Personally, I generally only play the custom maps in Starcraft 2, and try to stay away from anything with micromanagement.

Micro is the bane of my existence. I can do it, but I don't WANT to. My focus in RTS design has been into automating basic behaviors, allow for the assignment of hierarchies that work together smartly, more focus on economics and more static defenses and other such goals.

I think it might be important to differentiate between exploratory strategy and competitive strategy. Exploratory games want you to master lateral thought and mess around with combinations. Competitive games are the much maligned yet much loved clickfest style.

I vastly prefer exploratory strategy. Rather than dozens of strategies there are thousands and they are all a chaotic myriad of unbalance and mathcrafting. And even when you think you have a totally broken strategy someone else puts together a crazy out there combo of units and smashes you like its nothing.

I am currently working on a game that demo's the freedom of my GAE derived exploratory strategy engine while also serving as a test bed of new ideas. It has a dozen factions currently being finished, mostly just XML to describe their abilities and stats and I am planning to add at least a dozen more. They are small groupings of 1-5 buildings, 1-7 units, 1-10 upgrades, and 1-10 items with some other varied mechanics for customization as well. That way I can play with what is effectively 25 or more factions and see how they interact together.

Do I like to use farseers and ritualists to locate enemy bases and armies and crush them from afar? Do I use farseers and scouts and spam units to send a realistically sized army of spam to fake them out while backdooring with a small elite force? What about spam to take the heat while large numbers of cheap AOE fire damage floods in behind the lines?

How does having varied damage types even work in an RTS anyways? Most games just have attack and armor or maybe physical and magic damage. Can I change that? Should I? Can I mix military and economic functions? What if there is a woody faction that provides some ranged damage with debuffs and sells potions with useful bonuses on the side that you can't get anywhere else? Sure it's weaker than the other ranged faction but can the availability of poison and healing potions offset that?

Could it combine its poisons with a tinkerer faction to create better traps? Should I make cheap single target traps to cripple the enemies scouting or should I make expensive AOE traps like log or stone drops or spiked pits? Can traps "catch" monsters for my animal related faction to give it more variety than its normal spread of creatures can provide? If my enemy has the potion faction should I rush to prevent health potion supremacy or are they buffing their trap makers with damage types that don't scale to the late game but hit hard early?

I know a lot of people like Starcraft where you see emphasis on kind of unit and know exactly what to counter with, but I prefer a little more imperfect information in my fights.

In regards to creating different 'races' that have unique play styles, I think it would be a good idea to start with deciding on different styles of play that are interesting and make sense for the game and not worry too much about their specifics and how they balance each other in the beginning. I think this will give you more variation in play styles and avoid a template race that is then just modified for each additional race.

That approach would require a few things:

1. Most, if not all of the gameplay/unit data to be scripted/easily editable. Which, for an rts, is already a must imo.

2. Agile development or something similar that allows for play-testing/significant changes during most of the development cycle.

3. A LOT of play-testing. Preferably by a small group of experienced rts players that can evolve/document strategies as the game progresses(No reason these can't be developers, but time might become an issue)

For a non-competitive rts, focusing on the macro strategy level and not the math/direct counters is your best bet.

ie-Just because Race A has a unit that does a lot of damage, doesn't mean Race B needs one(copy) or even a heavy armor unit that can take the hits(counter).

Sometimes strategies just emerge from gameplay that are invisible when looking at the data values/code. That's why I stress creating unique (grounded) ideas, rapid prototype them, and see how they play together.

Its not that some strategies are hard to see in code. Its that no one person can see as many strategies as the thousands of hardcore RTS players who will play the game.

One person vs thousands.

Its not that some strategies are hard to see in code. Its that no one person can see as many strategies as the thousands of hardcore RTS players who will play the game.

One person vs thousands.

Although I agree that one person trying to define strategies is nothing against a horde of rts players constantly experimenting, that wasn't what I was trying to say.

By cultivating a small group of players during development that are play testing and discussing strategies for your unique races as they evolve, you can find balance in macro strategies that wouldn't be evident in unit vs unit comparisons. Hope that clears up my point a little.

[quote name='voidedLine' timestamp='1356735026' post='5015216']
For a non-competitive rts, focusing on the macro strategy level and not the math/direct counters is your best bet.

ie-Just because Race A has a unit that does a lot of damage, doesn't mean Race B needs one(copy) or even a heavy armor unit that can take the hits(counter).
[/quote]

I like that, no rock, paper, scissor as mentioned before. Look at MMORPG's most of them also have that triangle and it get's boring.

