large units(siege towers, ballistas, etc.) if you have enough men to crew it. People aren''t automatically ruslted up.

Hmm. Seige engines in historical/fantasy games are a funny issue - I expect them to function very differently to ordinary units. However, in a modern/futuristic RTS there are things like tanks and so on that would require several people to crew (as well as associated maintenance staff. I think that there should be a law of diminishing returns asssociated with growing - in a small village at least half the population might be combat ready. In a large city, the percentage of military personnel is much lower, and more advanced units require more maintenance staff.

quote:
now I propose that instead of setting the population like you would in AOE or something, you set the point value. And then assign different points to each unit. For example if you make the best unit in the game 100 then in a 1000 point game one army could only have 10 of them.

Starcraft did something like this... Its a good idea but I think that arbitrary population limits are a bit false, and on its own this isnt enough to encourage players to build a balanced army. But it can still be factored in... If at any given time we only have X people available to recruit, but unit A requires Y people to maintain, you have the choice... wait until more people are available to recruit, or build unit B, which only requires Z people to build. And you still have a training time...

quote:
Units should not be versatile. It either leads to superior units that dominate the game or inferior units that can''t be made stronger without turning them into superior units

I totally agree. IMHO the protoss carriers in SC are far too powerful - they are incredibly effective against both air and ground units. Compare this to the terrran seige tank - an excellent unit, very effective at ground attack, but very vulnerable to air attack.

quote:
Not that hard, perhaps you are referring to macromangement? Microing a couple spellcasters isn''t difficult at all. Trying to macro them... yeah that''s not easy. There are techniques to make stuff like this easier and occasionally a new interface enhancement comes out so it isn''t that bad. The problem is unwieldly armies. What is so much better about controlling three huge armies versus three men? That''s why I am in favor of smaller unit counts, it leads to better control without sacrificing diversity.

Yes and no. Special abilities like spellcasters, as you say, simply become unworkable when used in large numbers. I also find it annoying when you group stuff together and tell it to go somewhere, and then lose the whole lot because you didnt babysit them on the way, and they got confused by a narrow passage and half of them wandered off into an enemy base instead. I think this calls for better AI routines, interface improvements and better unit grouping. As for making the unit count lower, this isnt really solving the problem. I want to be able to control vast armies, but I dont want to have to control every single little man in that army.

quote:
yeah, there are a lot of fixes for this though. Making troops take longer to build for example, or making bases closer

Doesnt work. Increasing rate of unit construction is no better solution than increasing the cost of the unit. It is too easy to get around - In starcraft you can get a gigantic fleet of carriers built in a very short space of time, simply by building several stargates. Experience systems may go some way to helping, but I think it needs to be made clear to the player that his units have experience - if he cant see it, he doesnt care. I also think that if he knows that losing all his best knights will mean that he simply cannot build as many knights next time he recruits, however many Knight Training buildings he might have, he would be a little more careful.

quote:
just go with fewer units, that lets you focus on balance. If I am checking out a new RTS and I see them boast about how they have dozens of units per race I just hit the back button. I was playing a bit of Myth II a while back (got into it late, too bad) and there were so few units in the core set, like 6 I think, and everyone used the same set. It was such a good game though

I completely agree. 6 well designed balanced units is a hundred times better than 600 badly designed unbalanced units. In any case, players do not want to learn the strengths and weaknesses of 600 different units - they will learn 5 or six and ignore the rest anyway.
The only situation I might disagree with this is in a game like SC where different races have very different units. I think this is a great idea, which adds a lot to the longevity of the game. Must have been a complete bastard to balance though....

quote:
if you aren''t willing to scout you don''t really want to play RTS games. So much depth would be lost. Seriously scouting is a big big part of these games, the war of information can be just as important as the rest of the game. My most memorable games all involved intensive scouting

I am not completely convinced... In my opinion, scouting (like building) is a distraction from the real purpose of the game which is to control an army. Now imagine this situation... Say you are tank rushed by a huge force. You know perfectly well that this force is so big there must be very little defending the opponents base, leaving it vulnerable. Now, if you know where his base is, then you can punish him severely, but if you dont, then you just lose to his tank rush and get beaten. Just think how differently you would play if you knew what the enemy was up to all of the time? (and he knew what you were up to ) Maybe this info shouldnt be completely free, you have to do work and invest in something to get it, but I think that it should be a lot less tedious than it currently is.

