Gamasutra: The Art & Business of Making Gamesspacer
View All     RSS
October 25, 2014
arrowPress Releases
October 25, 2014
PR Newswire
View All
View All     Submit Event

If you enjoy reading this site, you might also want to check out these UBM Tech sites:

AI and MMOs - The Controversy
by Dave Mark on 09/06/09 04:52:00 pm   Expert Blogs

The following blog post, unless otherwise noted, was written by a member of Gamasutra’s community.
The thoughts and opinions expressed are those of the writer and not Gamasutra or its parent company.


There have been a significant surge of blog posts (and ensuing conversations) on Gamasutra recently dealing with MMOs. These discussions cover a number of topics but there are two consistent threads that run through them - player socialization and the paucity of adequate AI. I find this oddly fortuitous in that I am delivering a lecture at GDC Austin next week on the nexus of these two seemingly unrelated issues.

To paste from the GDC Austin site:


Cover Me! Promoting MMO Player Interaction through Advanced AI
Session Description 
Historically, PvE AI in MMOs has been a straight-forward affair. While this leads to predictability, it also leads to monotony. In online, team-based PvP games, however, much of the attraction is the dynamic nature of the engagement that necessitates that players read, communicate, and react appropriately to changing, even unexpected actions of their enemies. By leveraging more advanced techniques that are becoming common in FPS, RPG, and RTS games, the AI in MMOs can be designed to provide some of the more attractive and engaging elements of PvP games. This, in turn, can lead to more involved team play, greater replayability, and an increased sense of community in the game.
This lecture shows examples of some of the aspects of PvP games that are attractive to players, the AI techniques that can be used to replicate them, and the effect that inclusion of these aspects can have in an MMO environment. The attendee will leave with a variety of concepts that can be included in their own MMO designs.


For reference, the blog posts I refer to above are specifically:

Isn't Anyone Tired Of The Same Old MMO Classes Yet? by Edward Hunter
Bring Real Socialization Back Into MMOs Or Else! by Edward Hunter
An Alternative to Aggro, by Bart Stewart

Additionally, I would like to point to an article by Gregory Weir from back in June:

Analysis: Interplay in Left 4 Dead

While Edward Hunter's original post about MMO classes wasn't directly related to the topic, the comments took a significant turn when Jonathon Walsh said rather innocuously:

"The problem with PvE focused games is they are inherently working against their greatest strength. The draw of an MMO is playing alongside thousands of other players. PvE works directly against this by splitting up players into small groups and giving them a similar experience to what they could get by hosting up an RPG using a matchmaking service. The other problem with PvE content is that it will always be 'solved' at some point. Very well done randomly generated content and AI could change this a bit but that's definitely unfeasible at this point in time."

While I agree with his point that the draw of an MMO is playing alongside many other people, I disagree that PvE works directly against that. An instanced server does do so as he mentions. This is really no different from a very large game of Left 4 Dead. However, that is not the fault of PvE. Large world games such as back in the days of UO gave you the feeling of playing with or against many people but also the environmental monsters (AI).

Additionally, while I agree with his statement that randomly generated AI content could go a long way towards breaking away from the "solvable" AI encounters that we are used to now, I disagree with his position that this is "definitely unfeasible". That is not the case. There are plenty of techniques that can be used to create unpredictable, yet reasonable-looking actions in AI actors. I actually dedicate a significant portion of my book, Behavioral Mathematics for Game AI to these techniques and will touch (briefly) on some of them in my lecture.

Still, his points were well taken. The discussion continued in ernest, however, when Owain abArawn dropped this bomb:

"Christopher, one reason for the lack of choices in WoW, as you describe, is the cop out that developers all use of substituting hit points for AI. Raid levels aren't necessarily hard because the mobs are skilled or use effective tactics as much as they are insanely overpowered or nearly invulnerable. How much easier is it to give a mob 100,000 hit points than it is to develop an effective, adaptive, flexible AI. This is one reason why developers find themselves on a perpetual treadmill of providing updates, expansions, and additional content. Their current content is static and predictable. Perhaps a better solution would be to implement a dynamic unpredictable spawning system with mobs that have better AI. That way, your content refreshes itself, and players can never be quite sure what they will encounter, where they will encounter it, or what it will do when it is encountered." 

This is very true in many cases. In fact, this state of MMO AI actually harkens back to the early days of single-player AI - give the boss a lot of hitpoints and he is harder to kill than his minions. He is no smarter, however. You simply must repeat the same process that many more times in succession in order to defeat him. And that certainly doesn't make for engaging gameplay. While the spawning system has a little to do with it (and I cover this in my Austin lecture), the actual encounters are what needs to see the most significant change. Regardless, having the sense of the unknown does make for a significant change in how players must approach the game - as a group.

After Christopher Wragg argued briefly that he didn't believe changing AI would help the title topic (that of new classes), Owain responded again in support of the AI-based solution. He asserted that the whole reason that we are stuck in the tank/healer/dps class model is that the AI across many games has not progressed.

"I have used variations on that same formula in UO, DaOC, ShadowBane, Star Wars Galaxies, WoW, Vanguard, Age of Conan, Warhammer Online, and most recently, DarkFall, so I think I am pretty secure in asserting that all those games, over a span of a decade, have all used pretty much the same dogshit industry standard AI model. Raids are only different in the number of friends you need to drag into the party to compensate for the insane number of hit points/damage capabilities of your typical boss. The AI is usually just as stoopid.

Now consider how much success I'd have with those same tactics against a group of other players in a free for all PvP environment. I'd never get the first shot off before being charged, flanked, mobbed and pwn'ed."

This is exactly my argument in my lecture - compare the excitement and fluidity of a game such as Team Fortress 2 to the rote memorization and execution of many of the games he mentioned above. The point is, many of the tactics we use in MMOs against "stoopid" AI don't fly in PvP games.

Christopher Wragg responded:

"Now you miss the point, the lack of *choices* available to players has little to do with AI. If anything simple AI, raises the number of choices available to players (working on the concept of, if you are not being forced to react, you are instead free to act at will). For instance, given more complicated AI in a game like WoW, wouldn't raise the choices available to players, if anything it reduces the number of "correct" actions that can be taken. As such the classes would still have the same restrictions placed on them and more complicated AI would simply mean everyone dies."

This is where I really disagree. In fact, I believe that the exact opposite is true. Christopher claims that the lack of AI increases the number of choices available to players. I can't imagine a scenario in any sort of game where this would be true. I suppose I could see where Chris was coming from in this respect. Against simple AI you can do many different things. Theoretically, you could do an audition for an animated version of "So You Think You Can Dance" - but that won't bring you any closer to your goal of winning the battle. Usually the action set of things that you "must" do to win the battle is limited to a rather narrow "golden path". This is the phenomenon that gives rise to sites like TankSpot. Watch the video, play your part, win the battle.

By including good AI, however, you are significantly increasing the number of possible actions you may have to do in order to win the battle.It is a commonly known fact that the more potential responses that an opponent may take to your actions increases the branching factor significantly. Additionally, the more actions that an opponent can take against you (i.e. through good AI), the more potential reactions you must be prepared to exhibit.

I submit that PvP could be used as an example of the pinnacle of AI - the best we can hope for. However, in a PvP game (whether it be in an MMO or a game such as TF2), the gameplay is far more wide and varied because of the combinatorial explosion that results from many independent people making a wide spectrum of choices.

I also take issue with his overly simplistic (and almost completely binary) statement that "more complicated AI would simply mean everyone dies." This is actually a rather silly notion. One could have said the same thing about giving a boss monster 10,000 hit points. "How can we win against someone that is 100x more powerful than us?!?" That's not at all the case... it just takes longer. In the case of better AI, it is also not the case. The difference being that you may have to actually do more than one type of attack in order to counter the more clever AI.

