Hmm, lots of different ideas on what people consider vital to the genre. RNG I completely disagree with, considering it's used all over the place in other genres just fine. (Though RTS games have been simulating hits as far back as Total Annihilation) Same with Minimaps, that's also a red herring, World of Tanks, War Thunder, LoL, DotA. They all have minimaps. Heck, quite a few FPS games do as well. Age of Empires 2 had scale disparities, unless you think Castles were supposed to be about the size of four elephants, unless maybe AoE2 stands for something else? And defenses like minefields are interesting, but only if you think bases are at all vital to the genre...which they aren't.

I think there is plenty of room for innovation, but folks are often scared to go too far. Just look at Grey Goo, which is really a by the numbers RTS, which is not all that surprising since it's from former Westwood guys, who never really knew how to innovate anyway. Try Infested Planet instead, that was a really great single player RTS that doesn't play like other RTS games. The Eugen systems games (Wargame Colon SomeCodeWord) are fairly innovative compared to other games in the genre, though each sequel has been less and less so.

I do think there is plenty of room for a RTS MMO, or a F2P RTS, though most of the previous attempts were not the best stabs at it. I think trying to retrofit it onto a beloved franchise that isn't F2P is a bad way to go, and I think there are definitely better ways to go about it.

1) StarCraft at least lets you build static defences which do pretty much the same job. Eg. the Zerg have sunken colonies/spine crawlers, Protoss have photon cannons, Terrans have missile turrets, bunkers, and mines. There must be other games that allow this.
2) Is this really that relevant to gameplay?
3) Age of Mythology allowed up to 12 players, if I remember correctly.
4) Some RTSs allow saving the game partway through the match and resuming it later. I agree that for slower strategic games this could be nice, but for more fast-paced games leaving for even a few seconds means you're done, so the entire game would have to come to a halt.

1) Realism MIGHT help. Personally, I like a fortress when I see it in medieval RTS's. It would require different code to set up fences. It would have to fit with the story of the game too.
2) It's a source of new ideas. You can blow your mind considering realistic scales. We have only modern times to work with though. Looking at Google Earth.
3) Not as fun as FPS, which is the main competitor.
4) AI could take over. Something has to be done to make RTS's more casual drop-in friendly.

EDIT: Still, I don't know if I would play this over StarCraft.

1) StarCraft at least lets you build static defences which do pretty much the same job. Eg. the Zerg have sunken colonies/spine crawlers, Protoss have photon cannons, Terrans have missile turrets, bunkers, and mines. There must be other games that allow this.
2) Is this really that relevant to gameplay?
3) Age of Mythology allowed up to 12 players, if I remember correctly.
4) Some RTSs allow saving the game partway through the match and resuming it later. I agree that for slower strategic games this could be nice, but for more fast-paced games leaving for even a few seconds means you're done, so the entire game would have to come to a halt.

1) Realism MIGHT help. Personally, I like a fortress when I see it in medieval RTS's. It would require different code to set up fences. It would have to fit with the story of the game too.

2) It's a source of new ideas. You can blow your mind considering realistic scales. We have only modern times to work with though. Looking at Google Earth. Doesn't this look better? (View full size) I adhered to a real scale here.

3) Not as fun as FPS, which is the main competitor.

4) AI could take over. Something has to be done to make RTS's more casual drop-in friendly.

1) I'm not sure what your point is. It seems to me that you just like medieval RTSs.

3) Are you addressing my point re. AoM or adding a new one?

4a) Some RTSs do have the AI take over, but what if the AI does something that goes against what the player would have done? Not viable for high-level players, I think.

4b) Casual gamers are not generally the intended audience of an RTS, as far as I know.

When WC3 was announced that it would be in 3D, I was excited because now I thought I could zoom out to see the entire map. No. They didn't do it.

I disliked other things about the game, but this is one thing that Supreme Commander does pretty well -- you can actually zoom out until the entire map is visible, and when you zoom out far enough to make visual identification of units difficult they are abstracted as icons representing the type of unit -- I'd like to see that functionality expanded upon (to more capably handle groups of units and display other information) and included in more games.

(Note that Supreme Commander is the most recent RTS I have played by quite a long shot, so it's entirely possible I've missed this feature in other newer games.)

R.U.S.E. is like that.

[Edit:] Crap, did a Ctrl+F to look for 'RUSE', and ofcourse everyone spells it with the periods so I thought it was unmentioned.

I could be wrong, but part of the problem is probably that there's only really one current Triple-A RTS title with any mindshare these days, and that's Starcraft II. The Age of Empires series has been gone in its real form for years, along with Empire Earth and Rise of Nations. Thus the genre is poorly represented, providing few *major* options that a lot of people know about, and that one primary option feels very different from the more casual style of AoE.

It should also be noted that Ensemble Studios (and thus the AoE series) was eliminated not because of commercial failure, but because it evidently fit Microsoft's strategy (though I don't really get that), and the IP has not seen a major new entry since (Age of Empires Online was very unsuccessful. This is what they get for trying to turn AoE into a weird F2P social game thing)

I used to be a big fan of RTS, and I am one of the people who stopped playing RTS altogether.

(I will use Blizzard's RTS as examples, as they are the ones I have played the most.)