Hmm...... why tank push??
Some people tends to wall themselve inside the base and climb up the tech-tree real fast just to use some super advance unit. To counter these people fast, the easiest way to go is to mass produce some low-end unit and give these players some surprise since their base has less defense.

And you say there are dominant unit?
For a balance game, there''s no dominant units. Dominant it is, in the sense of unprepared player. IMHO Protoss Carriers almost rules everything in SC: it is easy to use, to attack, even maybe to defend. But with some Static fields, Dark Swarms, or some Lock-downs, your carrier-groups would be useless.

Of course, for an imbalance game, I can''t say anything.

And ....In a modern/futuristic game, perhaps you could also have a satellite that updates every so often (as it orbits the planet) revealing the enemy''s current positions (well, those that could be seen from the air anyway).... I think in modern/futuristic game, there should exist some intelligent or smart hacker units that scramble your satellite data so your maps won''t do you any good.

Anyway, I like your point. Especially (c)Troops have little or no value. They should have improvement in this. Good think''in!

"Doesnt work. Increasing rate of unit construction is no better solution than increasing the cost of the unit. It is too easy to get around - In starcraft you can get a gigantic fleet of carriers built in a very short space of time, simply by building several stargates."

ahh, I know what map you''re playing. That isn''t how SC was meant to be played, it isn''t the fault of the designers that the gameplay breaks down on that kind of map. Games need to be played on the maps that were used in testing the game. It never occurred to the designers that cannoning up and pumping carriers would be a viable strategy, and it isn''t, unless you are playing BGH. That is not the fault of the designers. Not trying to start an argument but you have to judge a game by the way it was designed to be played. That said carriers are a little too strong since 1.04 instead of being laughably useless like they were before. A minor downgrade is in order but the real problem is corsairs. Wraiths counter carriers quite nicely but corsairs completely devastate wraiths to a degree that is just too much.

As for making units take longer to build, it does work. However you need to make the unit production buildings cost more. That way if they try to build more unit production buildings they will have no money for units. SC has the ratio wrong, buildings cost about the same as the units they produce. So say I am playing Stifeshadow and I have a workshop upgraded to build three units simultaneously (650+350+350) I get a cost of 1350. My opponent only upgrades his once (650+350) and he has a cost of 1000. Now it is still early game so we have both only pulled in 2000 resources so far. With Gremlin Carvers costing 250 each I can afford four, he can only buy two even though he has more production capability. Now I can just run over to his base and rip him up, if he doesn''t have any spike traps he is dead. If he does a couple have spike traps at a cost of 50 each then he probably isn''t going to be attacking me. So you''ve got a lot of options (hard to explain, but the game has a really good early game that gets you into the action instead of making you wait five minutes) and no one strategy dominates.

Ok, first, I''ll make it clear that I''m not too brilliant at game design, I''m a programmer . I normally work on my own, do my own artwork (which sucks) and then once it is about finished, I find someone to do some better art for it, which I then put into my games. Because I work on my own, I also have to do the design, so normally I try and stay with fairly "tried and tested" things. Anyway, that''s the quick backgrounder

One of the projects I''m working on (well, the only real project I''m working on, but I''m doing a prototype for some of the concepts in it right now) is an RTS, it isn''t futuristic, it isn''t fantasy or medieval, it is completely abstract. As such, I don''t need to worry about realism, as long as I can keep the player familiar with the world.

