37 replies to this topic

[AltarofScience]

I am working on a game with supply lines. Basically resources are local and you have to work from stockpiles. So you place a foundation indicator with a stockpile. Then you have to physically move the resources in a unit to the foundation. But you can build in spurts and resources are spent over time so you don't need it all at once. And you also have some economics and moving weapons and armor to barracks and armories and so forth.

 

And you do indeed need to decide where to produce, and not only units, but do you process resources and create items on site of raw material gathering or do you ship it to a city. It would certainly be interesting to see how smart players are at dealing with those decisions.

[ShiftyCake]

@shiftycake

Why do you go in every thread referring to real time shooters? Real time shooter is not a genre. RTS means real time strategy. Shooters are FPS or TPS and they are all real time... This is the third thread you have done this in.

 

you are right, I actually typed it up all at once late at night and my brain did one of those switch word meanings things. Or acronyms in this case. Oh well, I'll have to edit my posts later and apologise.

[AltarofScience]

Its cool, it was just really weird cause you were making no sense.

[Dan Violet Sagmiller]

This goes to my second point. Decrease speed. And by speed, I don't mean how quickly units fight and whatnot, I mean actual movement speed. And increase sight ranges for "scout" units.
 
The idea is that you'll give players the ability to spot enemy movements from afar and think of a way to counter this.

 

Brilliant!  This is exactly what I was hoping for! I think that Dark Reign did this well.  the vision range to Movement speed ratio was very different in Dark Reign compared to Star Craft.  Some vehicles, that moved VERY SLOWLY, could shoot half way across a map.  Though there shots didn't do a huge amount of damage, you get a few of them together, and they could do huge amounts of damage.  Especially when the terrain held huge amounts of winding boulder ravines that these could shoot over.  sort of like siege tanks but not as powerful per shot, nor as rapid fire.

 

That also reminds me of a few other things.  You could build a look out post, right at the edge of the woods, and it would be hard to see.  pretty much it would only be noticed if enemy troops stopped next to it, or another player moved over it with their mouse.  Things could actually hide in the tree line in that game.  

[Dan Violet Sagmiller]

Buildings were insanely long to kill, and it was frustrating, but at least, it made it so that the concept of a settlement had a more permanent nature.

 

Good point.  Buildings don't typically collapse from gun fire, though they can be taken over with it.  But I agree about Warcraft 1, it did take a while.  same thing with walls that you could build.  Cheap, and excellent stoppers for oncoming troops, giving you time to react.

[AltarofScience]

Dropping actual move speed is something that I have been looking into. Most commercial devs and mainstream gamers don't like it though, because they want fast 30minute session games.

 

I have been looking into some sort of perception stat so you can hide units in tall grass or trees and so forth as well as reduced vision ranges. Not only for all factions but for specialist factions. Imagine a faction based on forest control. They see farther there, they hide better, they can move through normally impassible forest. Now moving in the forest is slower and harder to do formation, but if its big being the only one who can navigate it safely is a nice advantage, other groups having to take the long way around.

[Dan Violet Sagmiller]

I am working on a game with supply lines. Basically resources are local and you have to work from stockpiles. So you place a foundation indicator with a stockpile. Then you have to physically move the resources in a unit to the foundation. But you can build in spurts and resources are spent over time so you don't need it all at once. And you also have some economics and moving weapons and armor to barracks and armories and so forth.

 

And you do indeed need to decide where to produce, and not only units, but do you process resources and create items on site of raw material gathering or do you ship it to a city. It would certainly be interesting to see how smart players are at dealing with those decisions.

I'm starting to like the idea of having to follow the money.  I.e. supply lines to buildings.  Either by train, vehicle or building the factories in chains off of the supply mines/drills/etc...

[Dan Violet Sagmiller]

Dropping actual move speed is something that I have been looking into. Most commercial devs and mainstream gamers don't like it though, because they want fast 30minute session games.

