One thing I just started doing, was searching for RTS reviews on google. Surprisingly enough many people have decent ideas and speak them out when posting their reviews. You hear them saying in almoast every post: "If [insert name of game being reviewed] would have had [insert idea for the game], it would have been [insert very high number] times better!". Sometimes all they say is that the AI should have been better or this or that. Sometimes though they have decent ideas... just thought you guys might want to do some searching in your free time and see if you come up with any new ideas to add to your design.

Tazzel3d ~ Dwiel

P.S. still hoping someone replies to the previous post :D I''ve really liked this thread so far

One game that had customizable units was robo rumble. I liked being able to make units as I liked them. I didn''t like that everytime I wanted to make such a unit I was wasting time clicking the right combination and sometimes ordering a wrong combination.

What I think would be interesting is approach it a bit like an on-line collectable card game (bear with me here). There you have to build a deck before you play, so you have to make preperations beforehand, depending on the playstyle you intend to use. Some are geared toward one goal (rushing or defending and building up a superior force). I think this same concept could be applied to this system: you can customize your units beforehand. Choose maybe 8-15 units you intend to use. You could each give them their own AI type beforehand:
none- only respond to commands
defensive - run away when enemy units in sight or when being hit
guard - attack enemies when in sight, but hold ground
offensive - attack units that come near and persue a small bit
hunt - attack and persue indefinetaly

There could be pre-made units for those that don''t want to customize their own units for different purposes and then the player could make a selection of units that would make a ''deck''.

Do you want some really cheap, devastating units with good speed but nearly die when you poke them?
Maybe a few cheap ones that are fast to build that hardly do damage but can soak damage themselves and some of the devastating units?
Or some more expensive ones that are much more effective a little bit later in the game?

I think this would eliminate the in-game time wasted on designing units. It would have some mass production, but if the game is time based & very detailed, it becomes very hard to play.

Anyways that''s my rant of the day.

Thanx for the ideas. I understand that it would be very anoying to have to customize each unit that you create, and so thought that it might be a good idea to have the game save the unit''s specs each time you create a unit so that you can make another one very quickly next time you want one like it. Another icon would be placed in the array of default units to create. One thing that I just thought of, which is based on your idea of having customizable units which you create before the game. One problem with the predefined units that you create before the game starts is that in my game you can upgrade every attribute of a unit and so your predefined units might either have attributes not yet available or might even be out dated by the time you go to create them. A solution might be to have customizable ''templates'' for units instead. For example, you can create a unit which is called ''defensive tank''. It has the best armor available at the time (whatever that might be) and a gun that is as powerful as your current technology can handel. The AI though is the AI that costs 50% of that of thebest AI available. It has 5 small lasers that each have 10% as much power as your most powerful lasers. Do you see what I mean? Every attribute would be defined in terms relative to whatever technology is currently available. Could get confusing though... not sure... it shouldn''t be to confusing.........

Do you guys think that it would be a good idea to design all of these features now, in the design process, or create the basic game and then when you test it, add features that seem like they would help. Every game I''ve made som far follows the later approach, though I don''t know if it is the best one. Are specifc design issues best dealt with in the design process or afterwards when it is easier to evaluate its helpfullness and amount of gameplay it would add?

Thanx for all of the ideas/input/output...

Tazzel3d ~ Dwiel

A different way to handle it would be to allow scripting for the upgrades when creating a unit, so you could say "ok this unit will use type 2 lasers. when I research type 4 lasers, change them to type 3. when I get type 6, change them to type 4" or something simmilar. Not sure how you would handle pricing like that though. A relative number would probably be best. Instead of saying "50% of the power of the best" just say "level 5 lasers" or something easy like that from 1-10 where 10 = 100% and 1 = 10%. Makes it easier to understand. Every time you upgrade, it would effect just how good each level of whatever is.

Hey everyone,

Long time no speak Dauntless, I am off tripping overseas but I finally got a chance to visit the forums, good to see the RTS talk is still running hot.

I have read all of your posts and there are some great points brought up. I think we need to stear clear of forcing the player to do things with regards to unit balancing.

I think the "perfect" unit needs to be not the one which has the best force to cost ratio but the one that a certain general needs at the time. If your opponent builds only aircraft and you are good enough to scout this out at the beginning then your perfect weapon is a strong Anti-Aircraft tank. As Dauntless has pointed out countless times, games need to start taking terrain and such things into account, this should quickly stop the "perfect" weapon from being the one with the best power to cost ratio.

I agree with JNewt in that Warzone 2100 had a great way of designing weapons, you have your chasis and stick the weapons you want on it.

Anyways, it is getting late here so I will leave it at that, I hope all of your RTS games are going well and I look forward to seeing some early Alpha''s from you all soon .