He continued...

"If the enemy is smart enough to bypass the warrior, then what point having him, if the enemy is clever enough to kill the mage, the mage will die. The classes in WoW are not flexible enough to handle such a scenario. My point, is that without even any increase in AI or change to the enemy abilities, skill choice and use could be made far greater. This raises the number of choices a set of players could make before and during battle. As I was saying before, classes shouldn't be built with such harsh roles built into them. The players should be presented with a scenario, and then given the ability to work out their own way of surmounting it. Clever use of AI should merely alter the nature of the challenge.


If you change AI you MUST alter class as well to balance your game again."

Again (and I hate to pick on Chris, but he is illustrating an unfortunate meme in the industry - by players and developers both), this is a horribly narrow-sighted statement.

Using TF2 again as an example... each class has its own strengths and weaknesses. One obvious example is the use of the engineer's sentry. If we were to say "spies can go through our ranks and destroy sentries so what's the point of having them?" it would be viewed not as a statement of obsolescence of the sentry, but rather a call to arms to other players to make sure that spies don't go through their ranks. In fact, many times, pyros are called upon to patrol the less-active areas of the current battle specifically to look for spies. That is what pyros do well.

The same can be said for other classes. Many classes get mowed down by those same sentries... should we all give up on the battle? No... the preferred method for dispatching said sentries is for a medic and demo to go in with a deployed uber-charge specifically to take down the hardware and allow the others through. It is a big game of rock-paper-scissors. If, for some reason, the other team doesn't have sentries but is overloaded with pyros, my team will counter with heavies instead of demos. Heavies (especially with medics in tow) stop pyros in their tracks.

The dynamics of the team balance - the necessity for the different classes - is specifically caused by the fact that we do NOT know what we are going to be facing. If the opponents were to be completely static and predictable (analogous to poor AI), I could tell you exactly which classes we should have on our team and what we should attack in what order. There is one best way of doing it. The component of unpredictability and the surprise that can result is exactly what causes us to have to make decisions on the fly, communicate what we see, formulate plans, assign the proper person to the role, and execute this plan as a coordinated team - and this entire re-planning process repeats often - even multiple times a minute.

In the end, his final claim that altering the AI forces the designer to alter the classes to re-balance the game is not necessarily true. As I illustrate above, it just changes how we must use those classes. In my Austin lecture, I use the exact example that he mentions... that of an enemy flanking to go after the "squishies" in the back. All this means is that the fighters in front can't simply pull up a lawn chair and be content with their mindless button-mashing. They must be ready, if necessary, to rush back to the aid of their pals - or better yet, alter their tactics originally to be aware of the potential for flanking and prevent it to begin with. See how the increased AI can lead not only to interesting decisions but also more interaction between players?

Quick question to Owain abArawn regarding the following comment:

"Consider the following real world scenario. I am faced with two opponents. The opponent that is closer to me is armed with a knife. Behind him is a person holding a gun. Yes, the person holding a gun is a significant threat, but if I ignore the person right in front of me holding the knife and try to push past him, the person holding the gun never need fire because I will be killed by the knife wielder. That is a failure in intelligence on my part, and in an MMO, it is a failure in AI."

Seriously, dude... did you read my book already? That is almost straight out of chapter 14!

Edward Hunters's second post, Bring Real Socialization Back Into MMOs Or Else!, is more directly on-topic to my lecture in Austin. I particularly was fond of his observation that "... raid groups and 'Guilds' are not social situations in the MMO's today.  There is no difference, typically, in these social interactions and those you would find in your typical IM conversation or Facebook thread." The reason for this is largely because players do not have to communicate much game-related information to each other in MMOs these days. Again, contrast that to online PvP shooters such as TF2 where communication of relevant information is crucial to survival as an individual and success as a team.

Edward's main focus, however, was not on combat but rather on the "RP" aspect of MMORPG. We are agreed that oft-times MMOs are simply virtual chatrooms. This isn't as much my purview, however.

My new best friend, Owain abArawn, commented at one point,

"My comments apply to PvP-centric games, since the AI in PvE MMOs is typically strongly lacking."

I wish this wasn't mutually exclusive. As I mentioned above - and which is the focus of my lecture - I believe that if we move the AI in the MMO space to be more in-line with what we encounter in PvP, we can do a lot towards generating that community that Owain goes on to talk about in his comment.

In An Alternative to Aggro, Bart Stewart gently rails against the industry standard of "aggro" and goes on to suggest some relatively viable alternatives. While I won't comment on his suggestions here, I agree with his premise that the current concept of aggro is a horribly misshapen vestigial relic from the days when games couldn't afford to do expensive processing to determine something as simple as who/what/where/when to attack (not in that order, of course). The result is that all the enemies in MMO games are effectively "on strings". We tell them when to attack (when we are ready), who to attack (the tank), we know exactly how they will attack, etc. All sense of autonomy of the enemy is completely lost. We go back to the idea that MMO encounters are a "solvable" puzzle much like the little sliding number puzzles. All we need to do is push the little tiles around in the right order until we get an arrangement we like. There is no reaction necessary on our part... only control.

With loads of processing power available to us, many (but not all) single-player games have far more complicated algorithms in place that give a far more realistic look and feel to the initial and ongoing interaction with the player. (I give considerable treatment to this in my book and will give some quick examples of alternatives to standard aggro in my lecture.) The odd thing is that, for the most part, we haven't lifted these algorithms out of the single-player world into the MMO world. Really, the algorithms to deal with multiple units aren't that much more complicated than they are in single-player games. What's more, they can be entirely data driven, flexible, and adaptable to changing situations. Once again, if our goal is to have a more engaging interaction with the AI - not unlike what we see and feel in PvP games, we need to imbue our creations with more person-like thought processes. And, as we know, people are a little unpredictable. Reasonable, yet neither random nor rigid.

I was appalled at some of the comments that were defending the notion of simplistic aggro, however. (Johannes Smidelöv noticed this as well.) We have established memes like tank/healer/dps so deeply in our collective MMO-playing psyche that we use them as a defense against what caused those strategies in the first place. That is, current aggro models begat the idea of tank/healer/dps, and yet we defend the aggro system by saying "what will the healer do if he has to heal someone other than the tank?!?" It's almost Stockholm Syndrome-esque that we are so worshipful of the very captor of gameplay creativity that we are so willing to use our own mind-numbing boredom to defend it.

There were flashes of hope, however. Andre Gagne summed it up nicely when he mused,

"With a different type of AI the combat itself would fundamentally change and be, potentially, more interesting."

Amen, my brother.

Farther down, Maxime Doucet echoed the quick example that was suggested by Owain abArawn above...

"The AI should at least be able to compare both possibilities: should I attempt to take down the tank who's standing in front of me, or should I try bypassing him and aim for the healer, who's 20 yards behind him?"

Again, this is an example straight out of my both my book and my lecture. There are plenty of ways to do this that make the character seem reasonable yet varied. (Incidentally, one of the main problems with aggro is that it exposes the fact that the AI knows nothing of the concept of futility... e.g. "I will never kill the tank as long as the healer is buffing him up.") Maxime concluded with the observation,

"This would make the whole PvE encounters much less predictable and more adventurous, and way interesting for an experienced player as well, as the players would encounter mixed groups of different kinds of mobs with widely different AI behaviors and triggers."

I did get a kick out of what Matt Kane said... he looked at the reason for aggro from a slightly different angle:

"The real culprit here is the design of typical raid encounters. 1 big raid boss sitting still for 5 years picking his nose waiting for a group of 4-50 random guys to come and hack at his shins for 10 minutes and 24 seconds til he suddenly dies of congestive heart failure.

