Trending Games | World of Warcraft | Overwatch | Anthem | Entropia

    Facebook Twitter YouTube YouTube.Gaming Discord
Quick Game Jump
Members:3,835,960 Users Online:0

Show Blog

Link to this blogs RSS feed

The Lunch Break Blog

For those of us who would rather be leveling right now.

Author: cmagoun

MMO Combats Are Silly (or Do We Need More A or More I?)

Posted by cmagoun Friday August 8 2008 at 1:14PM
Login or Register to rate this blog post!

(Warning: This is a bit of a rehash of an idea I had a long time ago... still worth mentioning and posting here I think)

In almost every player's MMO life, in every game, there comes a time where combat becomes a rote exercise. Ok... group of three guys, start with the debuffs, then my AoE attack, then pick off the weak guy, if I get into trouble, try my hold and retreat... Teams are often even worse. Tank goes in and grabs aggro. Healer heals the tank. Cannons assist the tank and wait until he has sufficiently taunted the spawn. Crowd Control be ready for adds...

Now, there are slight variations game to game, and I still think that in the right situation, any game can produce a challenging and fun combat. It is just that after a certain point, the cool combats become few and the mind-numbing, "hit the 2 button, then the 3 button, then the 2 button, then the 4 button..." stuff becomes more common.

Often, I will hear talk, or read posts where people talk about "better AI" in a computer game. In fact, one of my friends was recently bemoaning the boring combat AI in CoH. "I feel that combats in CoH have become dull and repetitive, Chris. Certainly a better AI would be the proper solution." Well, he didn't really say it like that. In fact he was a tad drunk and so his critique came out more like, "Dude, the f'ing opponents in this game are f'ing stupid. Those f'ing programmers need to stop f'ing nerfing f'ing Energy Melee and write some f'ing AI!!! (burp)."

But here's the key: programming a computer opponent for a game is hard. Programming a computer opponent that can actually engage and challenge a human player with any level of skill is even harder. Chess programs are perhaps the most successful examples of this type of program and the techniques used there, not to mention the coding effort, and processing power involved would not translate well to an MMO.

The good news is that you don't really need a good computer opponent to run your MMO combats. You are trying to provide interesting and challenging combats; you aren't actually trying to win against the players. Players don't want to be beaten as much as they want to be surprised, challenged and given interesting situations to which they have to react. The current aggro-based system in CoH does none of this.

What if we change the system and give the computer opponents a few more behaviors from which to choose. When the players first engage a spawn, it usually behaves in the old, dull way. However, there is a small chance that they will react with a pre-set behavior, chosen randomly from a list (but based on the characters in the attacking group). In CoX, I can imagine that if a group had a boss in it, the boss might call out the orders to his minions, thus giving the players a heads-up as to what was coming.

Divide and Conquer -- All enemies in the spawn pick a random target from the player team and aggro onto him. Player tanks and crowd control will have to act quickly to save their less hardy teammates. The aggro is enough that a single taunt won't cause the enemy to retarget, but it will take multiple taunt-type powers, or holds to get the situation back under control.

Scatter -- The enemies aggro as normal, but also do their best to keep players at a distance and use ranged attacks. This spreads the fight out, making AoE attacks less useful. It also adds potential danger as the PCs spread themselves out to engage the outliers, possibly getting into 1-on-1 fights they cannot win, or wandering into other spawns. A disciplined team ought to be able to stand their ground and pick off opponents, but an undisciplined team might get themselves into trouble.

Spring the Trap --  The enemies aggro as normal, but also cause a similar-sized group to spawn nearby and rush to the combat. Within a few seconds, the new combatants arrive to make life difficult. Crowd control characters will have to pay close attention, find the new spawn and neuter it before any damage can be done. Alternately, for a nastier surprise, the enemies can teleport in, or deactivate their invisibility shields and appear right on top of the fight.

Use the New <blank> Rays We've Acquired -- The enemies aggro like normal, but instead of their normal attacks, they get a new attack of a random type that they don't normally have. Not a big deal in most cases, but it is possible that the new attack type hits a specific hole in the characters' defenses (could be psi in CoH). Players will have to grin and bear the new attack type, or use holds to prevent from being attacked.

Run for Help -- The enemies flee from combat to the nearest spawn. Once they reach that spawn, both spawns will turn toward the players and attack. Players will have to kill the enemies quickly, slow them, or hold them to avoid adds.

Bob, You Run for Help -- Similar to the above order, but instead, only one guy will run for help, while the rest of the spawn fights. If Bob reaches a spawn, it aggros to the players' location and Bob will continue to the next nearest spawn. Players will need to find Bob in the chaos of battle and deal with him quickly, or else the whole map could come down on them.

... and so on... I could think of more, but I think you get the idea. What do you think?


moptopRPG writes:

Dude, i have never really thought about, though i have become bored many a time because of edlessly the same combat. This is a great idea it would add a whole new aspect to mmo's.  It would definatly spark my interest and make me want to play longer. Two thumbs up Sir!