The game that killed my interest with RTS was Warcraft III. I didn't quite like the direction that Blizzard took with WC3. Micromanaging, heroes level system, completely imbalanced heroes, and it was hard to make a comeback. If you lose a battle, you lose the game. Without a comeback, it takes the fun away from RTS. Losing a unit is like losing a finger. It is not fun anymore, it becomes stressful.

Exactly the same thing happened for me as well. The stupid hero-and-creep crap ruined it completely. I haven't even bothered trying to find a new, good RTS ever since.

I haven't seen much in RTS games in the last decade that I didn't see twenty years ago with Dune 2 or Warcraft 1. More units, better graphics, a few RPG elements shoe-horned in here and there, but it just seems like a pretty stale formula. The golden age was the second half of the 90s, in my opinion (StarCraft, C&C Red Alert, Age of Empires, Seven Kingdoms), but after that I lost interest. I really don't like the hectic and fast-paced micromanagement. If they go the free-to-play or MMO route, I'm even less interested.

Outside the traditional gather-resources/build base/attack RTS subgenre, I really enjoyed the second generation of Total War games (Rome and Medieval 2). It's a much more deliberate pace, with the kind of one-more-turn mechanic that made the Civ games so addicting. In the real-time battles, you're only ordering < 20 units, and there aren't many special abilities that you have to manage - its all about picking your terrain and tactics. They're going away from this simple formula, and it in my mind, it's hurt them. Rome 2 could have been a really special game, and they kind of botched it, even if you just look at the gameplay and not the ridiculous stream of DLC that screams that they released the game with only about half of its content finished. Fortunately, the mod scene for Rome 1 and Medieval 2 is incredibly dedicated and very talented, so the games still have a lot of legs.

The other RTS-ish subgenre I've really enjoyed is the Close Combat series, although they are really more squad-level wargames that happen to be real-time. Apparently they are still working on these games, although it's moved from being published by Microsoft to being very much an indie-niche kind of deal. Company of Heroes is the closest thing I've seen lately, but it traded the realism for more traditional RTS elements, and though I tried to like it, it fell a little flat.

Just gotta mention that my favourite RTS of all time is getting rereleased next month : Homeworld.

One thing I loved in homeworld was the ability to set formations and "attitudes" for groups of units.

Shame that Cataclysm won't be included....

I saw also that a few people mentioned unlimited zoom and real time projectiles, which are great features to have in an RTS. (They’re also both present in my game.) Unlimited zoom is not always necessary, and it’s going to be much more difficult if the game has actual terrain to render, but it’s certainly a nice feature to have and it gets me to my broader point; that it makes it easier for the player to see the battlefield and manage his armies, without taking control away from the player. For me, that’s one of the core design philosophies for good RTS gameplay: make it easy for the player to do what he wants to do, as long as the computer isn’t actually making decisions for the player.

One of the problems I see with some RTS that hurts the genre’s appeal to more casual or simply more laid-back players is the artificially inflated APM required to execute basic tasks. The most blatant example of this kind of mechanical obstacle I can think of (apart from just bad pathfinding) was the selection cap in the original Starcraft, where you could only select 12 units at a time. If the player wanted to move more units, you had to select them in separate groups and give the order multiple times. That’s multiple physical actions (selecting and ordering each sub-group) to accomplish one theoretical action that the player actually wants (“move my army here.”) There are some other more common design decisions, such as the way production is handled, that can also contribute to the stressful and frustrating nature of RTS games. Games where the player has to pay for units up front before starting production demand much more attention than the C&C/Grey Goo/SupCom approach, where units can be queued up at no initial cost, and automatically deplete your resources as they are constructed.

Overall, RTS controls have gotten more intuitive since the early days, but I think there are still some ways to further streamline the process without alienating the enthusiast audience. One of my examples is the ability to draw lines on the map and have a group of units move into position all at once. Say you want to form up in a concave arc facing the enemy (a common tactic in RTS games) this can now be done with a single sweep of the mouse, rather than dozens of individual orders. I don’t see this as excessive automation, any more than say an attack-move command is, since the player already knows exactly what he wants to happen, and is inputting all the data points manually. (Video showing how this is useful:

)

Ultimately, I think core RTS game design can be viewed as a zero sum game in terms of how players need to split their attention. I consider there to be three primary activities in RTS gameplay: 1) strategic decision making (what to build, where to attack, etc.) 2) active unit micro (combat tactics, where each player responds to the other’s movements) and 3) passive micro/macro (tasks such as producing units, where the opponent’s actions don’t directly affect what the player is doing.) If you decrease the difficulty or effectiveness of one of these three tasks, the other two will become proportionally more important, so RTS gameplay design boils down to choosing what kind of tasks the player is going to focus on. I find that a mix of mostly 1) and 2) is the most appealing, although not necessarily optimal for an esport level of competitive play, but everyone has their own preferences. There’s quite a lot of other factors that go into making a good RTS of course: the story, the mission design, the uniqueness of the units, and so on, but this is the most concise analysis I can come up with for evaluating the core gameplay of an RTS and defining what makes that activity “fun.”

If any of you guyz know or recall RTS called Dark Reign (1997)

Absolutely loved that game -- unfortunately even the sequel lacked most of the awesome features.

I really liked the fine grained control you could have over your units through the settings for "pursuit range", "damage tolerance", "independence", as well as the more sophisticated behaviours for "scout", "harass", and "search & destroy".

I would love to see another game with similar control over units!