The essence of RTS?

These days the trend in FPS is to add more "complexity" with character classes imported from MMOs: tank, sniper, support, class-specific abilities, etc, etc. You see it Firefall, TF2, PS2, the upcoming Overwatch to name a few.

Still, there is one game I consider to be The One True FPS - Cube2: Sauerbraten. It's the exact opposite of modern FPSs: everyone has the same character and the same weapon and it's capture the flag all the way. Or rather, "Instagib CTF" is the most popular mode. Cube2 has survived 7 years without any update - not even a single new map - and has existed for even longer. It has also out-lived numerous variations based on its own engine, that tried to introduce "modern" features. By "survived" I mean that one can find a bunch of players 24/7. When you do the math that's a quite respectable active playerbase.

What's striking in this game is the simplicity of its mechanics: a rifle with a scope and some recoil that can be used either to jump higher or to go faster (by firing "backwards"), and... that's it. All the complexities of "modern" FPSs, the plethora of guns, stealth, armor, upgrades, tankiness versus DPS... Cube2 shows that maybe that's just distractions after all. Only good at having people complaining about "balance" and arguing which thing is better than another things in forums.

This makes me wonder: what would be the equivalent for RTS games? What is the minimal set of features for a simple yet enjoyable RTS? What is the essence of RTS?

Because the fundamentals of FPS skill are aiming, movement, and positioning, it is quite easy to restrict weapons and other mechanics and end up with a deep game. The weapons are somewhat separate from the core system mechanics, especially in arena shooters, where aiming is simply a matter of lining up crosshairs (except if grenades and rockets are involved), and the other two are mostly independent of weapon choice.

On the other hand, RTS games mix that a bit more. Arguably the fundamentals of RTS are scouting, unit positioning, economy/territory control, and army composition. However, two of those depend greatly on the units available to be used, so that involves determining a fundamental set of units. They also depend on whether the system uses a strict RPS set of interactions, or sets such interactions up to emerge from other mechanics.

If you had an RTS with only one unit (Multiwinia, for example), you'd end up with unit positioning mattering a great deal, but now army composition is irrelevant, and scouting is only relevant if Fog of War doesn't exist. I suppose that means army composition and scouting isn't fundamental, but in any RTS with multiple unit types knowing what composition would work best in a given situation is key. However, even in a game like Multiwinia, other game modes are added in to make sure there is something to do, given that unit positioning and sometimes territory control are the only skills that matter.

My point is that the number of unit types available (one or many) is more fundamental to RTS than the number of weapon types available to FPS. This means simplification still ends up with either a fairly complex game with a few fundamental skill tests, or you ditch all but one unit type and no real resource management and end up with a game that has only one skill test: unit positioning.

There are many RTS games especially on tablets that experiment with distilling the genre into simple parts. I mean, the obvious subject is the Tower Defense genre that focuses on defence layout, but there are others - see Galcon and Eufloria that strip away individual unit command and focus on where you put your resources.

Another good hyper-simple RTS is Swords and Soldiers - it focuses purely on spell-casting and unit selection. Your units cannot receive orders - you just build them and they march forward to the the enemy's base. Then you use spells to help them. There's only one building, and its function varies by faction. It's wonderfully simple and it works shockingly well, despite the fact that it stripped away 80% of the RTS genre.

Or Rymdkapsel, which is an economy-oriented derivative of the Tower Defense genre - it's something in between a God-game and a Tower Defense game.

Fun distillations of the RTS concept to fit into the tablet.

Yeah, but that's a core principle necessarily buried under the types of units available and the range of mechanics those units have to operate in. It's a good core principle, but it also demonstrates that there is far less separation of the layers of an RTS than an FPS.

quote: Direct Unit Control, Direct Base Construction, Direct Army Construction

Und dann: dungeon keeper.

My take on what defines RTS is "being a Real Time Strategy"

Let's go over those in reverse order.

Strategy means your means for winning is outplanning your enemy. It is all about planning. Even perfect-information turn-based no-economy games can be strategies, like chess. The important things seem to be that there is no "score", only winning, which you are left to achieve by outplanning your opponent in any way you deem optimal. The victory is final and there is no respawn; victory redeems everything.

"Time" adds a maxim that "time is the only resource". Typically this means that you can "grow" in such games, and other "resources" exist. But only one of these resources is truly, linearly, limited: time itself. Non-realtime games can satisfy "time strategy" conditions, e.g. Battle for Wesnoth does involve tradeoffs where you choose non-combat units early on to capture more villages, and thus yield you more actual combat units in the near future, allowing you to trade your time resource for more ingame asset resources on a certain planning horizon more efficiently, thus maximizing your win chances.

