I've been a fan of RTS games pretty much since I first figured out what a computer was. Chips challenge, Red Alert1/2, Age of Empire/Mythology, Warcraft1/2/3, Starcraft1/2, Supreme Commander, Tiberium Sun 1/2/3/4/5, Populous the Beginning, Dungeon Keeper... Great times all. But in the past decade or so I've been getting really into MMO's and more focused RPG's as well and got to wondering, what about an MMO RTS?

Unfortunately, when searching for the Genre, all I've found are browser and Mobile device games. Some of them are nice, like Tribal Wars, others are popular, like Clash of Clans, but most of them are cringe worthy and well, none of them are really RTS, nor are they MMO so much as Online Matchmaking systems.

Now, I've heard good things about EVE Online, Elite Dangerous and even Elder Scrolls Online's Siege Encounters expansion, where you can build empires of sorts with your guilds, but it's still a little too central avatar focused for what I'm trying for.

I guess what I'm asking is what would it take to make an expansive sandbox game world the size of one of WoW's continents or larger where the player could very literally build their own bases and go to war against other nearby settlements or bases with around 100 or so people participating at all times on the server?

Would that be possible? What Engine should I be learning more about for this? Would a significant number of people even be interested? I certainly am... I have a number of Ideas on how to organize the structure of the game, but I want to hear if anyone has any glaring arguments against the whole scheme.

Go look at Ark: Survival Evolved it basically does everything you are asking for and it even has dinosaurs. But in really short terms to answer your general question.

Is it possible? Yes.

You might also look at Rust, Dayz as they also fit your your criteria. One has penises and the other has zombies!

As to engine choice -- better to ask that in the technical section, but to be blunt do more research before you jump in headfirst at the shallow end of what is a very large pool.

Not all of the MMORTS games on the Wikipedia list of MMORTS games are browser-based or mobile games.

https://en.wikipedia.org/wiki/List_of_massively_multiplayer_online_real-time_strategy_games

And that's not a complete list. I vaguely remember games like this coming out a few per year for around the last ten years. I didn't notice Kingdom Wars on the list:

http://store.steampowered.com/app/227180/

If you cannot think of any games that fit your criteria, that just shows how difficult it will be to make a successful game in this genre. People have had this idea basically ever since RTS games and MMO games existed.

Let's say you have 1000 online players and each has 200 workers and they will have to pathfind every second. That is 200,000 paths per second. If there are 300,000 tiles that is 36+ kB of passability data. That is 547x547 tiles. 2 GHz/ 200,000 paths = <10,000 instructions per path. The lag might be unbearable. And if each path is 100 tiles that is <100 instructions per tile. I was thinking of using local avoidance instead of pathfinding and maze navigation (follow the left/right wall). Really I think if you look at the trend with MMOFPS it tends to lose all sense and becomes a giant sprawl and mass run around for miles until you get capped out of nowhere unexpectedly and have to spend another 5 minutes getting to the "front line". With MMORTS the game might be vastly unpopulated, dominated by a single clique, or lacking the finer details that make RTS fun.

I've been a fan of RTS games pretty much since I first figured out what a computer was. Chips challenge, Red Alert1/2, Age of Empire/Mythology, Warcraft1/2/3, Starcraft1/2, Supreme Commander, Tiberium Sun 1/2/3/4/5, Populous the Beginning, Dungeon Keeper... Great times all. But in the past decade or so I've been getting really into MMO's and more focused RPG's as well and got to wondering, what about an MMO RTS?

Unfortunately, when searching for the Genre, all I've found are browser and Mobile device games. Some of them are nice, like Tribal Wars, others are popular, like Clash of Clans, but most of them are cringe worthy and well, none of them are really RTS, nor are they MMO so much as Online Matchmaking systems.

Now, I've heard good things about EVE Online, Elite Dangerous and even Elder Scrolls Online's Siege Encounters expansion, where you can build empires of sorts with your guilds, but it's still a little too central avatar focused for what I'm trying for.

I guess what I'm asking is what would it take to make an expansive sandbox game world the size of one of WoW's continents or larger where the player could very literally build their own bases and go to war against other nearby settlements or bases with around 100 or so people participating at all times on the server?

Would that be possible? What Engine should I be learning more about for this? Would a significant number of people even be interested? I certainly am... I have a number of Ideas on how to organize the structure of the game, but I want to hear if anyone has any glaring arguments against the whole scheme.

If we are talking about "technically possible" ... Yes, with caveats.

- Some engines are supporting terrain streaming out of the box (Esenthel does, Unreal Engine 4 AFAIK)... others will need extentions to be able to do so (I think there are third party assets for Unity). That means you might need to do some legwork if you happen to choose the wrong engine.

- Even if the engine supports streaming... if you use huge pregenerated maps, that map might take a HUGE amount of space on the disk... I had one case where a single map would take 20 GB of disk space (map was huge). Not a problem for memory (its streamed to memory), but if you have 5 maps of this size, you are asking the player to give up 100 GB of disk for just your game.

- Procedural generation can ease some of this pain (apart from what I will talk about below).... but if you create maps or parts of it on the fly, you are now asking the players PC to churn extra things. Thus you need to be very careful to do this at level startup (which means you need to write it to disk somewhere), and not churn while the CPU is fully busy keeping the GPU busy, running AI and all the other CPU bound subsystems.

If we are talking about "can I do it, and how long will it take?", question is "what is your goal?"...

- If its just about huge open maps with no or procedurally placed features... sure, no problem other than the time and technical challenge to develop and test the procedural generation algorithms and test them, and all the optimization needed to stream that huge world in and out of memory, and all the systems needed to make sure you only process AI close to the player at full rate (there is plenty of design theory on how to handle that), and same with graphics (depending on the engine there is more or less legwork involved).

Look at ashes of the singularity, besides the DX12 buzz (which seems kind superfluous to me besides the technical interest in DX12) they have concentrated on their cutting edge engine to render tons of units and cut a lot of corners when it came to level design. Hence big maps, almost void of details. THAT type of map is achievable without any problem.

- If we are talking "beatyfully crafted game world with hand placed details", question would be do you either have access to several million dollars or a team of 50+ people, or hundreds of years time to build it yourself?

Filling a huge map with handplaced details will take ages... giants like blizzard were able to do that, the result are quite stunning worlds with a lot of well placed details. You will most probably not be able to reproduce that, even the best procedural generation might only come close.

TL;DR: Yes you can do it, but its a technically challenging subject that will take a long time to pull off given more than just basic expectations if done procedurally. Yes, there certainly is a market... but be aware that if others didn't do it before, it might be not because there is no market, but because it is challenging to create.