Break the Diku! (Panties)

[Nonentity]

Panties! Yeah, I'm done. I just wanted to write panties at the beginning. If I'm an attention grabbing whore.. okay, well I am.

So, Diku-style mechanics state that there are a couple of core types of classes. Let us list them out so we can think of how to break it. Why? Because I want to break it. I want to find all the Night Elf Rogues named 'Solidsnake' and 'Staburazz', throw them
in a pit, and burn them.

I do like classes. I know of the whole dynamic skill system vs. class discussion, and that's a whole other can of worms. This is just under the assumption of a class based system.

Yeah, I realize this thread has more then likely be done millions of times, but I want to see what the brilliant (heh) minds of F13 have to say on the matter. All things considered, I'm relatively new here, and I'm genuinely interested.

So without further ado, those are:

The Holy Trinity:
- Tanks - Meat shields, who piss off mobs so they get beat on. Heavy mechanics to manipulate the aggro components of the Diku system. The 'tank' part of 'Tank and spank'.
- DPS - Generally the non-meat shields, who are the 'spank' part of 'tank and spank'. Get in close and use a dagger/sword/lazer stick/whatever to poke shit. Alternatively, stand far away and use guns/arrows/spells/rape beams. Usually has abilities to lower their aggro on a given mob.
- Healer - People who are suckers for pain. Like to stand back and watch little green bars on a UI slide around while pushing buttons. Gotta keep people from dying, after all.

Then there are the few that have been added to the formula (sometimes rolled into the above classes).

- Crowd control - Whoops, we pulled two baby dargons! Gotta keep things under control. That's what these guys do. They dance around, fear stuff/stun stuff/mesmerize stuff.
- Buffers - Like to install a series of complicated UI mods to keep track of half hour long buffs, hour long buffs, 10 minute long buffs, and so on. Gotta keep the raid buffed to spank the undead dragon lich kings, right?
- Pullers - As erotic as they sound, they generally use a ranged weapon to pull apart roving bands of giants, and have some sort of 'feign death' mechanic to keep the raid from dying, and having a plethora of hate tells sent to them.

So, that's what we have. Every class in every Diku-centric game has been based on a combination of the above. Taking into consideration that we can change the core mechanics of how a Diku works (please, please do this), what can be done differently? I mean, there are various levels of abstraction (I'm looking at you, Blood Healer from Vanguard), but it all comes down to the basic CORE things.

Is it really a consideration that the Diku has to be broken into an unrecognizable form for this system to be changed? Yes, I completely think so. Do something completely different.

But at the same time, you want a level of familiarity. Wasn't there an MMO that was also a screensaver, where you played like, a deer or a buck walking around a mystical forest, doing emotes at each other? Fuck. There's abstraction, and then there's asanine.

The system can't just be changed into a sci-fi setting so the things don't apply. I mean, shit, look at the info from Tabula Rasa that's been coming out recently. It's supposed to be this big grand sci-fi pseudo-shooter, but if you look at the class trees available on the website, it's just "Hey, the Commando! These guys absorb all the damage, while using rocket launchers! And hey, the Biotechnician! It's a player with basic healing abilities, and a poison injector!" Jesus.

I think the basic considerations about the game have to be totally different.

So, lets say it's a game where PvP conflict is the endgame. Yeah, there are plenty of high-level systems and discussions for how this should work, but lets think of core concepts here.

Do you PvE to get to the endgame? Do you have 'aggro'? Are there a series of classes that are really good at doing one thing? You're gonna get a Diku if you start traveling along that mindset.

First off, what basic considerations need to be made in order to change it up?

[bhodi]

This should (panties) still be in (panties) game design/developmet/(panties)

[Nonentity]

This should (panties) still be in (panties) game design/developmet/(panties)

Hm... You're right.

Rookie move.

[ajax34i]

So, is this an exposition of your ideas, or is it a call for developers to make a change (and in the process, come here and tell you how they would change things)?  Or, is it one of those "dikus need change, discuss" posts?

[Nonentity]

So, is this an exposition of your ideas, or is it a call for developers to make a change (and in the process, come here and tell you how they would change things)?  Or, is it one of those "dikus need change, discuss" posts?

More along the lines of option 2. If a developer wants to post, be my guest, but that's not the point of this.

I just know there are people with strong opinions on the topic, and I'm interested in what those people have to say. I've read a few ideas, but I'm just trying to get a broader idea.

[Slayerik]

First off, it would need to be a skill/item based system.

If you equip *tanking* armor, you are the tank for today. If you equip your lightweight magic leather armor you are the kiter guy. As long as you have spent enough time working those skills they are available to you.