Fri Aug 08 2008 4:25PM Report
Bopper writes:

I think it's a pretty good idea. In my university lecture on gaming AI we actually discussed this. You can program an AI to be smart and kick the shit out of you but that wouldn't be any fun. Instead, you should incorporate more "randomness" to the fight to keep things fresh and interesting. Some of those preset AI routines you described would be a cool idea.

Implementing them, however, would be pretty damn difficult. Not knowing what an enemy will do, or even worse, a group of enemies, could make things too chaotic and random and would result in a huge spike in difficulty level. In certain situations, and in controlled environments this could be a lot of fun though.

Some enemies in WoW have some basic AI routines that mix things up, but these are very few and far between. Some mobs will run away when low on health, some will call out for allies, some will trap you and hit you ranged, and so forth. I would like to see more dynamic fights and more random AI patterns. Making the mobs slightly easier but compensating for that with more unpredictable AI routines could spice things up.

The more the mind experiences a certain event, the less exciting it gets, it is basic psychology. This is a fundamental aspect that I think should be more properly understood by developers. Mix things up! Make things more dynamic, more random! Doesn't have to be everything but at least something. If I know exactly what is going to happen in a given situation, then why am I even bothering to do it?

Another idea is having groups where the "boss" monster needs to be killed while the others just get resummoned or cannot be hurt, or have the group disperse and run for help requiring you to stay on your toes, or other pre-scripted variations to make things a little more dynamic.

Fri Aug 08 2008 4:31PM Report
midnightgame writes:

dude ur silly not mmorpg u u u u u hahaahahah u u u u u

Fri Aug 08 2008 4:36PM Report
midnightgame writes:

oh by the way im just ur ordinary kid who is the bast gamer in the hoole world thats all

Fri Aug 08 2008 4:38PM Report
cmagoun writes:

Bopper, I do think something like this could spike the difficulty of the game at times, but I am not certain that is a bad thing. CoX especially could use a little variability in the difficulty... and let's face it, soon after these random routines were implemented, groups would learn how to counter them.

The nice thing is, even a dozen or so of these would make combats more exciting and then with each update, one or two more could be seeded into the behaviors of the computer opponents.

Fri Aug 08 2008 4:46PM Report
Bopper writes:

Would be cool to see something like that, as long as I'm not the one who has to program it and balance it :D

Fri Aug 08 2008 5:39PM Report
Anofalye writes:

You forget the basic.


Peoples play with levels and everything because they can see, asset and judge the improvement of their character.  The system could be as perfect as you want, if you don't feel the progression, the developpment, the becoming, the players won't stick.


What is the driving and itching desire going to be?  What will make the player want to come play for an extra hour?  The developpment must be something the players can not only asset, but desire deeply.


Realism and challenges are welcome side order, but never at the expanse of the main course.  Games need to be entertaining and fun, and if realism and challenge can be entertaining and fun to some extand, they are not that by definition.  Advancements and progression is fun by definition.

Fri Aug 08 2008 6:41PM Report
HumbleHobo writes: Step 1: Make terrain matter. Melee attacks launced from low ground, or in a marshland will have less power. If you just add a simple "high-ground" bonus to damage, it adds depth. Better yet, make everything run slower up an incline, and faster down one. Just adding those simple things will change strategy for each encounter a little. Keep people on their toes. Step 2: Make different strategies viable, instead of: "Clear every single group of trash mobs so you can fight the boss". Strategies such as sending a group member around back to draw the enemies out, then the rest of the group sneaks in the entrance unchecked. Fri Aug 08 2008 7:42PM Report
Loke666 writes:

Funny, I think other points are dumber... Like the fact that MMOs dont have hit locations, something most Pnp RPGS have had since early 80s.

Like the fact that I doubt any MMO dev had even held in a sword, there is a big difference in how a sword damage someone and if it pentrates a specific armour, A guy in full armour is almost impossible to defeat for an unarmored guy unless he got a polearm.

And like humble said, terrain modifiers should be in, and weather modifiers too. Try firing a bow in rain.

And the level system is stupid, in real life a guy in a good armour with a good sword will defeat a better train person with a knife and no armour almost everytime.

I'm not saying the system should be 100% realistic but MMO combat is still where the pen and papers game where in the 70s.

Sat Aug 09 2008 3:07AM Report
PSoponyai writes:

It's all a matter of opinion and in the long run, numbers decide the outcome.

What I mean with this is that, sure it would be fun and interesting to battle clever enemies for a few, but to most it would be simply too much. Think about the kids that play the games.

Developers do things for the profit, so if something makes most people fed up with it, they won't implement them.

As for behaviors, many games have them. Even for example Diablo 2. Many mobs attack in group, and they do flee when they are loosing. Or think about the zombies, they truly are mindless and they just come at you.

But take Lord of The Rings Online for another example. Wolves always howl for help, when I attack them and if there are other wolves around, they come and help each other.

But when all is said and done, we are humans and as such we always want more. No matter how cool or new a system is, in time you will get to know it and end up bored with it.