The boss has no goals. If you wouldn't have come along with your group of buddies, he would have probably continued to pick his nose for another 5 years.

So designers compensate for the fact that they develop enemies and encounters with no purpose by imposing standardized, shallow enemy motivation.

Instead of attempting to conquer the world, your terrifying end boss to the dungeon you've been trudging through for hours is actually playing whack-a-mole with one of your UI elements.

A good start to solving boss mechanics would be endowing bosses with motivation and purpose (beyond dialogue text)."

To sum up, he suggests aggro is a bastard child of the simplistic design of MMOs to begin with. That is definitely beyond the scope of my column... but I do want to suggest that there may be a "chicken and egg" phenomenon going on here. Designers may not even be aware of the fact that we can do nifty things. I would suspect that they design their worlds and encounters in this way because they haven't been aware that there are other things that we can do with AI to enable their creativity. That is largely the purpose of my lecture in Austin, by the way... not to "solve" anything but rather to get the creative juices flowing.

That being said, I hope that those of you who have commented or even simply read the referenced threads and are also going to be Austin will stop by and take in my lecture. Please make a point to introduce yourself. I believe you will find it entertaining as well as informative. If my audience is even remotely as spirited as these discussions, I'm going to very interested in taking questions at the end although I should only have 5 minutes or so. Regardless, because I am on day one of the main conference we can take the rest of the week to hash it out.

See you in Austin!

Dave Mark
President & Lead AI Consultant
Intrinsic Algorithm LLC

Related Jobs

Red 5 Studios
Red 5 Studios — Orange County, California, United States

Graphics Programmer
Red 5 Studios
Red 5 Studios — Orange County, California, United States

Gameplay Programmer
Gearbox Software
Gearbox Software — Plano, Texas, United States

Server Programmer
Forio — San Francisco, California, United States

Web Application Developer Team Lead


Christopher Wragg
profile image
Shiny, an entire post "mostly" dedicated to breaking down my points forgetting most of them are in reference specifically to WoW and class options and diversity of tactics.

In essence, friend you should give lessons to the media about how to twist things out of context. You seem to be on a "rail against bad AI" kick, and far it be it from me to stop you (I do, in essence, agree), but I can't help but take some minor offence at the obtuse manipulation of my points.

Anyways back to the topics at hand. Complex AI *always*, reduces the options open to the player, that is one of the main points of it. The more reactionary, the more options available to it, and the stronger you make it, the fewer *correct* options the players have. Now as you say my examples are quite simplistic, but if you have a mob that does nothing but move to the closest enemy and attack you have a huge number of tactics available to you, you have the greatest number of "golden paths" open. Do you have an interesting encounter, maybe not, but we were talking about diversity after all. WoW itself seeks to reduce the number of options not through AI, but through increasing the power of the mob and making it "immune" to various sets of abilities (which I'm personally against).

Your argument is also flawed when you start mentioning pvp, for instance, pvp is different and complex each time because of an unknown combination of enemies, it's like running into a boss in a game and the boss is different each time. This requires you to bring a more flexible array of skills and tactics with you as you don't know what you will face, it doesn't increase the diversity of effective tactics against a mob (the same golden path exists, meeting it's requirements is harder though). A cunning (per unit) AI rationale doesn't replicate this, it merely represents interaction on the field, which SHOULD NOT be unpredictable in it's entirety, this doesn't lead to interesting matches, this leads to frustrating matches, because the player CANNOT form a viable strategy whatsoever (it's like a game that asks a player to form a strategy for picking a number off a [truly] random dice roll.)

Also you took points from an RPG and started talking about the interplay of class in an FPS. The secret to an FPS is that there is no real class, TF2 implements it but in no restricting fashion (aka TF2 pretends to have classes), you can simply switch class at need. So rather than building a character or team to face a challenge, all you're doing is adjusting it on the fly, the nature of the two require entirely different rationales (not to mention an FPS is far more subject to actual player skill to overcome clever AI, which is a balancing factor in and of itself that is less prevalent in RPGs). Also my AI-class alteration comment holds true even for TF2, if we alter the AI of one class in such a way that it is now capable of taking apart a class it couldn't before (so in a rock/paper/scissors scenario this is like rock all of a sudden being able to win against both paper and scissors), then something needs to be changed to balance this out (the addition of another class or the alteration of another). Along these lines TF2 didn't START balanced, when it was naught but a glimmer in the eye of a dev, it still had many iterations of clever tweaking and class balancing to go through before it hit prod, it's not like it's balance is some clever function of AI alone.

Another interesting concept to point at is emergent AI. You know, a bunch of really simple rules that AI follows that actually turn into quite complex interplay between players and AI simply because of it's simplicity. Simple AI is nothing to be scared of, bad AI is, two different concepts entirely.

Either way, it's ultimately naive to say that AI adjustments don't affect class balance, it's also naive to state that complex AI will necessarily lead to more interesting and "fun" game play interactions. Sure it *can*, but take it with a pinch of salt (it can also make your game clunky, inefficient, boring, difficult and frustrating even laughable), basic AI can be just as interesting, if not more so, railing against it without understanding it seems like "cutting your nose off just to spite your face".


Anyway hope I don't come off too aggressive, had to defend my points, was always brought up with the notion that if you can't defend your opinion not much point having one =P

Jamie Mann
profile image
When it comes down to it, most games come down to a risk/reward structure: player manages to successfully complete activity X, player receives Y - Y can be anything from in-game achievements (e.g. experience, equipment, money), external achievements (e.g. points, awards) or just satisfaction.

The key there is the fact that the level of reward has to be balanced against the risk. Making the risk too low devalues of the reward; if it's too high, then the players may give up in frustration, having deemed the reward insufficient.

Which brings us nicely to the point: game designers and developers may want to implement nice, complex, shiny AI: but is this what the players actually want? There are many opportunities in the MMORPG arena for PvP combat, but the vast majority of players opt for PvE instead: how much of this is because the tank/healer/dps model is simple to understand and apply? There's generally enough variety in the scenarios to keep interest up, which combines with the team-work to make for an enjoyable experience. You could even argue that PvE is a meta-grind of sorts!

lee cummings
profile image
I'll try, once again, as I did in the “Aggro” thread to outline what seems to be wrong with the way you guys are attacking the concept.

But first... I'm confused by the following statement:

“The reason for this is largely because players do not have to communicate much game-related information to each other in MMOs these days. Again, contrast that to online PvP shooters such as TF2 where communication of relevant information is crucial to survival as an individual and success as a team.”

Really? You really believe that a group of players fighting each other in TF2 need more real time information passed between them then, oh lets say, end-game WoW guilds fighting cutting edge content? Or a hundred planning a co-ordinated attack on a keep in Dark Age of Camelot? Or dozens of players participating in end-game PvP in Aion? This absolutely staggers me, and I don't for one minute believe that you've been on Ventrillo or Teamspeak with a few dozen friends facing challenging end-game content in an MMO.

Your attack on class roles confuses the hell out of me – roles didn't appear in games because of the concept of aggro – roles appeared in games because we, the players, need those roles. They exist for a multitude of reasons. They help users get a grip on game play in what can be a very complicated game world to just jump into (see: branching class trees which go from simplistic to specialized classes over time), they give us individuality and purpose during game play (how much fun is it being just another nameless grunt on the front line, versus being a specialized class?), they let designers define better experiences because some things set in stone to balance against (see: issues some MMOs have when players push one stat or skill far enough that they break encounters – the players have essentially created a highly specialized subclass – tanking monks in EQ, for example). And what is the ultimate proof that we need class structures? Because even in games where we don't design them, they appear anyway as soon as users get to grips with the game (see: high-skill team based first person shooters where even though classes such as “the defender”, or “the scout” don't exist, users gear up their characters and act in that way anyway).