Doolwind

Ok, so lets see if I''m understanding you here.... If I''m not, this method seems to me to be very basic and understandable:

you a scale from 1 to 10 where 1 is equivelent to 10% of the best currently available level, and 10 is the best currently available level. One thing though I''m trying to stay away from is having certain attributes ''maxing out''. I hate how in most RTS games, you get to a certain place in the game where your units are the best they can ever get. You''ve upgraded EVERYTHING. This is SO unrealistic when you think about it. In realaty, countries spend a lot of time/money just keeping up with the rest of the world''s technology let alone surpassing everyone. My point being that in real life, there is no point where technology is maxed out... or if there is it... well it doesn''t matter, you know what I mean.

Am I right about replacing a percentage with a relitive number between 1 and 10?

glad this thread is still going thanx everyone! There''s still a lot of potential to be gained here!

Tazzel3d ~ Dwiel

I''m saying to replace % with 1-10 yes, but I think you missed what I meant.
10 would represent the best weapon currently available. Every time you get a new, better weapon, that becomes weapon level 10 and everything else is moved down.

You would probably want to branch out a little, though - instead of just ''level 10 weapon'' you might have ''level 10 missile'' and ''level 10 slug'' and ''level 10 laser'' etc.

Level 10 is always the best, so every time you upgrade, the newest best thing becomes level 10, and what used to be level 10 is now level 9, etc.

There doesn''t have to be any kind of cap at all, its still all relative to your current technology.

OOOHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHH
I see!!!

thanx for the clarification.
So there would be no way to clarify that you want to use the cheapest level then? Well, you could, but the cheapest/smallest level would always be 10 levels under neath the best one..... I''m not sure of this is important... I''m just looking at the pros and cons of each method. With the percent method, where 1 = 10% and 10 = 100%, you might not have as much precicion with which level you want.. ie say you have researched 100 levels and you want say the 85th... that''s not an option even though there is a big difference between 80, 90 and 85. with the other method, when you get to any level higher than 10, you start disabling levels as you create new ones... Maybe give the user an option?.... maybe code both and then see which one would be used more often in context?

wait... ok I like your idea better now because normally technology becomes obsolete very quickly, there is no need to create level 1 of 100 technology when you can make level 80 for about as cheap. old technology is old technolgy. period. Depending on the curve I use to determine the cost of a technology based on time its been known about, I will increase\decrease the number of level available... That''s all part of the balancing though. Anyway I like your idea, and will most likely implement it.

one other thing I have been thinking about is if it would be feasible to allow the user to give conditional commands. technichally feasible and also feasible to assume that the user will be able to understand it and use it to their advantage.

I was hoping on allowing the user to give commands such as:

go to position x, y
IF you see an enimy unit, avoid it

or:

go to position x, y
IF you are attacked, don''t keep going

I was thinking maybe have a list of events that could happen, ie. the unit sees an enimy unit, the unit is attacked by an enimy unit... etc. and then have a list of actions that the unit could ensue. such a list might include avoid the unit, attack the unit, follow the unit, etc... keep in mind that all actions might not correspond to all events and visa versa....

what do you guys think?

Tazzel3d ~ Dwiel

A good solution for conditional commands (at least in response to the enemy) is to have a button that represents the current state of the unit, things like "agressive", "defensive", "ignore enemy", "hide if spotted" etc. Clicking on it changes it to the next one and right clicking changes it to the previous. Have a different icon for each state and its easy to tell what that one unit will do when it meets an enemy.
In:
aggressive mode, it will attack any enemy it sees.
defensive mode, it will attack only if attacked.
ignore enemy mode, it will keep going even if shot at.
hide if spotted mode, it will move stealthily and if seen, it will run away and try to hide.

The reason why I was trying to think of ways to better the player''s control of the units was because if I am going to be making location, terrain ect have a large effect, users will get very annoyed when they can''t get the units to do what they want them to do. When the stradegy involved in the gameplay is more complex, the method of command will most likely need to be fairly more complex than one with less strategic gameplay. So either the AI will need to be very smart, or the user will need to have very good control of each unit. I guess the AI should also be more complex in order to be able to either take more complex commands, or interbret more basic commands and employ stradegy them selfs.... Again I think it will be easier to tell how this should work once the basic engine and gameplay is up and running, but might it not also be more difficult to implement into a system that has already been designed?

In your opinion, how much of these details should be decided upon before coding of gameplay? How do you normally tell if something should be decided before coding or while in the proccess when it will be easier to test and play around with, but possibly harder to implement?

Thanx!

Tazzel3d ~ Dwiel

P.S. I love the near chat room feeling this thread has gained over the past couple hours...