I'm beginning to play with the idea of writing a non-technical concept design document for a game I've been tossing around in my head for some time. However, I want to see what a few other MMORTS players feel about the following topics:

"Grinding" -
Not endlessly, per say, but perhaps the more units you kill or objectives you achieve in a game, the more kinds of research becomes available, areas become unlocked, etc. I realize this stirs up some pretty heated opinions, but I was curious as to what they were.

Teams -
I recently re-activated my Planetside subscription, and realized that the basic foundation of the game has potential as an MMORTS. One of the main complaints of a MMORTS is that new players can't compete with veterans, and get gobbled up as soon as they enter the game. What if newbies began fighting for a side in a conflict, capturing static objectives and engaging enemies from a number of other sides as a unified force while they progressed through the game in terms of research, infrastructure, and experience?

Later on, they could break off from these governments to form their own associations in a frontier area, much like EVE's 0.0 space. I'm not sure whether or not it would be a good idea to let these veteran players raid the newbie factions or just let them fight their own wars, though. There would have to be incentives for older players to renounce their allegiance to their previous faction. Maybe a tax on income that they would shake off?

---

You see this concept with many of the more political and strategic MMORPGs. Shadowbane is very similar, where newer players progress as characters initially, and then proceed to PvP once they've learned the game and amassed some assets. Of course, Shadowbane's grind is very, very boring, but its concept is sound.

Basically, why do MMORTS games have to be a free-for-all from the get-go? Why can't the free-for-all be an endgame feature?

"Grinding" - my friends that play World of Warcraft talk about this. Nobody likes it. To me, someone who has little time for gaming, it says "don't play this." There should be a better choice.

"Teams" - the most successful MMO games build a lot of tools for players to build a sense of community. Chat, guilds, there are others. Players stay in the game long after they would normally leave in order to keep in touch with their "friends" in the game.

"Veterans" the trick is to make newbies unpalatable to veterans, then they'll leave them alone.

In my ideal MMORTS, there is no end-game. Just a series of skirmishes, advances and retreats.

How do you keep the game interesting after a period of time, though? That was my complaint with BoundlessPlanet. After awhile it was just the same units, the same two resources, the same structures, everything.

Also, how do you make newbies unpalatable to veterans? They'll always take up space and resources, and they'll always exist as some large player's neighbor and potential threat. The only way around this is to make the game less territory-based and more experience-based, where the concern is less about how much land you control but how many objectives you complete. As such, then newer players could be worth less for a veteran to engage than another veteran.

Grinding doesn't make the game interesting. New content makes the game interesting. Your friends make the game interesting.

As for veterans vs newbies. Make veterans tech use a different resource than newbies that isn't near where newbies spawn. Make a police force that stomps someone who attacks someone too many levels below themselves. There are lots of ways to make newbies unpalatable. The trick is that for every rule you make, you have to think about how someone is going to exploit that rule. And someone will find a way.

I'm not the biggest fan of invincible AI and protective bubbles for newer players. Systems like that seem a little too artificial for me. I like realistic methods of player control, like mortal AI-guardians that you can still kill, but would take a lot of fighting to do so. The resource idea doesn't ring true with me, though. As soon as the newer players out-grow their initial tech-1 resources, that places them at the far bottom of an immense and threatening food chain when they try to branch out into tech-2 resources. It would be a race to see who could monopolize as much as the tech-2 resource as possible to exclude newer players from ever breaking into that tech tree. How would the newbie's tech-1 units even compete against the units the veteran with the tech-2 resources can produce?

As for grinding, I agree with you that it doesn't make the game inherently fun, but it does add a sense of accomplishment to the game, especially when cooperating with friends. That's why so many MMOs become addictive. It comes down to an issue of "Just one more baddie before I log... Wait, let me just level, then I'll go to sleep... I'll level, and finish up that quest real quick, then I'll go to sleep..."

With current MMORTS'es, there's less of what I call "exponential potential," where the more baddies you bash, the better you get at bashing baddies, and the better baddies you can bash. Sure, the more land you conquer, the more resources you have to field troops with, but that growth levels out quite quickly. It's what bored me with Boundless Planet. After a certain point, there was no profound difference between having 20 or 50 CCs.

All of that is partially intentional and beneficial, though, to help curb the newbie vs. veteran issue. You don't want your veterans becoming so powerful that nobody stands a chance. At the same time though, players reach a 'maturity' level, where they can stand up to most other opponents effectively, quite quickly. What is there to keep these 'matured' players in the game when they're pretty much equal to everyone else? Aside from community, of course. I'm talking about game mechanics.

I was thinking that the tech-2 resources would be in a different location physically removed from newbies. Like a one way "star gate" to borrow a phrase that lets them go to different planets.

I think the key to making an MMORTS interesting is content. PvP is good, but an entire game it is not. At least in my opinion. There needs to be stuff to do, goals to reach, quests to complete. Tech trees to climb. AI's to kill.

Give the higher level players more challenging goals and access to new types of missions. There will be some that still bash newbies, but every game has its griefers.