It's astounding to me that we've got to this point in the conversation without defining what aggro is. It's a state, governed by a ruleset, which decides a course of action, which at some level has a numerical value, or at least a flag saying that I'm attacking X, not attacking Y.

Which has to exist – an enemy needs to attack something, and follows rules to pick that something. And while I wholeheartedly agree that we can always make these situations better, I haven't seen one example of how to accomplish that. While you can scream “amen brother” from the rooftops in support of the guy who says things should change, having no concrete ideas how to establish that change means you're just noise.

A common example posted in both these threads is that an enemy could know that having a healer present should cause him to go after that healer, since it will make the whole encounter easier to “beat”. Lets also stay with spaceships, because that seems to be fun. Yep, I'm going to use the concept of healer even though it's so last week.

Attacking squad goes after my healer. That's all well and good, but now I need to be able to react to that in a number of different ways, or game play gets very boring, very quickly.

I shuffle my formation to physically protect my healer ship, putting some metal in between him and them. Pretty cool idea. Or another ship who wasn't acting as a healer in this encounter has to change mid-stream and take over healer duties, while the first healer has to change it's configuration and can now take damage. Ha! That'll show them.

Or I use an attack skill which shifts around their formation so that, for a short time, their damaging attack craft are way at the back, and their support craft are at the front, turning the tide of the battle until they can get back into formation. Or on a smaller scale I shunt their strongest attacker out of range and tether him there, buying some time.

Or I find some way to deflect some of the incoming damage off my healer craft and direct it to another craft, or crafts in my formation. Or I focus instead on a rapid counter-attack, forcing some of their attacking units into a more defensive state, potentially salvaging the situation.

Now, here's my issue with people raging against “aggro”. Every example which they give for game play situations which would be an improvement over current PvE MMO design already exist. They exist, and are built on top of various “aggro” systems because aggro is simply a tool to direct an action. It's not a crutch, or a flaw, in the same way that shooting a weapon at an enemy in a FPS is a crutch or a flaw. I suspect that a lot of the people commenting against aggro think it's simplistic because they've only seen simplistic examples of what can be done (and I can't believe I have to type this) when systems deciding who an AI entity should attack are built on a system of numerical values which can be modified during game play by either player or AI actors. It's a simple system, which is simple for lots of good reasons, but which, time and time again, has been used for exhilarating encounters and complex game play.

Killing rats at level 1 in any MMO who act as stupid as mud do so for exactly the same set of reasons that 40-70 person raid encounters exist which take weeks for the best players in the world to get to grips with their intricacies and overcome them.

So, back to my example above with my squad being attack by an enemy squad, which is focusing on my healer. I find all the examples I gave as game play replies to that “attack healer” assumption to be simple, but quite interesting in their own way, and cover a lot of what could be done to make an encounter interesting to players – from counterattacks, to player role changes, to utilizing line of sight / collision, and so on.

However, all of these examples already exist*, and they all exist on top of the aggro systems which are being attacked.

Once again – aggro is little more than selecting who should be attacked based on a set of rules, and letting players and enemies change those rules during combat to create a much more interesting and reactive experience. And I've yet to read one comment which outlines a system to replace this (and no, adding story reasons to explain why enemies act the way they do is not a replacement, just an enhancement), where the game play outcome of a particular encounter would appear significantly different to the end user who doesn't know how the numbers are being crunched behind the scenes.

Christopher nailed something really nicely in his above post when he said:

“You know, a bunch of really simple rules that AI follows that actually turn into quite complex interplay between players and AI simply because of it's simplicity. Simple AI is nothing to be scared of, bad AI is, two different concepts entirely.”

“Aggro” is a very simple rule which leads to very exciting and, perhaps more importantly, easy to follow, and understand (and ultimately overcome), game play.

“I shuffle my formation to physically protect my healer ship, putting some metal in between him and them” See: one of the many MMOs which have unit collision detection allowing strategies like these to exist.

“Or another ship who wasn't acting as a healer in this encounter has to change mid-stream and take over healer duties, while the first healer has to change it's configuration and can now take damage.” See: multi-role classes in many MMOs, such as druids in WoW

“Or I use an attack skill which shifts around their formation so that, for a short time, their damaging attack craft are way at the back, and their support craft are at the front, turning the tide of the battle until they can get back into formation” See: MMOs such as Atlantica Online which uses formation as an important part of game play, and let users modify their, and enemy, formations

“Or on a smaller scale I shunt their strongest attacker out of range and tether him there, buying some time.” See: Any game which allows more freeform movement of player and enemy units during fights, and prevention of movement during battle, for example, City of Heroes.

And so on – you can find many more examples of these in my book.

Timothy Ryan
profile image
Please excuse me while I put my troll hat on: This blog reads like a long-winded rant that should have been posted as comments on other blogs rather than compiled as a blog entry. It's very difficult to get anything of value out of it when all the threads of the conversation are jumbled together. I also find highly suspect any authority on a subject who uses his own lectures and books as a reference. It's academic chicanery bordering on narcissism.

Dave Mark
profile image
Interesting... I thought I was writing an opinion piece (as most of these "Expert Blogs" are - including yours, I believe). Apparently some were expecting a doctoral thesis. Even so, if I was trying to write some academic paper on the subject, I wonder who I would cite as a reference on the subject of "should AI be better in MMOs". I'm not familiar with definitive research on the subject. *shrug*

Regardless, I didn't realize you needed any "credentials" in our business. Methinks that you may be challenging your so-labeled "academic chicanery" with what appears to be "academic snobbery." Considering that I am not even remotely an "academic" either, this amuses me even more.

And as for referencing my own lecture (which I will not deliver for another 9 days) , I suppose that if I had simply posted the entire thing here, I could have saved the airfare and hotel to Austin, eh? But that wouldn't make the CMP/Think folks very pleased. Oh well.

As for the rest of the comments above, I appreciate the conversation. After all, I did title this discussion a "controversy", did I not?

See you next week, Tim. I'll buy you a beer.

Owain abArawn
profile image
From the threads Dave Mark cited from last week, I don't think the chief complaint was that agro was bad in and of itself. As I pointed out last week, it is the combination of weak AI and poor class design that is at fault.

In attacking a group, it would be good AI for a mob to go after the healer, but it should be bad AI for it to ignore the Warrior standing between it and the healer, given the example cited above. Since Warriors are badly implemented however, there is little penalty associated with ignoring the warrior, and no collision detection to physically block the mob, so the Warrior has to employ an artificial Taunt to make up his lack of realistic effectiveness. If a player really wants to be an ace Tank, the warrior should devote skill points, talents, whatever, to make that taunt more effective, but in games that feature both PvP and PvE, the warrior gimps himself for PvP if he optomizes his built for PvE.

Here lies a stark example that highlights the dual failures of mob AI and PvE class design. Ideally, players should use the same builds and the same tactics for PvE that they do for PvP. It should be a design goal for developers to construct their AI for PvE to mimic the variability, the flexibility, and the unpredictablility of PvP. If that were possible, there would be no need for 100,000 hit point bosses, because mobs in general would be quite challenging enough, thank you. And there should be NO skills or moves that are used for PvE only (Taunt being the most glaring example). If Taunt doesn't work on other players, it shouldn't work on a Mob, and no other example better demonstates the failure of MMO AI as currently implemented.