 

true, but those are also based on completing a battle/war in 30 minutes or less as well.  This game is on going, with battles lasting longer periods, with or without player activity. That is a significant change to the modern RTS, and I believe would account for allowing the change.  

 

I have been looking into some sort of perception stat so you can hide units in tall grass or trees and so forth as well as reduced vision ranges. Not only for all factions but for specialist factions. Imagine a faction based on forest control. They see farther there, they hide better

 

This sounds quite good.  I know there are often upgrades about better vision.  But what about better hiding.  In starcraft, a group of troops left alone just stand there.  they don't stay near cover, or get down, or try to hide them selves.  (not including zerg burying, which you have to activate yourself)  but perhaps additional training can be given to troops to help them hide their information longer.  

 

"Knowing is half the battle" right?  this would be a great way to start putting that to use.  The art of deception should be part of war.  The AI Bar will tell you what things appear to be, and what is absolutely known.  Which will often be different.

 

they can move through normally impassible forest. Now moving in the forest is slower and harder to do formation, but if its big being the only one who can navigate it safely is a nice advantage, other groups having to take the long way around.

 

The skill to use certain items of scenery sounds like a good idea.  This could also include learning to use ditches/objects for cover, instead of just standing in the open shooting.

 

Perhaps Accuracy can play a bigger role here as well.  Just because you shoot at a target doesn't mean you'll hit it, and not just in a "not dealing as much damage" kind of way, but "shooting things behind it/next to it" kind of way.  So surrounding your enemy is great, so long as your guys have good accuracy, otherwise you might deal as much damage to your self as to others.  Also information/tactics use that the AI can provide info on.  with such fun info as "38% chance your troops will shoot them selves."  

[Yrjö P.]

Dropping actual move speed is something that I have been looking into. Most commercial devs and mainstream gamers don't like it though, because they want fast 30minute session games.

I don't think "30minute session games" are something designers and gamers just mysteriously "like". I think it is a nearly unavoidable consequence of making a good RTS game.

A good RTS game has to feature substantial average probability per time unit of winning by superior strategy. If that opportunity is absent, it means players do not get the chance to make meaningful strategic decisions. Then the game is a bad strategy game, or not a strategy game. If we assume the game does offer such opportunity, and further assume it averages out to some constant per player, then the probability of a 1v1 match still going on at time t is something like exponential distribution squared, P(t)=(1-e^-t)^2. That drops very fast.

I have been looking into some sort of perception stat so you can hide units in tall grass or trees and so forth as well as reduced vision ranges. Not only for all factions but for specialist factions. Imagine a faction based on forest control. They see farther there, they hide better, they can move through normally impassible forest. Now moving in the forest is slower and harder to do formation, but if its big being the only one who can navigate it safely is a nice advantage, other groups having to take the long way around.

FYI, the Wood Elf race in Warhammer Fantasy Battle is pretty much exactly like that.

Edited by Stroppy Katamari, 09 January 2013 - 05:47 AM.

[Orymus3]

Some vehicles, that moved VERY SLOWLY, could shoot half way across a map.

 

I would actually reduce the attacking range though. While trebuchets and the likes are what drives a turtling player out of their base, it does come with its set of drawbacks: it essentially becomes a building. If you have a mobile unit that can fire halfway across the map, even moving slow is not a sufficient drawback. You need the opponent to know the origin of the fire, and that they must strike this position. If the enemy can move back too easily, then there isn't any right play, and the opponent will just build the SAME unit.

You want units to counter one-another, not be dominant.

 

Everytime I approach RTS design, I try to do the following first:

Look at the gameplay mechanics, and insert some kind of infantry/pikeman/archer/cavalry "rock paper scissor" dummy units and see how they interact.

None of these units make it to the end, but they are essential in verifying my gameplay mechanics support this form of natural counters.

I don't give them artificial bonuses (such as say, infantry has +1 dmg vs cavalry) but these counters just naturally ascend from the actual abilities of these units.

Someone not totally up to speed with the design would not figure it out, but it transcends as is.