Completely agree with you about content. So far, though, Ballerium has been the only game I've seen to give the players AI to fight against. It's a very good thing. I don't know if you've played enough of it, but the game has 'hives' that produce wandering, menacing NPC enemies that attack anything they can see. There are different types with different levels of strength. You can either camp the hives to see what they spawn and kill baddies once they do to gain EXP for your troops, or destroy the hive itself to gain 'karma' (a player-specific stat that controls how many units you can control, among other things) and magic items for your troops. Killing the strongest hive is quite difficult to do solo for a newer player. It took me some army development and quite a bit of micro-managing strategy to pull it off without losing more than a unit or two at worst.

Anyway, what I like about that AI system is that it doesn't try to replicate another player. MMO's that use AI to make it seem like it's just another player bother me. AI will never replicate the reflexes and versatility of another real player within the conceivable future of MMO gaming, so it should be treated very differently if included. If we're talking about space (which is my favorite venue for a MMORTS - I'd be drooling if I heard news of a Homeworld-esque MMORTS being developed), then AI should be reserved for random alien strikes or somesuch. You shouldn't have to wonder if that other nation you just stumbled upon is an AI or not, because if it's an AI, you'll just develop strategies to defeat it and end up being disappointed. In Boundless Planet, with the right timing and hit-and-run attacks, I managed to eliminate several well-defended bases from an offline player by distracting specific AI units and long-ranged artillery. Units would never get close enough to hit me while I took out power plants and oil rigs. Then once they were dead, everything ran out of fuel and died. It was boring, since it was the same strategy every time.

I've never seen quests in an MMORTS so far, but the idea of 'missions' given to you by the empire you belong to (think EVE or Planetside) to conquer X and Y territories would fit the bill. Making a MMORTS without a focus on PvP is a little counter-productive though. To balance that out, I've suggested cooperative PvP, once again like Planetside or Anarchy Online, where you work initially with your faction and other players that choose that team out of three or four or so when they wish to join the game.

With goals and tech trees and everything though, you constantly reach the same problem. The longer the tech-tree, the more of a disadvantage newer players have against older ones. Isolating newer and older players does work, but then it starts to feel like you're just grinding initially to get to an endgame.

Apparently, I really like discussing game theory. :D

The most homeworld-esque MMORTS game is Time of Defiance. Go check out our video review to see what its like.

Like you, I don't think you should be able to confuse an AI for another player, however at the same time, it is a good way to give the player something to do at their own level. I played the command and conquer AI a lot over the years, and it was always able to give me a run for my money (until Generals came out and sucked wind.)

I always enjoyed cooperative play against the AI with friends when we had LAN parties (I actually supplied the lan and all the computers, but I used to have too many PC's laying around.)

MMORTS games are so new, they are still trying to find out the rules for the genre. I think they have a long way to go towards developing community - which is the key I think to long-term success.

The folks in World of Warcraft have managed to figure out how to give advanced players stuff to do without spending all their time killing newbies. It should be possible for a MMORTS.

I played Time of Defiance and enjoyed it, I may actually pick it back up again sometime soon. (I really liked the review, by the way. I saw it after I had played the game, but it made me want to give TOD another shot.)

I had a hard time getting used to the slow pace of the game. I wasn't sure if it was a good thing or a bad thing. It kept me alive overnight, but waiting half an hour before a colony ship arrived at a new island to take over got old eventually. I just don't have enough things to do to pass the time in small half-hour blocks, and I don't want to leave units sitting there not doing anything. Also, part of what made Homeworld great was that it was the first truly 3D RTS. Not 3D in terms of graphics, but 3D in terms of strategy. You can't attack from the top or bottom of an enemy force in too many other games. Homeworld also has a bit of a different feel to it than TOD, but it's a little abstract and hard to explain at 4AM. Maybe I'll see if I can put it into words in awhile.

The MMORTS is indeed a very young genre. It'll learn to stand on its own though. We're already seeing some very influential archetypes that will hopefully be built upon. Boundless Planet, I believe, is the simplest and purest example of the MMORTS as a genre, with Time of Defiance along those times too. Not necessarily the best, but I feel it has the most core aspects a MMORTS in the current state of the genre should have. By comparison, Shattered Galaxy (and Dreamlords from what I've heard - if you've played it) really breaks down into a regional game-matching service. Limits on how many individuals can fight eachother in the same combat environment are not consistent with the MMO aspect of an MMORTS (not including limits for the sake of lag, though). Mankind is in a world of its own. It has a whole economy aspect that other games don't even approach. Of course, I have my biases, but I'd personally like to see MMORTS games go more in the Mankind/Boundless Planet/Time Of Defiance direction rather than the Shattered Galaxy/Project Visitor/Dreamlords direction.

The difference between AI in C&C (which one was it, Red Alert 2? Tiberian Sun?) and the AI in MMORTS'es is processing power. It's fine to play against a handful of AI opponents when it's just your computer, but when you need a server to process a ton of user clients and then load it up with a whole bunch of AI entities to handle too, it becomes either very slow or very expensive. That's what holds AI back in MMO games.