This is admittedly an unattainable goal, but I think game designers could do much better, because the current AI hasn't changed much in MMOs for a decade, as previously mentioned. Instead of more eye candy, MMO games would be better served by devoting more cycles to to the AI, and less to shadows, particle effects, mip mapped textures, and so forth. When I can play a game with a character that can function equally well in both PvP and PvE, using the same tactics and gear for both, that will be REAL progress.

And Timothy Ryan? Man, what a kill joy! Lay off my new best friend, Dave Mark. I think he is right on the money in the points he has raised.

For my part, I base my observations on the study of military history, and the evolution of stategy, tactics, and warfare down the centuries, from the Pellopenesian Wars to the Mobile Infantry of Heinlein's Starship Troopers. Part of that study began as a Cadet at the Air Force Academy, and was furthered as a F-4 Phantom fighter pilot in the US Air Force. Air Combat in paticular is an excellent example of how strategy and tactics shift and evolve from second to second. There is nothing canned or predictable about it, nor should there be!

I look at computer games in general, and MMOs in particular, and they really should be much better than they are given the revolution in computing power of recent years. I can only hope that at some point, developers will be more willing to 'push the envelope', and cease being content to maintain the status quo. We have enough WoW lookalikes, thank you. Let's advance the art, shall we?

Owain abArawn
profile image
Oh, and in answer to Dave's question, "Quick question to Owain abArawn regarding the following comment: ... did you read my book already?"

No, but what book is that? I think I may need to check it out, but I don't think I saw a title mentioned in your post. Do you have a link?

My observation is merely common sense applied to game play. Alas, common sense is frequently lacking in MMO mob AI.

Dave Mark
profile image
The book, "Behavioral Mathematics for Game AI", is actually linked from my blog profile.

And you hit it right on the head about "common sense applied to game play." There is much we can learn - and subsequently design and program - by simply observing and quantifying how we as humans (or orcs or whatever) make decisions. What is important? How does that compare to the importance of other things?

Ted Southard
profile image
I think that Dave is making an extremely good point here: After a while, we need to step back and take a good look at what we're doing as opposed to what we can be doing, and effect change where it can be effected in order to bring out progress. Aggro is one of those things that we can progress...

There's nothing wrong with liking that particular style of MMO- everyone has tastes. But it's a bit disconcerting to see the story of the monkeys and the banana being played out here. We can do better with MMO AI, and one of the ideas that I myself like and have been doing a bit of work towards is similar to the Cultural Tactics model that Bart Stewart talks about in his blog on this site. And yes, you can say that aggro would still be a part of it in essence because it is a formula calculating just who should be attacked and when, but that is an argument regarding semantics due primarily to the fact that aggro as is used in MMOs today has not advanced as such, and so should be regarded in its present form because as a term, aggro is insufficient to explain how an opponent would come to that conclusion. Dave is also right in that AI should model not only correct behavior, but incorrect as well, as that would give the players a wider variety of behavior.

The defense that it would require a rebalancing of the classes is partially correct in my opinion. I don't think that it would completely break the system because players will find ways to use their classes to great effect, as they do now with all the spreadsheets they produce and the levelling guides that put the lie to the greatness of class/level methods, but on the other hand, the classes would still need to be changed at least slightly in order to give the players the ability to make some of the changes they will need to make when dealing with the smarter NPCs.

As Owain, I base my own opinions on the time I spent in the USMC back in the 90s and the tactics and thinking that are involved in ground warfare, as well as experience I have in travel (occasionally, I aggro creatures in real life). The things that I have noticed is that throughout history, humans have been able to deal with the unknown fairly well, from encountering new animals to the week-to-week advancement and changes to military tactics that occur during wartime, and MMOs do not offer that. What they do offer, or try to offer, is that kind of experience- The Hero Experience- but heros must be challenged, and they must experience failure. MMOs do not offer that.

That said, it's not impossible to accomplish a much better design using aggro and better rulesets. But at the same time, it is entirely feasible to create richer AI for MMOs which give the players both The Hero Experience as well as a challenge that keeps them on their toes. My own opinion on this is that once the better AI is in place, the next natural point of weakness will show itself to be an antiquated class system which pigeonholes players into being unable to change their own characters over time, but that's a subject for a different thread.

Anyway, I think we can all agree that aggro needs to be reevaluated and improved or replaced by something better.

Owain abArawn
profile image
Ted makes some excellent points. To pose the problem according to his experience, Marines in a combat situation use a form of agro as well. Your squad is currently egaged with enemy infantry, and you face not only enemy riflemen, but machine gun nests and indirect fire from a mortar emplacement as well. What are your priorities, how do you deploy your forces, and how do you direct your movement and your fire to counter the forces arrayed against you? How do you change your tactics and your plans when you discover that a T-72 main battle tank is now advancing on your position.

THAT is agro management, and the strategies and tactic you select to deal with the various threats you face determine whether you will survive, or not. The problem with agro as implemented in your standard MMO is that it is too simple and predictable. You can't try to destroy the mortar emplacement without taking both the enemy infantry or the machine gun nests into account, and everything has to change at a moment's notice with the advent of the T-72. You typically don't see that in MMO AI, which is how players can defeat insanely over powered bosses. The tanks taunt them, so they focus their attention on the player that poses the least threat (!!?%$#?!) to them, while the nukers are frying his liver, and the healers are spamming heals on the tank to keep in on his feet.

So what happens if the Tank falls? Typically a party wipe, which demonstrates that the Boss, if he had an effective AI that would ignore that silly taunt, would be FAR better off to bypass the Tank, and engage the real threats.

Would an advanced AI force a change to class balance? You bet it would, but that is in no way a bad thing, since the crucible of PvP, on MMOs that feature free for all PvP, already demonstrates that builds for PvE typically fail catastrophically when used in PvP. Defending current MMO AI is an exercise in defending the indefensible.

Ted Southard
profile image
"How do you change your tactics and your plans when you discover that a T-72 main battle tank is now advancing on your position."

Hmmm... Swarm it and hit the hatch with an Eagle Cocktail after disabling it with a huge unruly mass of field telephone cable to derail the treads (which they probably don't have these days with those new-fangled radios) would be my own choice. But then you run into problems like shown in Saving Private Ryan, where half the squad at the end was cut down while swarming the Nazi tank because the tactic makes you stand out like a sore thumb and the Nazis are probably pissed that you're trying to rain all over their parade...

But that last part is the problem, isn't it? Why would Nazis allow my squad to swarm the tank, even if there were a larger force available for them to shoot at? Tactical reasoning? Sure, there's that. But just like my squad, the main motivation for carrying on in a battle comes down to emotions. Sometimes, the only way out is through, and my squad is funneling its fear into aggression aimed at the tank, while the Nazis justly funnel their fear into getting us off the tank, because if we succeed, then they're likely to be very bad-smelling toast. It's not all about dispassionate tactical and strategic goals. People in fighting holes don't pull the trigger for God and country, but because they don't want themselves and their friends shot to pieces, and if that means that they need to accomplish the tactical and strategic goals, then all the better (because the other option of running away means desertion, which means courts-martial and possibly death).

What is needed for proper AI is a fast set of rules to figure out the emotions and character traits of the NPCs (since the player is good not only at his/her own, but also imitating their character's in context), and that means realizing, just as scientists have, that intelligence is related to the emotional spectrum of the organism, not so much the ability to problem solve (birds make tools too). It's not enough that AI simply choose the best course of action, but to do it in a way that reflects convincingly (which does not mean accurately) the NPCs state of emotions at the time.