[shadowomf]

You need the opponent to know the origin of the fire, and that they must strike this position.

Actually if your game is based on our time or the future it should be no problem to give the player that information. Even today it is possible to detect incoming artillery and calculate where it came from. I guess using radar or some other high tech stuff.

 

That's why artillery isn't static anymore. It strikes and if counter artillery is expected it will change it's position. Just like snipers that move after a shot.

Self-propellered artillery makes much more sense for such situations.

 

I found it even more interesting that some artillery systems are capable to shot multiple shells and all hit the target at the same moment, preventing the units in the target area to take cover before the next shell hits.

They adjust the detonation used to shoot, then each shell is using a different trajectory (the first one takes the longest trajectory the last one the shortest). That way some system can hit a target with up too six shells simultaniously.

 

I would only make the strongest artillery units incapable of movement. They should however have a longer striking distance than all mobile units, else the will get shelled, the attacker withdraws out of range before the counter artillery hits and then start the whole thing again.

[Dan Violet Sagmiller]

I would actually reduce the attacking range though. While trebuchets and the likes are what drives a turtling player out of their base, it does come with its set of drawbacks: it essentially becomes a building. If you have a mobile unit that can fire halfway across the map, even moving slow is not a sufficient drawback. You need the opponent to know the origin of the fire, and that they must strike this position. If the enemy can move back too easily, then there isn't any right play, and the opponent will just build the SAME unit.
You want units to counter one-another, not be dominant.

 

Actually it worked out really well.  it would take a while for the shots to get to you, and resetting for a second shot takes a while.  Plus you have to have visibility there (like a Terran ghost setting up a nuke)  if it was your base, then you need to look for anything with visibility near by.  If it is against your troops, you just move them, even a little, and the Artillery would miss.  (not an area attack, only a targeted impact)  second, you didn't get to see where they where from, which created their own issues, but you could fire at targets you couldn't even see either.  For instance, you could target a black area of the map, and the shots would go there.  You would have no idea of a successful hit or not, but it worked.  

 

Because the Artillery were about half the speed as anything else, and took so long to fire, it was great in bulk when well defended, but easy to evade, if the target wasn't hit by 10 of these at once.  Regarding the origin, you know the direction they came from, and would start targeting with flying units that could get there pretty quick.  The Artillery could not shoot flying units.  In the game play, I enjoyed this mechanic and found it well balanced in the game.

 

Look at the gameplay mechanics, and insert some kind of infantry/pikeman/archer/cavalry "rock paper scissor" dummy units and see how they interact.

 

That's a good idea, and in general I would agree with it, but I've toying with a different idea now.  So far, Star Craft and others are based on rock scissors paper technology, but the handling of each is almost identical.  I've been considering different modes of game play depending on the race.  So its not just weapons/defense balancers, but also tactical balancers, in the sense of how things move and work together via AI, and Player control.  

 

Over all, you sound like you have very sage advice, and have some good experience in the RTS Area.  Thanks for your insight.

[AltarofScience]

Dropping actual move speed is something that I have been looking into. Most commercial devs and mainstream gamers don't like it though, because they want fast 30minute session games.

I don't think "30minute session games" are something designers and gamers just mysteriously "like". I think it is a nearly unavoidable consequence of making a good RTS game.

A good RTS game has to feature substantial average probability per time unit of winning by superior strategy. If that opportunity is absent, it means players do not get the chance to make meaningful strategic decisions. Then the game is a bad strategy game, or not a strategy game. If we assume the game does offer such opportunity, and further assume it averages out to some constant per player, then the probability of a 1v1 match still going on at time t is something like exponential distribution squared, P(t)=(1-e^-t)^2. That drops very fast.

That is not right. People specifically say that they want games around 30 minutes for scheduling reasons. Also 30 minutes is a function of map sizes and gameplay style and number of units. Do you really think its a coincidence that the optimal game would fall right into a good value for the average person's schedule?