"Real" adds that the time resource expended on moving ingame assets is also the same time expended on actual planning and transmitting orders, and that pieces move in parallel.

Surprisingly enough, if you squint, this is a pretty robust definition. For example, things like DoW2 and DOTA don't qualify the first two criteria either hardly or softly, and even the scary part of shooter games played strategically (as setting up proper patrol routes in Quake to minimize probability of enemy gaining powerups) or shooter/strategy hybrids are put into their proper places.

A possible flaw is that this says nothing about the number of pieces you control. Theoretically you could devise a "real time strategy" along these lines where you only had one strategically growing, capturing, expanding, planning and countering "commander unit" directed in FPS fashion. But then, is controlling a battleship a strategic task?

Dota softly doesn't qualify on "strategy" because you have respawns. Softly because i'm not sure about how respawns are score and not asset generation for achieving victory.

Number of pieces under control (or of uncontrolled pieces on board) is precisely one thing the above description avoids.

You can play ZK dota-style but this is typically a rush, a process in which you choose to optimize spending your time on a very low planning horizon. This is so with all rushes.

The thing on which DOTA doesn't fit most, i think, is the "time strategy" part. Sure enough you can grow and invest time differently but i'm not sure there are tradeoffs similar to the comm rush tradeoff described above. You either farm efficiently or you do not farm efficiently. But here's a huge disclaimer of me not knowing much about DOTA game theory.

quote: Dota softly doesn't qualify on "strategy" because you have respawns. Softly because i'm not sure about how respawns are score and not asset generation for achieving victory.

I disagree. Respawns are just a game mechanic. I think what you really want in your definition of 'strategy' is that the current options and gamestate need to be significantly affected by what happened in the past.

Quake (for most people) would then not be a strategy game. You spawn, run around collecting weapons and kill each other largely independent of your assets 5 minutes ago. It would not be unreasonable for high level powerup patrol Quake to fit my definition but I do not know enough about it to say for sure. It sounds like the past (except for score) would be forgotten in a sufficiently long Quake game.

Fighting games would not fit this definition of strategy. Your options in a fighting game do not change over a long timeframe (exccept when your health is too low to do certain things). There are many decisions in these games and in short timescales your choices are constrained. The distinction is that your options are not likely to have much correlation with the game from 10 seconds ago.

In effect I have taken a while to say that 'strategy' is long term planning. For long term planning to be useful the game has to be structured such that actions will have effects over large timescales (relative to the game pace).

The distinction between "score" and "only winning" seems too narrow and hard to generalize. This seems like a bit of the definition which could be removed because the change in options already covers it. In may games score just happens to be something which you need to gain to win but which has little effect on your future options.

If you write enough about 'strategy' you will eventually try to define 'tactics'. To preserve symmetry I would say tactical decisions are those that have their particular details forgotten in future gamestates. Many things that look like tactical decisions could actually be optimal actions and micromanagement.

Anarchid your definition of time is interesting. It seems like every game is a 'time game' (whatever that means) and it would be good to see a counterexample or at least a contrived toy example. Perhaps I am overgeneralizing time and your definition of time really depends of fluff.

My definition of strategy here still includes many games which people would probably not want to call RTS. At this point we should remember that names like this are rarely accurate. For example many games (Quake?) could be called a Multiplayer Online Battle Arena but when people say MOBA they mean Dota-like.

RTS tends to actually mean:

• You have many peons that you can order around.
• You tell these peons to collect resources in a generally exponential-return way.
• You make a base or some other static structure and take territory which contains the resources.
• You can tell your peons to destroy the other players stuff. Fights are generally slippery-slope.

Kernal Panic sounds like a decent contender but the mines and artillery complicate it a bit. How about Liquid War?

what abou sudden strike? no base options etc (dunno about ressources)

Yes. There is no pressure to optimize spending time on achieving some max value when you just oneshot targets of opportunity which cannot grow.

There is no value to maximize.

Soo the revised definition list would be

"Strategy" - The game state is significantly affected by events which occured in the past
"Time" - the game features a significant amount of problems which involve maximizing value per game's time unit
"Real" - the time unit used by the game is the same as used for planning and executing, and pieces move in parallel

quote: Now the question is if you think MOBAs should be included into that definition or whether you need to refine it.

I'd settle for a sliding scale by declaring that the outcome is a float, not a bool.

However, when people say "RTS" they usually mean something that is not MOBA so i think the thing about "growing" is usually exponential (to some form) in an RTS. So a more robust definition would include that.

Yet again, not all forms and not all exponents would suffice to define it. e.g. a dota hero which grows exponentially in power does that exponential growing, but doesn't feel RTS-ey.

Maybe "number of controlled pieces on board" is just a mandatory thing after all. Well, or just go the feel/expectation route, but that feels like giving up.