I used to like the UO method of forced 700 point skill cap, but now I just kinda look at EvE and think if properly implimented there is no need for a cap. You have skills, but you cant use them all unless you have the right stuff equipped. Staff and robes for magic. Or maybe you can use lvl 1-3 spells in Plate, level 1-6 spells in leather, and level 1-10 in robes. Have CC abilities be level 7+. Have very limited bag space you you cant just fully swap your gear whenever you want. Or dont limit it and have PKs make you wary of doing that anyways :)

Your skill with swords opens up special attacks, and damage modification. If you are using a shield, it enables a taunt ability if your shield level is high enough.

Eventually, one character could be a mage in one encounter, train his skills towards that for a long time. Once boredom sets in, he decides its time to go for medium armor and bows and dual wielding.Eventually, he masters that. Depending on his gear set he can do either in a dungeon crawl or PVP.

No classes, just items and skills. And PVP. Lots of PVP.

Expanding on this:

1H Mace class - Opens unparryable Bash attack(can be dodged or miss), all maces causes stamina damage
1H Sword class- Opens undodgable Slash attack (can be blocked or miss), Opens some high damage thrust attack

Bow class - Aimed shot, Rapid fire attack (accuracy at the expense of 2x firing speed)
Xbow - Aimed shot, heavier damage than bow
Heavy Xbow - Aimed shot, goes Thunk

A skill tree would be pretty standard

Armor -
   Light Armor - Anyone can wear

   Medium Armor - Specialize for fast movement (dodging), less dex modifer, etc
   Heavy Armor - Specialize for faster movement (dodging), less dex modifer, etc...maybe even specialize as far as specific types?

I dont know, I think you all get the drift. Some deeper skills would have a level pre-req....lets say Heavy Armor requires Medium Armor 3 or something.


Throw in slow offline training and you have Eve. But I say include a slow active skill gain.

[Nonentity]

First off, it would need to be a skill/item based system.

No!

Okay, maybe. I do know the merits of both class and skill based systems, and I like both. But for the sake of tactical transparency and this argument (you target a guy, you have SOME idea what he's gonna do based on class), I picked a class-based system. An item-based system might be interesting, or some sort of a system where you are given a class based on what skill and itemset you bring into the fray.

If you equip *tanking* armor, you are the tank for today. If you equip your lightweight magic leather armor you are the kiter guy. As long as you have spent enough time working those skills they are available to you.

True. Using what I mentioned above (where an item/skill build nets you a 'class' type of title), you could work something out. It might be worth also having some sort of a system where you can bring a few skills from another 'class' into play, just to have that small concept that you don't know what they're going to do.

For example, you throw on your platemail of asskicking, a big axe, and a shield. You have all sorts of damage blocking and stunning abilities. It gives you the 'Tanky Guy' class title. But, you also threw 3 to 4 other abilities on your bar, like a fireball, or an Cone AoE knockback. Those would be unknown to whoever is fighting you. Sure, you know basically what he's going to be doing based on his class, but he has those few things that he can do that throw you a curveball.

I used to like the UO method of forced 700 point skill cap, but now I just kinda look at EvE and think if properly implimented there is no need for a cap. You have skills, but you cant use them all unless you have the right stuff equipped. Staff and robes for magic. Or maybe you can use lvl 1-3 spells in Plate, level 1-6 spells in leather, and level 1-10 in robes. Have CC abilities be level 7+. Have very limited bag space you you cant just fully swap your gear whenever you want. Or dont limit it and have PKs make you wary of doing that anyways :)

From what I've read about Fury, they're doing a similar thing. You can eventually learn everything, and then in 'safe areas', you swap out equipment and skills to get a 'build'. That still doesn't indicate what kind of tactical transparency you'll have, though. I mean, in UO, if you saw a guy running around in plate, you had no idea what he was gonna do. He could be a platemage, or pull a poisoned katana out, or whatever.

The reason I say classes is that it's easier to do full changes to one thing, without affecting a variety of people with the build. But as I said, that's another discussion entirely, heh.

Your skill with swords opens up special attacks, and damage modification. If you are using a shield, it enables a taunt ability if your shield level is high enough.

Eventually, one character could be a mage in one encounter, train his skills towards that for a long time. Once boredom sets in, he decides its time to go for medium armor and bows and dual wielding.Eventually, he masters that. Depending on his gear set he can do either in a dungeon crawl or PVP.

Having a variety of skills you can train up alleviates the problem slightly, but that still comes down to the fact that you're just a bowmage. Still similar mechanics. Dungeon crawling or PvP, it's still falling into the same trappings you're caught into.

No classes, just items and skills. And PVP. Lots of PVP.

At least we're on the same page as far as PvP is concerned. :>

Ultimately, we need to get away from fantasy for a change, but even as CoH/CoV shows, they still know how to put the tankmageheal into a superhero game.

But hey, this thread can dual wield. I guess. Class vs. skill and Diku tankmageheal discussion. Yay!

[Merusk]

As long as there is character-healing, there will be a "tanker" and a "healer" class.  People are stupid in this regard.  Add-in a character that is 'balanced' because it trades HPs for Damage, and you begin the trinity anew.