Enraged people are not logical, and cold-blooded warlords recognize a taunt when they hear it. Frustrate the warlord with tactics that make his energy expenditure futile, however, and he becomes angry, opening him up to a taunt and defeat. Trickery and dialog become valid features of combat, and suddenly you have dozens of new avenues of approach in dealing with issues.

All of this is extremely feasible. But at the same time, it requires many changes to things that we may or may not be very used to.

Oh, so here is your fast-personality-trait AI, in blog form. I call it AE (artificial emotion), but whatever floats your boat: - Personality-based AI - NPC conversation systems (this is old and I'm well past that stuff, check the site for more information on Interrogative, which I developed for my MMO Epic Frontiers).

Jonathon Walsh
profile image
Interesting stuff, I didn't realize the Class discussion picked back up. I had abandoned it after people started talking about DnD rulesets. I'm also completely unsurprised to be called out for some of my points, I tend to get pretty bitter and pessimistic when talking about MMOs.

On the subject of AI and MMOs... How many people have tried Darkfall? I don't recommend playing it but it's interesting from an AI point of view. In Darkfall the AI tries to be more intelligent. It runs when wounded, gets help, swarms the player, and dodges side to side like a real player will, kites, and it's incredibly annoying and insanely frustrating (as a side note this could be said for most of Darkfall). The better AI does nothing but become a huge pain for players as they now have to try and corral the AI into a state where it can be defeated. Then repeat that exercise 1000s of times so the player can get their skills and bank accounts up. I think this represents the double edged sword of universally more PvP like AI in MMOs. At some point as players we want the AI to lose, and if we have to kill massive amounts of monsters we don't want to struggle with each and every kill.

What I think needs to happen is that we need to see a huge expansion of AI variety and goals in MMOs as well as expansion in game play. Some AI, especially sentient AI, should take a smart PvP like approach to encounters. They should call for help, prioritize threats, flank, etc. Other AI should probably play more like Diablo. Cutting through 1 monster in Diablo is rather dull, the AI is incredibly simplistic. However cutting down 30 of the same monsters repeatedly can be frantic, fun, and quite challenging. If MMOs also adopted this type of AI they could effectively combat the double edged sword of implementing AI that tries to hard to survive. Essentially we need to delve into what the NPC is there to do and provide appropriate AI for the situation. NPCs that are meant to be farmed but are wily and difficult to kill is as bad, if not worse, than challenging AI that is just pumped up with an overload of HP and Damage.

However AI changes aren't enough... Going back a bit to the class discussion I think MMOs will always have problems as long as they use hard counters for class roles. Right now most MMOs put people into very hard or strict roles as discussed with generally no flexibility in a group situation. They create the problem where if the AI is smart and somehow attacks that healer then it's very likely game over. If classes were more able to dynamically play around with their roles then you could have room for smarter AI. A perfect example comes across with classes like the Druid and Paladin from Wow. These classes are supposedly flexible, both can tank or Heal. The problem is it's an absolute OR situation and one that requires an investment of time to change. A Druid can heal OR he can tank, he can never swap effectively between roles during combat, likewise with the Paladin. If a Druid could tank then switch to a healing role in combat then you open a lot more in combat dynamics. Now it's OK for the tank to lose aggro if he's needed to perform healing duties.

Another example would be the rogue. Generally for rogues dodging is an ineffective way of taking blows. If you change it around though so that a rogue could effectively give up combat damage to become very focused on dodging then suddenly it's 'ok' for him to tank monsters at times. As it is now if the rogue is being attacked everything is going horribly wrong, but if the rogue can switch roles then suddenly it's a new dynamic to play with. You keep the same class roles but you open up the possibility for people to shift roles allowing for AI to be interesting rather than a pain. At the same time you keep the roles exclusive but flexible in combat, so a warrior is either DPSing or defending but not both at the same time, and limit gear effectiveness so players can remain effective at both roles. This way you open up room for the AI to be truly dynamic without causing players to resent your game.

Owain abArawn
profile image
Jonathon, speaking of flexibility for classes, Darkfall actually does well in this regard, since it has no classes, and uses skills instead, more like Ultima Online than WoW in this regard. Running mobs in Darkfall became trivially easy to manage once I picked up archery skills, which made running a very bad choice for the mob since I would just pull a bow and nail them rather than chasing them down. And if I ran out of arrows, I could pull a mage staff and magic missile it instead. Try THAT with a warrior in WoW.

With respect to classes, none of this is exactly heresy, either. Consider the flexibility inherent in the archetype for fantasy role playing games, The Lord of the Rings. Gandalf is a wizard, yet at times he is an accomplished swordsman, and also uses his staff as a quarterstaff in melee combat as well. While Aragorn is primarily a swordsman, he is also adept with the bow. He typically wears no armor as a Ranger, but at Helm's Deep, he wore chainmail, and in the assault on the Black Gates of Mordor, he wore plate. Legolas is an archer, but also duel wields Elven swords. Circumstances and skill sets are the driver, not a rigid class system.

Obviously, then LOTR, at least in the books and movies, isn't bound to the class model, but adopts the skill based approach of Ultima Online instead.

"NPCs that are meant to be farmed but are wily and difficult to kill is as bad, if not worse, than challenging AI that is just pumped up with an overload of HP and Damage."

NPCs that are only in the game in order to be farmed are part of the problem. How far removed from the Hero Experience, as Ted Southard says, can you be when with each mob you kill, you are grubbing for gold and other 'ph4t l3wt'? Leonides didn't lead the 300 to Thermopile because the Persians had a good drop rate!

Every fight should be a challenge, otherwise everything degrades into the mindless grind that people complain about, but as seen in comments in this thread, some go to great lengths to defend. How dare you seek to deprive me of my simplistic predictable walking little loot boxes!

I think that some folks must be falling entirely too far into a state of complacent comfort with the current crop of MMOs when people start talking about player resentment in response to the suggestion of developing realistic effective AI.

Jonathon Walsh
profile image
Yeah, I didn't really get around to that but you do make a great point about how a skill based system can open up flexibility for classes. Games like UO also seemed to offer more flexibility as well.

I am surprised that you found it easy to nail monsters in the back though, I found it rather frustrating to do so in Darkfall. The mobs would run in very erratic and hard to predict patterns making hitting them difficult. That combined with the challenge of staying stocked with arrows made it even worse.

Also yes, farming mobs is a big MMO problem, but if that's not going away then bringing in smart AI isn't going to help any either. I'd love to see an MMO that focused more on the journey and other epic encounters than farming mobs. I just think smart AI can make farming mobs even more of a grinding chore than dumb AI. So if you keep farming mobs but fix mob AI then overall your game is going to suffer. You need to solve both, neither, or at least make the mobs that have to be farmed dumb. Darkfall was the game that proved this to me, it was agonizing hacking away at goblins and other mobs where it took 3 times as long to kill them as it should have because of their AI yet you needed to kill just as many as you would in a game like WoW.

Ted Southard
profile image
"The mobs would run in very erratic and hard to predict patterns making hitting them difficult."

Does Darkfall use FPS-like aiming? If so, that opens up lag issues, which get exacerbated by an MMO architecture.

"I'd love to see an MMO that focused more on the journey and other epic encounters than farming mobs."

AI isn't just for NPCs and MOBs. Remember that when adventurers enter the Wild, the environment itself becomes an adversary. Luckily, the amount of AI needed to accomplish this is rather minimal. Going to the next level, you can leverage RTS AI in an MMO to control issues such as the kinds of quests given out to players in an area, the activity level of bandits or enemy NPCs, etc.