On the other side, many beat'm up games also have vastly different characters and not all are balanced. The developers try to make it still fun by putting a bigger selection in the game.

If there is one character that is a bit stronger in a selection of 3 available the developer has to do something about it (or everybody will use the one character).

However if you have one or two unbalaced characters in 20 characters, the chances that the players will find the proper tactic against thoose two are much higher. The developers don't have to explicitly do something about them, they can pretty much rely on the creativity of the players. Of course the players do need to have the means (in this case many different characters) to do this.

I think, while a lot of people think of RTS foremost as strategy games, they usually rather tend to be evolution games, where all player start the same (or similar) and as the game progresses, you research/customize/unlock new units, building etc.

This makes an RTS replayable to me. in C&C you always started with a few soldiers and build it up to tanks etc. and C&C 2, 3 etc. extended those possibilities.

what I miss in RTS is to have really strategy, at best you have some tactics, but I'm not a game design expert to pull out ideas for how to let the player make great strategy.

[quote name='Krypt0n' timestamp='1356874133' post='5015743']
what I miss in RTS is to have really strategy, at best you have some tactics, but I'm not a game design expert to pull out ideas for how to let the player make great strategy.
[/quote]

Hard to say, but I would have thought that there is a bit strategy involved. Of course there aren't many and often they don't influence the game as much as just being faster than everybody else.

But if you think about it, sometimes in Starcraft 2 and many other games one player tries to kill some of the enemy workers early in the game.

Now it's clear the tactic employed is hit and run. But almost all the time the player is having a strategy in mind. In this case, weaken enemy economy early on to weaken his army in the mid-game.

Sure this is a very simplistic nevertheless efficent strategy. The problem is that most of the time games don't allow for anything more elaborate.

I'm not even sure myself if it's possible to ask the player for a deeper strategy in this genre. When you take a look at round based strategy it seems much more deeper. However imagine doing the same in real time.

Strategy tends to be a matter of scale. Which is problematic when players refuse to spend more than an hour per game. The economies are so simplified, and the military too, that it's really totally impossible to use strategy.

You can say that your strategy is to expand quickly and suck up the resources you can, but if there is 1 natural expansion for each player and only 2 other expansions, that seems kinda pretentious.

Starcraft games tend to have 1 or 2 raids and maybe some probe harassment and then one or two large battles. How do you have long enough to establish a strategy.

A lot of this deals with the environment too. The Russian Strategy was to move back and back and back stretching out German supply lines and exposing them to winter. Starcraft style games don't even have food or supply lines, much less winter.

Strategy takes root in complexity. Dominions 3 has strategies. You can spend the game finding magic sites to farm gems for summons and spells. You can expand endlessly leaving no defenses except in an our ring on your border lands. You can expand using your pretender and/or blessed troops or by purchasing local troops. Do you focus on mages or the much cheaper ordinary units? Some people are really into Super Combatants and some use multiple thugs in place of one SC. Other players use astral duels to assassinate enemy mages and some people use seduction units to steal commanders. There is a stealthy scout system too where scouting is actually more than running a guy into an enemy base for a brief glimpse.

Most real time strategy games don't have the depth for strategy intentionally, not that they couldn't but because the audience has shifted away from the kind of people with the patience and desire to play deep strategy games.

[quote name='aattss' timestamp='1356717762' post='5015118']
There should be completely different gameplay if you decide to have multiple races.
[/quote]

- I like this idea. Like not just different units with different attack/defense styles, but a completely different approach to game play. Perhaps different a robotic race, where you can micromanage a lot, but they aren't very creative, I.e. they don't fill in gaps well, so there are certain areas of defense they are weak in. Then Humans. You can perhaps order squads around, but what they do is up to how much you've trained the quad. I.e. A weakly trained group might have poor aim, and shoot at individuals, where a better trained group might have groups of 4 in the squad each shoot at individual targets to bring them down faster.

With the Robots, you simply make the upgrade, and all robots are instantly upgraded. if defense towers go out, all your robots take on only basic defense and start returning to base. If the humans lose satelite views, the squads will finish their missions and then return.

Not that I'm going with Humans Vs Robots, but I do like the idea of different perspectives of game play and character management. There is already something like that in Starcraft, as the Protoss Carriers have lots of little drones, that you don't control, you just control the carrier's location.