I like the mission idea, Taz. I wonder if there could be a way to make a mission template and save it between games. This would allow the players to create combinations of orders that they could use during many games.

The problems with modern RTS games are hard to solve. IMHO, there are only two flaws with RTS games as they are today:

1) Twitch
Most RTS games are not strategic at all but games wherein the production of more units is the be-all-end-all. It doesn''t matter how tactical you are in the maneuvering and utiliziation of your troops. All that matters is that you produce more guys faster than the competition. While it''s still fun, it''s not strategic.

2) The Master Faction
There is always a faction in an RTS that beats all others with dumb strategies: Soviet Tank rushes, Protoss Carrier Stomping, and so forth. Blizzard has done the best job of balancing the power between their games'' factions but they are still imperfect.

Like I said, these aren''t easy problems to solve. Part of the twitchiness of RTS games is inherent in the fact that they''re Real Time. You run the risk of losing some of the "umph" of the real-time aspect of gameplay if you take some of the twitch out. And if you don''t make the races feel powerful (in their own rights) then nobody will want to play them. It''s a delicate balancing act that takes years of work and playtesting to accomplish.

I completely agree with you about the point that play testing is very important here in the ''feild'' of RTS games. I am going try to mix my love Genetic Programming in here to solve this ''kind of'' and for AI at the same time... Basically, I am also very interested Genetic Programming and will hopefully use it to create the AI for my game. By the nature of GP, it will hopefully find many of the gameplay quirks and loopholes that normally require human testers. This by all means does not mean that human testers are un-necessary, it just means that it can be used to reduce the amount of human time needed for testing. I am a little more optimistic with this GP stuff than most people though so it might turn out to not work at all but, its worth a try.

anyway, it seems that most of us here are on the general consensus that the stradegy in most RTS game are quicke lacking if present at all. Hopefully Me and now it looks like maybe some others, or just one other, will be able to create a game that requires some extra stradegy... Luckily, it looks like most of us all want the same kind of things so I guess I''ll get programming ;P

Dwiel

I'm working on a RTS as well, with some friends, I liked a lot of the ideas brought up in this thread as many of them were already in my design doc.

I just have a few comments though about some of the ideas.

"I would like for a player to be able to defeat an army of say 100 units with a meager 10 if strategy is used by the player with 10 and not with the one of 100... I REALLY would like strategy to be a focus..."
Well the problem with this is that both teams have very similar units so 100 units with a little strategy is better then 10 units with very good strategy, unless the 10 units can stay out of sight. Now the problem with this is that "out of sight" is normally an engine rule, not a strategy; so the 10 units can manage to say out of sight then the 100 units would also be able to stay out of sight.


For the topic on balancing, ya its very hard so my team decided to drop to just having 2 races, a 3rd is just too much work.


Custom Units: i also had this idea but we dropped it because it makes the game less memorable. When you think of warcraft or starcraft you can picture 2 or 3 races and know every unit for each of them, maybe not by name but by look. This kind of thing makes people remember your game after they stop playing it.


"Oh, totally random maps so that no one can memorize them and have an advantage in multiplayer"
Well if you do that there is no strategy based on maps, its like an if-and-only-if. Also, people will never have a favorite map; and the more favorites you can give a player the more they like the game.

"moral; this was in close combat, wouldn't you be worried if you were facing down a panzer with a rifle?"
Okay, this should never be done in a RTS. It sounds like a cool idea but its the biggest pain in the ass when a player doesn’t have control over their units. I think it was a postmortem for C&C that i read and said something along the lines of "we changed the script for a harvester so that when it was attacked it would run away. Then we decided to put the new script on a developer's computer without telling him. The next day we heard the developer yelling at his harvester because he couldn’t control it once it started running away after it was shot at." or something along those lines.


Moral idea 2: the unit has worse aim or something. This is ok, as long as the player knows it, but dont count on them figuring it out on their own.


Ok, thats it. I dont mean to be too critical or bashing any ones ideas, I simply wanted to show the other side of the coin.


Tazzel3D, Good luck on your game and try not to get bogged down by feature creep.

PS: Custom Units - also it doesnt allow the players to have a favorite unit and doesnt allow your game to have flag units. If anyone says the word "Peon" who doesnt think of warcraft right away? I sure as hell do.

[edited by - TheDarkening on January 13, 2004 7:22:35 PM]

Thanks for the great post and comments Drakening. I think you made some good points, and we should make sure that we dont go around adding ''gameplay features'' just for the sake of adding features. Its a very delicate situation. I really liked my cutomizable unit idea, but as it is turning out, I am not sure how good the idea is in practice... looks like its one of those things that sounds good conceptually, but then once implemented or attempted to be implemented, doesnt really work out as planned. Maybe I will work on a middle ground somewhere.