Dominions 3 and other TBS games NEVER end in 30 minutes. Is that because you cannot make significant strategic decisions? Have you ever noticed that a certain segment of gamers tend to like the super long games more? That preference couldn't possibly correlated to their larger amount of free time  could it?

I have been looking into some sort of perception stat so you can hide units in tall grass or trees and so forth as well as reduced vision ranges. Not only for all factions but for specialist factions. Imagine a faction based on forest control. They see farther there, they hide better, they can move through normally impassible forest. Now moving in the forest is slower and harder to do formation, but if its big being the only one who can navigate it safely is a nice advantage, other groups having to take the long way around.

FYI, the Wood Elf race in Warhammer Fantasy Battle is pretty much exactly like that.

I'm sorry, I was not aware we were talking about table top war games...

[Dan Violet Sagmiller]

For instance, you could target a black area of the map, and the shots would go there.  You would have no idea of a successful hit or not, but it worked.

 

I correct myself, the opponent down the hall yelling "How the f..." does give you a clue as to whether or not you've made a good shot.

[Dan Violet Sagmiller]

To help clear up some confusion, this has nothing to do with the length of a battle here.  this is related to discussions elsewhere about an everlasting battle front RTS, where it doesn't end.  But I'm not looking for info on that either.  Primarily I'm looking at interactions between players on a tactical level, and how to bring back slower combat that works.

 

Dark Reign was slower movement, and it took about as long as a Starcraft map, about 15-25 minutes for a good game.  still very enjoyable.  Not nearly the marketing budget Blizzard had, so they took over.  (plus Dark Reign was only humans/vehicles, and did not provide the separate player mechanics so enjoyed in SC.)

[Orymus3]

That's a good idea, and in general I would agree with it, but I've toying with a different idea now. So far, Star Craft and others are based on rock scissors paper technology, but the handling of each is almost identical. I've been considering different modes of game play depending on the race. So its not just weapons/defense balancers, but also tactical balancers, in the sense of how things move and work together via AI, and Player control.

 

apologies if I didn't put the "etc." at the end.

Support units, and the likes all need to be prototyped as well. The idea though is to think of units by their basic function within the game. The "why would I want X unit?". The answer can be to counter another, or it can be because its cheap, or it can be that, if you have a wall of y and a cluster of z, X actually complements them and make them stronger.

[Dan Violet Sagmiller]

The idea though is to think of units by their basic function within the game. The "why would I want X unit?".

 

That is probably the most brilliant statement I've heard on this.  It would make sense to engineer my players characters with some basics.  for instance, develop a few characters for each race, of fairly uniqe types.  primarily answering to build time and movement issues.  but then, I'll go through each character for each race, and compare it to the other races.  Is there something that can resolve the issue that this character creates, regarding time/cost to create, abilities etc...  

 

for instance, an enemies soldiers need to be taken out, and we need ways to heal our troops, so in addition to standard soldiers of our own, we developed Mine setters, and medics.  Then the opposing side looks at mines, and generates devices to set off mines prior to troop engagement.  With mines taken being matched up, the first side then develops grenades to have similar effects, but with instant targeting, instead of presets.  So side b employed badminton players to volley the grenades back.    

 

A bit of a joke, but I get the idea.  it makes sense that that is how the military would proceed, why engineer something if it isn't resolving a challenge.

[Kenji Kousagi]

Improving a bit on standard RTS unit AI to cover basic commands like kiting and falling back when low on health. You could even do in game research options to improve AI or use a veterancy system to aid. I like an earlier post where if X unit survives long enough against Y race, it is just generally better against it.

 

I think ff12 included a programming for dummies style of combat. It relied on a series of easy to grasp if/then statements. (ex. If life is <50% then cast cure on self. This would be two things, life <50% and cast cure on self) The game's battle scheme was easy to grasp and allowed for a decent amount of options. I would at least suggest giving the combat system a look.

 

The player could set these up mid game or even use presets he/she made. You've got sniper units tasked to move to next nearest cover after firing and infantry that will only throw grenades at hardened targets (conserving ammo if the game uses that mechanic).