What I''ve done, is taken a traditional AOE type RTS, and changed a few things:

Resources
The main ''resource'' in the game is the colour green. Every object in the game (objects being units, buildings and terrain) has a maximum capacity to hold green, so a bucket of green paint can hold a lot of green while a red bucket can''t hold any.

Any object which has green in it can be (thinking of verb and failing) mined by another unit as long as the unit can hold the extra green. Mining the green, is just a simple matter of moving up next to the object and instantly, the green is transfered from the object to the unit.

To build a unit, the structure which builds the unit must have enough green inside it to be able to pay for the unit, if it does, then the unit is created and the amount of green needed is subtracted from the structure''s green.

One of the units is the spanner, by putting a few spanners together, you can build structures, so one structure might cost 4 spanners, so you would put 4 spanners in the same position and the structure would be built.

(Yes, that is very confusing reading back over it, with all the green and so on.... but I can''t put it any other way)

Because green is so easy to mine, it makes stealing it from the enemy possible and often necessary.

There is also a short supply of green, meaning that wasting it is quite foolish

Tech tree
No tech tree as such, but when some units gain more experience, they might be able to upgrade for a cost (either time, or green, haven''t decided yet). The upgraded unit might not be unique either, a Type 1 unit might cost 50 green, a Type 2 unit might be 100 green. But if a Type 1 unit makes 5 kills, there might be an upgrade to Type 2 for 10 green or something like that.

Intelligence
Not sure now. I was going to have something similar to the traditional fog of war (like from C&C times, where once explored it is fully visible for the rest of the game) except after a few minutes, areas which haven''t been explored are ''fogged'' again.

Now I''m wondering whether to have fog of war or not, I think it adds another aspect to the game as different units can have different lines of sight, and recon missions can be done.

What would you guys recommend?

Trying is the first step towards failure.

ragonastick:

Wow, that has to be one of the most original concepts I have seen for a while. Its difficult to say what you should do about the fog of war... my reasons for suggesting it be removed are as follows....

1. It isnt terribly realistic, and if you want realistic strategies to develop you want to aim towards realism. This obviously does not apply to your game, since realism is not what it is about.

2. Knowing the terrain.. I dont believe that it is possible to plan a decent strategy without knowing something about the terrain you are fighting over. By including a fog of war, you are making scouting and exploration a very important part of the game (even more so if the fog regrows as the game progresses). Personally, I think that half the reason that scouting is encouraged is to justify the existence of a few fast weak units which are otherwise almost completely useless. (vultures anyone?) At least make it possible to use some other mechanism for scouting (eg the Starcraft Comsat station) that doesnt require quite so much hassle as babysitting a unit all the way around the map. Note that vision ranges are still relevent without a fog of war... spotting units, selecting units to fire
at, etc.

DerekSaw:

quote:
Some people tends to wall themselve inside the base and climb up the tech-tree real fast just to use some super advance unit. To counter these people fast, the easiest way to go is to mass produce some low-end unit and give these players some surprise since their base has less defense.

Surely a tank rush early in the game is no more tactically interesting than a tank rush late in the game? Its still a tank rush.

quote:
I think in modern/futuristic game, there should exist some intelligent or smart hacker units that scramble your satellite data so your maps won't do you any good.

Interesting idea... I'm going to have to think about how all that will work... I think the maps of the terrain shouldnt be effected, but mucking around with radar coverage might be interesting....

AP: Ive used it on many maps very successfully. The problem I have with most RTS's is that there is very little strategy in the individual battles. Imagine a game of starcraft, protoss VS protoss. Now imagine that one side has 12 zeaolots, the other has 6, and there are no other units left in the game. The side with the 12 zealots is being played by a lobotimised monkey which has been trained to recognise enemy units and click on them, the other side is being played by a military genius. Who wins? The answer is simple - the lobotomised monkey. I want a game in which the military genius can win.

Edited by - Sandman on April 30, 2001 7:16:54 PM