I agree in that we should also not make units poor fighters or scared of battle if moral is low... UNLESS, we dont make them make to drastic of change, AND the change is very easily predictable. I think that if you alter the units speed factor when moral is low, that is OK as long as you say in the stats box somewehere: "Speed: 20 -3(low moral)", etc... Just so long as the player isnt left guessing as to why his units are doing what is expected. Still, I think that is important that this sint overdone.

One problem that I am running into is units'' scale. In most RTS games, the units are much bigger than they ''should be''. It is unimportant in most of these. In the case where we want terrain to influence gameplay, this is unfortunately not acceptable. One solution is to make the units much smaller and leave the scale of the terrain alone. This ok, unless we would like the players to be able to see their units...

About the only other option is to change the scale of the terrain. By doing this, we nearly force ourselvs to also allow for the camera to be a varying height off of the ground... I am thinking about either allowing for the player to change the zoom, or for their to be some choosable zoom settings. maybe a close-up, a middle and a way out. I even thought it might be cool if we were able to be in the zoomed at view, and then click other parts of the screen to show a close-up of that portion. You could have maybe 4 closeups each showing the movement of a specific group, or just a close-up of your various bases. This way, you can view the big picture, and at the same time, control small units precisely.

Well, what do you think about the scaling?

Thanks for all of the input!

Dwiel

"If anyone says the word "Peon" who doesnt think of warcraft right away? I sure as hell do. "
The first thing that pops into my mind is the phrase "Stop poking me" Probaly one of most memorable game lines ever...

moral reminds of rts game I was working on a number of years ago. In it one of the factions had a unit known as an empath. The empaths where passive units that increased the effectivness of friendly units and decreased the effective of enemy units within a radius around it. Moral could work in more or less the same way. You decrease or increase a units effectivness. For instance -1 moral could mean a 10% decrease to all stats. While +1 could mean a 10% increase to all stats. Which could be enough to effect the out come of battle as a larger force of weaker troops takes down a much more powerful demoralized enemy.

My rts also had customizable units. Basicly units where seperated into diffrent chasis with slots. You choose what when into each slot. For instance there was an infantry chasis with a primary weapon slot, a secondary weapon slot, armor, headgear and backpack slot. You could also select aditonal training for units to recive such as field medic or scouting. Of course diffrent items had diffrent stats and costs so you had to choose what was best for the unit you where making and for your budget.

You could also reaserch new equipment, chasis and skills as the game went on.

-----------------------------------------------------
Writer, Programer, Cook, I''m a Jack of all Trades
Current Design project
Chaos Factor Design Document

If you want to check out a RTS that uses customizable units then check out "Impossible Creatures", i have the demo on a PC gamer CD so im sure you can find one on the net somewhere. I didn’t like that game though because the combat felt so fake.

Zooming:
Well if you make it so that you can zoom out then I’d always be playing it fully zoomed out. The more I can see the easier I can coordinate my units. I''d only ever zoom in to take screen shots.
If there were a reason for zooming in then I guess i would have to. For example, an enemy ranger cant be seen if the camera is X zoomed. If that were the case though you should make it have like 3, or so, levels of zoom so i know when i can see enemy rangers and when i cant. If you do zooming then make sure you do some level of detail stuff or your fps will drop.

The scale:
Ya... I eventually had to convince myself that each guy really represented like 20 or so and that’s why they were so big and could take so much damage. It would be cool though to have units scaled closer to the world''s scale but in order to make it look good you might have to increase the number of units the player can produce and lower their health or keep the game fairly zoomed in. Just got to make sure that you don’t have vast open fields with like 20 ants running around on it

For the game I''m working on, we have an idea of our scale but for the most part we are just going to play with the scale later and see what looks best.

TechnoGoth, is this RTS you talk of finished? If so do you have a link to it?

Id also be interested in seeing that RTS if it is finsihed... I would be just as interested if it wasnt complete too, especially if youde be generous enough to share the source

Dwiel

Someone mentioned scale
What would make things interesting is to have a proper representation of scale in battles. It all depends if you want to make a "realistic" RTS (i.e. based on real war, like Age of Empires II) or an imaginary RTS a la Warcraft....

Anyway, my point is, having artillery being able to shoot actual 2 kilometers away instead of the 50 meters they usually manage at best in current RTS''s would be good. Having battleships the size of battleships and not that of fishing boats, etc... Maybe it would make a RTS a Real Time Strategy game and not a Real Time Close Combat game Of course that creates gameplay and technical issues, but let''s dream nonetheless

Anyone remember Total Annihilation and the effect Big Berthas (extreme long range cannon) had in the game ? Some good strategic playing was involved in making original use of them (against human opponents of course). Just building one was a challenge in itself, often enough subtle diversions were created to keep the opponent busy from finding out about its construction (within range of his base, if possible ), and that could make the game very psychological. But to achieve such subtelty in gameplay, Total Annihilation games had to last long (more than 1 hour at the very least, 2 hours being a good average on a medium sized map), and that length and complexity put a lot of players off. Can''t have the cake and eat it !