Christopher Wragg
profile image
@Jonathon Walsh

While in essence I agree with what your saying here, there are fundamental balance issues with just allowing WoW characters access to all aspects of their class. For instance, Druids, Paladins and *maybe* shamans, would be all that ends up in a raid. While there would be far more interesting combat, if druids could switch between 3 things their class does well, damage(ranged and melee), heals and tanking, there isn't much need for anyone to play anything other than a Druid. It, in essence, becomes too flexible. there has to be a limitation of some sort, after all it's limitations that breed creativity.

I feel like a broken record but I love the way Guild Wars manages skill and class roles. Considering the way you can multi-class, and considering there is no real restriction on weapons, or abilities (though there is on armour) players can build any one character to meet virtually any challenge. But, considering one can only carry 8 skills onto the field at a time, players must carefully select what their character does and how they will do it. Want a warrior who can switch to healing others every so often, sure you can do it, but at the penalty of not being able to do either 100% effectively. How about a healer who can tank, again doable. Want to build a party of players designed around crippling your opponents abilities and not bring a healer at all (instead everyone manages themselves), again doable.

Once this sort of flexibility exists, and there are numerous counters to, well, anything. Implementing clever AI is merely the next logical step, because both AI and players have the abilities to meet the challenge it brings. It also allows players to bring their own flair to the game, why be a plain old warrior when you can play a necromancer who amplifies his damage with curses, then uses a sword and various vampiric abilities to heal himself during combat.

Bart Stewart
profile image
Time to jump back into the fire, now that I'm finally home from Dragon*Con and have access to an actual computer again:

1. The problem is not aggro. Aggro is just a simple mechanic. I believe the true problem is twofold: A) The mechanic of aggro as a simple target selection subroutine has been expanded beyond all reason into "aggro management" -- magically dictating the attention of a hostile NPC. But that play bears no resemblance to what people who know combat (like the folks posting earlier in this thread) understand as "tactical" behavior. B) Developers of new MMORPGs appear to be blindly copying this aggro management-based model of combat into their games instead of consciously developing a tactical combat model that is specifically keyed to and enhances the rest of the game they're making.

2. So getting rid of "aggro" (defined simply as target selection) is probably not necessary. But getting at least a few MMORPG developers to discard "aggro management" in favor of some other model of NPC behavior selection may be crucial to the long-term viability of the MMORPG game segment. In that case, on what basis should developers design a decision-making model for opponent NPCs?

3. I suggest that a good place to start would be to consider gamer psychology. What is it that gamers want? Here's one theory: Some gamers want to beat the game, and expect to be able to beat every game. Some gamers don't mind winning but take most of their pleasure from improving their understanding of the rules of the game. Some gamers don't care much at all about following the formal rules of the game, but show up either to tell stories with friends or to manipulate the gameworld and/or other players (meta-gaming). In brief: players want games that deliver fun through Persistence, Perception, Persuasion, or Power.

4. Early RPGs were played by people who enjoyed solving puzzles and acting out roles. They cared about Perception and Persuasion, and the features of the games they played reflected those interests. Today, those people have for the most part been discarded by developers making MMORPGs almost exclusively for Persistence-oriented gamers. Today's typical MMORPG players expect to be able to win by demonstrating that they refuse to give up. They don't want to have to think or feel; such things just get in the way of proving who's the best at rule-following. I've more than once observed Persistence-minded players energetically trying to define storytelling (persuasion through emotional understanding) and puzzle-solving (perception through intellectual understanding) as "not fun," and warning developers against wasting any time on such stuff. And the folks making new MMORPGs appear to be happy to do as they're told.

5. But if the above is a reasonable way to perceive player motivations, the true lesson is that there is no one "best" NPC AI on an absolute scale -- there's only which NPC AI is likely to be most satisfying relative to the kinds of gamers you want your game to attract. Persistence-minded gamers, whom I believe now make up the majority of MMORPG players, don't want a "tactical combat gameplay experience" -- they want a game they can beat by throwing themselves repeatedly at it. They like the aggro management game precisely because it bears no resemblance to tactical action in which rapid and intelligent adaptability to surprising changes in the environment is paramount. What they want is a system of rules that will allow them to win. And they like the aggro management game because it reliably gives them the victory through sheer bloody-minded persistence that is their preferred form of interactive entertainment.

6. There's nothing inherently wrong with Persistence as an innate preference for how to enjoy a game. And there are a lot of such gamers. It would be commercially foolish to deliberately design a game that they couldn't enjoy.

7. On the other hand, there's nothing inherently wrong with Persuasion or Perception as innate preferences for how to play a game, either. A combat model that those gamers can also enjoy should be more attractive to more paying customers than a purely Persistence-based model. Why drive away a meaningful number of potential customers if you don't have to?


All of the above is why I like the "cultural tactics" model I proposed in my own blog, and part of which was mentioned in passing in Dave Mark's original blog post. In brief: a cultural tactics model of combat would define numerous cultures as having one or two archetypal attributes (aggressiveness, cowardice, honor, sneakiness, etc.). Any NPC (whether persistent or randomly generated) would belong to a culture, and would inherit the archetypal attributes of that parent culture. In a combat situation, NPCs would attempt to behave according to rules of action that are associated with each cultural attribute -- an NPC from an aggressive culture would tend to attack the nearest enemy; an NPC from a sneaky culture would tend to use abilities, gear, and environmental elements that give them the element of surprise; and so on.

This model of combat decision-making allows NPCs to appear to make choices (including but not limited to targeting) that are perceived as intelligent because those choices seem to fit what players know about the culture to which the NPC belongs. That (I think) should create predictability satisfying to the Persistence-oriented gamers in that when you know the culture, you have a pretty good idea how an NPC from that culture will behave. At the same time, a cultural tactics model of combat also permits some surprising behavior (to satisfy the gamers who want their Perception to be rewarded) because the specific environmental conditions where each combat occurs will vary, and because some cultures will be able to recognize and exploit features of their local environment. Finally, culturally-guided NPC behavior helps to tell a story (to satisfy the gamers who want opportunities to display their abilities of narrative-based Persuasion).

In summary, I think aggro management will eventually prove to have been an evolutionary dead end, supplanted by a combat model that satisfies not just Persistence but Perception and Persuasion as well. As a starting point for finding that superior system for combat decision-making by NPCs, I offer "cultural tactics," but I expect there are even better ways to model PvE combat in MMORPGs. I hope developers will consider those alternatives.

Dave Mark
profile image

That was absolutely well-said. You, sir, get a free pass to my lecture next week. (The GDC Austin pass, travel, and hotel are up to you.) In fact, I want your butt sitting by the podium next to me.

The "cultural tactics" thing is a good idea and is actually just an extension of modeling personality, mood, and even emotion in game characters. This was the subject of a lecture I did at the AI Summit at GDC with Richard Evans ("Black & White" and "Sims 3") and Phil Carlisle (University of Bolton). We pointed out that a properly designed model can either statically or dynamically reflect many different types of characters.

In your case, modeling subtleties in tactical or behavioral approaches to combat is really not difficult at all. It would lend many shades of difference to the various types of entities - even providing differences between agents of the same type (for variation). However, each of these different types can be distinct... i.e. "this is how ABCs attack/defend and this is how DEFs attack/defend." The fun part is, much of this can be built on the same underlying data-driven combat model. (read: "designer tunable")

Anyway, I appreciate all the commentary, folks.

Owain abArawn
profile image
I agree with much of what Bart says, particularly with respect to his 'cultural' approach, but even that should not be a static attribute among the various types of mobs in a game. The degree of agressiveness among Orcs, for example, could shift during a single engagement, or over an extended campaign as dictated by the current state of their morale, for example. If they perceive they are gaining the upper hand in a fight, that might push them into berzerker frenzy. Otherwise, if they are getting their butts handed to them, their will to fight might waver and break. That is just in a single fight. I could see this same dynamic take place over a larger scope. If they are being defeated regularly on a larger scale, their culture might be modified so as to make them more wary. Pushed far enough, they might go into 'cornered rat' mode and respond with insane desperation. Alternatively, if they experience wide spread success, that might influence their global 'culture' as well, triggering wave upon wave of attacks for conquest and domination.

