Hello there, I just recently had a great spark of an idea for a Real Time Strategy. It almost got me chucked out of a lecture because I was doodling lots of ideas and getting rather excited. The idea I don’t really wish to shout out until I have worked out most of it and even then I won’t give too much away apart from the story line because I rather want this idea to be made, even if it is just small time. But this isn't about me "rambling" on about an idea I won't tell you, I wonder if you could help me by supplying me with articles, forums, online books, or general information about RTS games, I play them a lot but my opinion is well limited and possible bias. Basically I want to know about the real complex things, I know how the games work; sure you build a base and then destroy the other team etc. I want to know about how certain units are at an advantage and how others are not, some are fast and some are strong, how do you create balance between "races". That sort of thing and in case any one is interested in the idea I’m not telling you just PM me or something. Cheers, Ian.

Well, you're basically asking how to design a game. The short answer is: math & statistics.

What the designers on my team do is:

1) come up with a general combat chain (a basic rock-paper-scissors)

2) come up with a general combat mechanic: starcraft has a weapon v. armor scheme. Lord of the Rings has a strict unit type v. unit type scheme. basically a certain thing will make one unit obliterate another (making it the paper to the other's rock)

3) come up with some basic personalities for your races: is one side's units like the zerg (cheap & weak) or are they like the protoss (expensive & powerful) or are they any other crazy combination. Every side will have a strength and a weakness.

2) create a giant excel spreadsheet with units for each side and their damage/armor/unit cost/tech tree investment cost numbers. balance those numbers so that each side has the same effective fighting force.

It's an n-body problem so you just have to work it out. Balance can be achieved by adjusting any one of the many variables. If it's too strong you can either increase cost, reduce weapon damage, reduce armor, reduce speed, or any combination of the above.

It may help to create a "unit worth" equation that is some function that takes as variables each of the unit stats and produces a magic number. A unit in that function would be balanced with another if their magic numbers were equivalent. This isn't easy to do and won't give you 100% accurate results but you can find outliers with it.

After that you play-test the crap out of the game. That spreadsheet just gives you the 90% of the way there numbers. After that you need to tune and tweak the numbers until all your testers consider the game balanced. After that you release it and the public will find 10-20 horrible exploits that you never noticed. Then you do a balance patch. Then more exploits -> another patch. etc. The better you do the initial planning the less exploits will be found.

-me

Thanks, thats exactly what i needed, so whats the first ever thing you would try to establish in your Desgin Documents?


Ian

First thing you need is what style of game. Always helpful. Since you already know this is going to be an RTS, you have to ask yourself: How should the game play?


There are a number of ways to build an RTS. Do you want to micromanage everything to make things work best? (As in Starcraft) or do you want to sit back and play it as a 'Big Picture' game, (as in Supreme Commander) or even a very different take, and make a Mixed Time Strategy game, such as Total War Games where you have Empire Mode, and then battle mode.

Then you need to ask what kind of units do you want? Lots of 'big' differences in units, like Starcraft's units (there isn't much you can say for a zergling and a Marine being similar) or more like Total War games. A unit of spear men are much the same as other units of spear men, which are much the same as sword men, but nothing is actually 'the same'.


Other things to consider are:

will 'units' be single men, as in Starcraft, or formation units where you get a number of men in each 'unit'?

Should battles be fast and over in seconds? or slower, more "Push of War" style, where even if things don't look great you can make good quick choices to turn the tide in your favour.

Is every unit you build of type X the same as every other type X? Like Starcraft, if you build a marine it is the exact same as every other one you've built, and if you get an upgrade for your marines they all get it. Or do you want your units to be able to gain experience or upgrades on a per unit basis?

Should your user deal with resources and unit and structure building at the same time as combat? Or take a Dogs of War/Ground Control route, where you get supply points after each battle to rework your forces for the next round of combat?

you've given me alot to think about and lots of great idea starters, i didn't realy want to let slip but i feel like telling you this game is going to work with actaully no units, so no ground control units like marines, and no vechiles ethier. its ganna work on a offence and deffence and econmical building system with optional upgrades and research. it's set in space and well thats all i want to reviel really.


Ian

Even if it is in space you'll still have to have some sort of units. Space Ships or something, or do you plan to use some sort of teleporter system and nothing really 'moves' but just changes location?

whether there are units or not the principles are the same. if you have different races they must be different in some interesting way. what makes choice interesting is tradeoffs. Do you give up individual unit strength so that you can produce tons of cheap units? Do you give up superweapons so that you have very powerful high-end units?

The choices you present in your RTS game are going to be the cornerstone of the design.

Once you have the choices outlined, the broad-strokes of difference between the races it's all the same balancing math techniques. Create statistical models of combat, evaluate the numbers, theorize about what number hooks will create the differences in behavior, combat, whatever. There's no "standard" methedology, you just need things to be: numerically oriented, provide meaningful choice to the player. A lot of things that players tend to think are interesting and cool ideas actually aren't in practice because they destroy a balanced game design. Your role as the designer is to make sure that everything is logically and numerically consistent and to realize that by adding even a tiny feature it can cause dramatic ripple effects of change in the design.

-me

Thank you!
This I understand, and I’m getting a great deal of useful information. This has not how ever put me off but more spurred me on. If you want to know how I plan to make this work you will have to PM as I don't feel right just telling everyone my plan because I consider it good and other people might too, that you can understand. I was thinking having optional tech-tree instead of "races" so like you have basic units, then you do a research or build something and it opens up new buildings, research, etc but closes others. The "space" part had nothing to do with not having units, it was a personal and more interesting choice NOT to have units, in this way the game becomes more tactical. I would greatly enjoy continuing this chat but in a more private situation.

Thanks for all your help, this topic is going in my portfolio for future references.

Ian.

I wouldn't bother with a spreadsheet, or even bother with spending tons of time before release trying to balance everything. With RTSs, it's all about how units work together, so the only way to balance is to actually play the game and see if one side has any unfair advantages.