Sat Aug 09 2008 7:20AM Report
Bopper writes:

Terrain and weather modifiers take up a lot of processing power and require a lot of real-time calculations on the server's part. The costs of this system would greatly outweigh the benefits. Nice ideas, but they would work more with a multiplayer FPS game and not an MMORPG.

"Realism and challenges are welcome side order, but never at the expanse of the main course."

I agree. Realism is not a priority in MMORPG's, at least not to the extent that it hampers the funfactor or gameplay. We aren't talking about realism necessarily though, just some more varied AI routines for single/group encounters to spice things up.

"Step 1: Make terrain matter."

This is an interesting idea but again brings along way too many problems that outweigh the benefits. Having a system where using cover and terrain would be nice, but it is hard to incorporate in a meaningful and fun way in MMORPG's. It was a planned feature for ranged combat in AoC which was ultimately dropped. The cover system was also tried for Tabula Rasa which turned out to be more of a gimmick than anything. Again, these ideas would be better suited to a FPS or more combat oriented game instead of the traditional MMORPG.

Another problem is that this would cause a tremendous amount of bugs, i.e. not being able to hit mobs at all, mobs doing way too much damage to you and players abusing the system.

"Step 2: Make different strategies viable, instead of: "Clear every single group of trash mobs so you can fight the boss"."

I absolutely agree with this and hope to see more of it in future games. This is something we tried to address with cmagoun. Changes in encounter mechanics and specifically enemy A.I. routines would result in more varied and dynamic battles, ensuring some sort of variety.


Funny, I think other points are dumber... Like the fact that MMOs dont have hit locations, something most Pnp RPGS have had since early 80s.

Like the fact that I doubt any MMO dev had even held in a sword, there is a big difference in how a sword damage someone and if it pentrates a specific armour, A guy in full armour is almost impossible to defeat for an unarmored guy unless he got a polearm."

These kinds of realism and strategic concepts are very hard to implement in a MMORPG and would hamper some of the gameplay mechanics that reward progression. For the sake of balance and fun things like these need to be overlooked. Hit locations would require you to either physically target an opponent which is very taxing on a server to calculate and would result in lag. It works fine in FPS games since they are designed and operate on a more "real-time" basis.

"And like humble said, terrain modifiers should be in, and weather modifiers too. Try firing a bow in rain."

While this is true for real life, try telling a character class whose damage is 80% ranged that his character can't operate effectively simply because of the weather :D This would lead to some pretty pissed off players.

"And the level system is stupid, in real life a guy in a good armour with a good sword will defeat a better train person with a knife and no armour almost everytime."

Again in real life this makes sense but you just can't do this in a game. Level progression is a fundamental gameplay mechanic that rewards a player for achieving certain things within a game. I agree that games do require a skill-based factor but having this would defeat the purpose of levelling or advancing in the game.

This mechanic works well in a game with no levels that are more skill-based and require more real-time skills (such as Counter-Strike) but they would not translate well to a MMORPG.

Age of Conan is a nice example where the gameplay is more reliant on real-time factors such as manually attacking, pulling off combos, actively blocking, and so on. It's also a classic example of what this kind of system can lead to when you have many players fighting at the same time. In PvP the lag affects the combat in such a way that even though you may be swinging at a character, the server is registering the character as standing 2 meters to the left of where your client sees him, therefore rendering your combo ineffective.

Until we have the internet infrastructure (faster servers, T1 lines for everyone :D) to allow these kinds of real-time and intensive server-client interactions to take place, we won't be seeing these kinds of things in a MMORPG. Personally, they are not too important to me anyway. In a FPS for example it is a different story.

Sat Aug 09 2008 7:50AM Report
cmagoun writes:

Lots of comments...

Anofalye: I think people play MMOs for lots of different reasons, but I understand your point -- most people don't want their MMOs to be hard at the expense of character progression. I agree. However, there are players that want challenges and more commonly, there are players who occassionally want interesting stuff to happen to them. While I think randomizing the combat behavior of enemies has the danger of spiking difficulty, I think that danger could be mitigated by design and rarity -- the typical spawn would follow the typical behavior and we could let the player decide how often they wanted the challenging mode to occur.

HumbleHobo: All good ideas and I long for the day when we have MMO combats that reward positional tactics over aggro management. Hard to do... I know, but it would be great.

Loke666: Actually, if you look at PnP games today, they often lack such "realistic" combat features as you describe. The trend in PnPs as I understand it was for more complex systems in the 80s, swinging back to a more narrative, story style in the 90s. In many cases, that style persisted... until D&D 3.0 which had a crunchy system. D&D4 continues the crunchy trend, but not in the way you describe. D&D's tactics are very abstract -- more akin to a collectable minis boardgame than a "realistic" wargame.

Mon Aug 11 2008 10:00AM Report
cmagoun writes:

Ohh sorry about the crazy bolding... not sure how that happened

Mon Aug 11 2008 10:01AM Report writes:
Login or Register to post a comment