This would fit well with the aspects of combat that Ted Southard discusses, "...the main motivation for carrying on in a battle comes down to emotions. Sometimes, the only way out is through, and my squad is funneling its fear into aggression..."

This would result in much more complex behavior from mobs, and how much more rewarding from a game play point of view would this be? If you want a simple encounter that permits you to auto-attack a mob while playing solitaire in another window, maybe more complex interactions are not your cup of tea, as with Bart's example of "...typical MMORPG players expect to be able to win by demonstrating that they refuse to give up. They don't want to have to think or feel; such things just get in the way of proving who's the best at rule-following."

If that is all you know, that is what you come to expect. But when I look at the popularity of a game like Left 4 Dead that features more complex AI modes and interactions, I can't help but feel that there is an untapped market there waiting to be served. Perhaps because I come from a PvP-centric orientation, it seems that most of the players I know, particularly those in my current guild, would all appreciate greatly an MMO that replaces the same old grind with an AI that approaches the challenges offered by PvP.

Jonathon Walsh
profile image
Darkfall does feature FPS like Aiming, the lag on the abilities generally isn't too bad but the projectiles are somewhat slow moving (spells) or arc (arrows). I enjoy the extra interactivity in combat, and it works well for PvP but when you're just trying to kill some mobs for spare cash or skill ups it becomes a much bigger chore than it should be.

I certainly didn't mean to suggest that changes wouldn't be needed to open up more flexible roles for a game like WoW. It's definitely something that has to be done from the ground up. Don't think that hybrids are the only classes that can fit multiple roles though. Even 'simple' classes like warrior and priest can fit two roles in WoW through speccing and gear. I think Guild Wars is a great example of edging in on more flexible classes. It does still have the problem of limiting your role in combat preventing a more dynamic group play, but it's still way more flexible than WoW.

The cultural tactics model is definitely something I'd like to see in action. A great post all around, I think you really nailed the heart of the issue with AI in MMOs.

Kevin Dill
profile image
I did a writeup once upon a time about using positioning in the place of a mechanic like agro. The idea was to build real advantages into the game mechanics that would work the same whether your opponent was a mob or a hostile player. So the tank would physically place himself between the mage/healer and the enemy. Maybe he'd have the ability to deliver a massive stun if you turn your back on him - pull you around and throw you on the ground in front of him or something. Maybe the mage would have the ability to knock the mob back 10 feet. Maybe the rogue has to actually sneak around behind enemy lines, and has the ability to deliver a massive attack if he can pull it off - but it's a risky move because he's out and exposed with enemies all around. Maybe at higher levels the priest can elect a temporary barrier around him letting him get back to safety.

Obviously, lots of work to do to flesh something like that out - but it could lead to some really fun gameplay. And as I suggested, one really nice thing would be that the PVP tactics would be at least similar to the PVE tactics if it were done correctly - that is, mobs would act the way they did because it was the smart thing to do, not because of an arbitrary mechanic that didn't apply to players.

Ted Southard
profile image
@Kevin: Positioning is a great tactical feature. If you walk into one of the many little natural valleys in WoW, you'll find NPCs (MOBs) positioned all over, including up on some high-ground spots, but if you initiate combat, they close distance. If the NPCs had an awareness of positioning and their own group's strength, the ranged NPCs would take high-ground while the melee fighters kite the player into the valley and surround and pin the player down until they get annihilated from the ranged bombardment.

Conversely, players would send in combined forces with Rogues acting as stealth units to hit ranged NPCs while warriors hold the middle and more mobile classes flank and envelope. This is what makes a base a base: A single player can't just walk in and pop a few NPCs for quick money. But because the AI is so lacking, they've created the Raid system with overpowered bosses to get that feel, or the player just winds up aggroing that half of the planet all at once upon entering the zone.

Procedural content generation can also be used with spawnpoints hidden in buildings (similar to those mobile spawnpoints disguised as hives in SWG) to generate an "appropriate" response- if the NPCs are capable of such, which is an important point- in effect, giving the players a real good battle on their hands. I don't know any player worth his salt not willing to give something like that a shot over repetitively beating MOBs over the head like baby seals for loot.

Dave Mark
profile image
re: positioning, spawning, etc.

Fun with influence maps! You will see what I mean at the lecture next week, Ted. (Complete with nifty programmer art and Powerpoint animations!)

Maxime Doucet
profile image
I'm still going through all the feedback on the original post but you hit the nail on the head with this:

"... raid groups and 'Guilds' are not social situations in the MMO's today. There is no difference, typically, in these social interactions and those you would find in your typical IM conversation or Facebook thread." The reason for this is largely because players do not have to communicate much game-related information to each other in MMOs these days. Again, contrast that to online PvP shooters such as TF2 where communication of relevant information is crucial to survival as an individual and success as a team."

With advanced AI, lesser use of Aggro and various other changes in the battle dynamics - such as replacing classes with a skill-based character progression, damage tables instead of hit points, "cultural targeting" as proposed by Mr. Stewart and various other devices - the gameplay would be enriched to the point where the players, as small groups or as whole communities, would have to socialize and strategize in order to take down a much more complex, organic, unpredictable and dynamic AI system.

John Hahn
profile image
@Bart Stewart "Persistence-minded gamers, whom I believe now make up the majority of MMORPG players, don't want a "tactical combat gameplay experience" -- they want a game they can beat by throwing themselves repeatedly at it. They like the aggro management game precisely because it bears no resemblance to tactical action in which rapid and intelligent adaptability to surprising changes in the environment is paramount. What they want is a system of rules that will allow them to win. And they like the aggro management game because it reliably gives them the victory through sheer bloody-minded persistence that is their preferred form of interactive entertainment."

I agree with this assessment. I have a large group of friends that all play WoW. They have their own little guild and they hang out on their private ventrilo server all the time and go raiding and stuff. Years ago, before any of them starting playing WoW, they all used to be regulars at a local comic book/ccg shop and they all played Magic: The Gathering, other card games, miniatures games, etc. These guys love learning about the rules and tactics of winning a game and using those tactics to win based on a clear set of defined rules. In the card game world, the randomness comes from shuffling the cards. Each individual card is static and never changes. There's no dynamics or emotions in the card itself. The dynamic comes from what cards the player puts in their deck before the game and what order they play the cards they happened to have drawn. However, the rules of the game are clear cut and unchanging, which is completely unlike a real war or real battle, much like chess or a board game.

Coming from that kind of mathematics driven card gaming, it's easy to understand why this kind of player would want a game where everything is driven by rules and mathematics like "aggro management". In that kind of game, you, as a player, can totally plan out a raid boss fight mathematically if you want to take the time to build the spreadsheet. The boss has x-hitpoints and does y-damage per second. We have z armor with b protection and c weapon with d attack damage per second. Mathematically, we can win if we use this tactic. Some people like that kind of analytic strategy and thinking, particularly people who come from card games, miniatures games, board games, or chess. Other people want a more fluid and visceral kind of game experience like you get from TF2 or CoD4.

Some select few single player games have managed to capture both the analytic and organic seamlessly such as Fallout 3, but this is rare, IMHO.

Emory Myers
profile image
"Theoretically, you could do an audition for an animated version of "So You Think You Can Dance" - but that won't bring you any closer to your goal of winning the battle."

Don't be so sure -