The AI I liked best was Red Alert 2. I didn't like Tiberian Sun, mainly I think due to glitches like missions not ending because you haven't destroyed one tunnelling tank. I didn't finish that one even though I thought the story was cool.

I still have the Homeworld expansion pack to finish, plus Homeworld 2. I'm working on Age of Empires III right now. I don't get as much time to play as I'd like and there's always a new game to play.

If designed for it, there's no reason that AI players have to be on the same hardware as game simulation. In the MMO genre, often the games are built from a number of tiny services that each run on their own (connection, login, chat, gameplay, ai, database, replication, patching, etc.) When the MMO has a small user-base it can run on a single PC, but when it grows, the server farm can grow as well with some services running on multiple PC's at a time, all talking to each other to make sure players see a consistent world.

Some articles I've read talked about harnessing the extra power on your players' PC to run AI's.

I'm personally a fan of the one-world MMO paradigm. Like Eve Online has. It might actually end up being a world for each continent due to network latency, but I don't like the limited number of slots per server thing.

Currently EVE Online is the only mainstream MMO to have only one non-test shard, to the best of my knowledge. Anarchy Online has only ever had two, as far as I know, so AO comes in second in that respect. The way EVE does it is that different regions or clusters of regions, or even individual systems depending on population density, have dedicated servers for their handling. Other MMOs can't do this so easily, because they emphasize the 'seamless world' without loading different zones. EVE is anything but seamless. You travel from system to system through jumpgates, and there is always a noticable loading time between jumps. Much easier to divide zoned regions between servers than it is to divide seamless tracts of land if you want it all to be processed as one world.

And yes, AI can be run on separate servers, but assuming you have even one AI for every five or six players, that's still a lot of AI decisions to calculate. Games nowadays are designed to save processing power for the sake of saving money, which is why AI is limited if present at all. AI on the scale of an MMO is very taxing, so good AI would take an extra server or so in the farm to operate. I remember reading that Ballerium has three servers devoted to AI. That could get quite costly, and their AI leans more towards basic AI than complex.

The problem with distributing processing back onto the client is that it exposes the game to exploitation. You always have to assume that anything within the client computer's memory is in the client player's control. The less you give them to process, the less power they have over the game.

Granted, making a seamless world isn't easy, but there are a few articles on how to do it. One of the Charles River books on MMO games has one.

The C&C games I used to play had 6 AI players simmed on moderate PC's for the time (400MHz Celeron) plus doing rendering etc for the actual game. I would think a decent rack mount server (quad 3Ghz P4) nowadays could run many more than that.

As for farming processing out to the clients, one of the MMO designs I looked at did that for EVERYTHING. It had the virtue of being infinitely scaleable because of its distributed design. Today, with big datacentres running MMO games, eventually their scalability is limited by bandwidth. Its design had every calculation tripley redundant with hosts comparing the results from each other and a central system allocating tasks to widely dispersed players and a lot of encryption so it would be virtually impossible for someone to tamper with the game and actually effect other players. The downside of course was that it was pretty complicated. And that says a lot when you consider how complicated a regular MMO is.

That Ballerium number is interesting, because I read they have a total of 4 servers, which means that 75% of their server farm goes to AI. Yikes.

Games nowadays are designed to save processing power for the sake of saving money, which is why AI is limited if present at all.

Do you mean saving development time? Or electricity and server costs?

The problem with that checking and re-checking is that it curbs some of your efficiency since you need to devote processing power to do that checking. Still, I guess the end result is better than the sacrifice. Then it comes down to development time, though. We aren't lucky enough yet to have teams the size of those developing games like Vanguard and Warhammer Online working on MMORTS titles. If only, right? Someday.

By processing power I mean electricity and server costs. It isn't just about the server box and the juice powering it though. More servers mean you need more people managing and looking over them, which is even more expensive. We just don't see many MMORTS games with that kind of budget yet. The fact that one person can homebrew a MMORTS as a hobby that competes with all of the other ones out there shows that the MMORTS genre is still a bit of a 'do-it-yourself' frontier in gaming.

Also, the Ballerium server number I have is just from reading the page occasionally. I know they added two AI servers recently. I don't know how many those servers were added to, though.

Comparing the results from 3 servers can be reasonably efficient. CRC checks of some of the server state data sent periodically would not require much CPU and should be adequate to detect tampering. While CPU is a limited resource, it is free CPU time. The development time is the real killer for that idea.

The 2 guys working on Boundless Planet took on a lot to make an MMORTS, they have been working on it for 3 years.

It will be interesting to see what the guys at Ensemble come up with. They're RTS masters used to big budgets and teams.

Yeah, that's true. Development time is very precious. Fans like us are very impatient. 8)

Boundless Planet is very impressive, considering. I treat it more as a 'proof of concept' piece to show the giants out there that an MMORTS is more than viable in terms of production and execution.

I'm excited for Ensemble's project, though admittedly I know very little about it. Are there any other details aside from "Ensemble is making an MMORTS!" out there?

The only info I've seen on Ensemble's project is what I've posted here. They're 'noodling' on the concept.