First step in breaking DIKU is getting away from the D20 or D&D mindset.

Why does a character have to have more HP than another?  Why do damager-dealers have to have crappy defenses?  Why do magic-users have to be shitty at melee combat? "THEY JUST DO!!" is the answer because of this legacy.  Getting away from it seems to be impossible for most designers, as they grew-up in D&D's heyday, so its pretty engrained in their psyche.

The very idea that a damage-dealer can also be a big burly guy is an 'horrible, stupid notion.'  Seems rather at odds with the guy checking your ID at the bar, yes?  We don't hire skinny, spry folks to do that job because, "Dude, he'll fuck someone up, just look at how scrawny he is!"  So why do the same shit in games?

Next question: Why are there HPs in the first place?  2000 hp or 20, a stab to the right part of the neck should be fatal.  I don't care how big and burly and how much "stamina" you have;  If I pierce your jugular, you're going down in seconds.

   HPs are an artifact of D&D, but in D&D you only had - MAYBE 50 or 60 of them. They were meant to keep things simple for the GM and to keep the game flowing smoothly.  When D&D moved to computers, it was a direct translation of the rules.  When it moved to MUDs, it was another direct translation of those game rules.  Nowadays they are an abstraction taken to an absurd level, simply because players have been taught to OOh and AHHH over big numbers.

Computers & processing have come a long way from simple number crunching. We can do absurly complex calculations on the fly, interpreted from data given by a few key presses or the gesture of a hand.. however we as consumers seem to insist that things in games must only be addition and subtraction.  Even shooters started out with this silly formula - and some still retain it.

There's a start, something to chew on for now, but it's farther than you'll see games evolve away from DIKU for a good, long while.

[Nonentity]

As long as there is character-healing, there will be a "tanker" and a "healer" class.  People are stupid in this regard.  Add-in a character that is 'balanced' because it trades HPs for Damage, and you begin the trinity anew.

First step in breaking DIKU is getting away from the D20 or D&D mindset.

Why does a character have to have more HP than another?  Why do damager-dealers have to have crappy defenses?  Why do magic-users have to be shitty at melee combat? "THEY JUST DO!!" is the answer because of this legacy.  Getting away from it seems to be impossible for most designers, as they grew-up in D&D's heyday, so its pretty engrained in their psyche.

The very idea that a damage-dealer can also be a big burly guy is an 'horrible, stupid notion.'  Seems rather at odds with the guy checking your ID at the bar, yes?  We don't hire skinny, spry folks to do that job because, "Dude, he'll fuck someone up, just look at how scrawny he is!"  So why do the same shit in games?

Completely true. A step in the right direction would be to throw out the idea of a 'tanker' or 'healer' altogether.

Every class has the capacity to do a lot of damage in different ways. Every class also has some way that they can heal themselves. (For example, take a look at some of the screenshots people have been taking of abilities from Warhammer Online from Gencon 2006 - One of the Dwarven Engineer Morale abilities is 'Bite the Bullet' - based on your morale, it heals health. That is cool.)

So, from that idea, you have the base concept that everyone is very lethal at what they do, and it starts looking up for a PvP game.

I can hear the cries now, though. "But Non, there's no tanker or healer! You can't have group strategies! FORMATIONS!11!" Yeah, well, spread those abilities around the classes. Don't give them all to one and call it a day. Give any given class a level of odd flexibility, and then multiply that by a ragtag group of people, and you have an interesting PvP encounter.

Next question: Why are there HPs in the first place?  2000 hp or 20, a stab to the right part of the neck should be fatal.  I don't care how big and burly and how much "stamina" you have;  If I pierce your jugular, you're going down in seconds.

HPs are an artifact of D&D, but in D&D you only had - MAYBE 50 or 60 of them. They were meant to keep things simple for the GM and to keep the game flowing smoothly.  When D&D moved to computers, it was a direct translation of the rules.  When it moved to MUDs, it was another direct translation of those game rules.  Nowadays they are an abstraction taken to an absurd level, simply because players have been taught to OOh and AHHH over big numbers.

Computers & processing have come a long way from simple number crunching. We can do absurly complex calculations on the fly, interpreted from data given by a few key presses or the gesture of a hand.. however we as consumers seem to insist that things in games must only be addition and subtraction.  Even shooters started out with this silly formula - and some still retain it.

Herein comes another big factor - people who aren't as big on twitch games, and PvP combat, love to number crunch. It is inevitable that you will get number crunching in a game.

I can't tell you how fucking sick and tired I am of hearing the other Shaman in my WoW guild do their theorycrafting about DPS numbers, percentiles, etc.

God, I don't CARE. Give me a set of abilities that do fun things, and let me mix it up in PvP and see what I can do with them.

I won't lie, though. I love the numbers, because at this point in time, there's no viable alternative. It's deeper then an FPS (though less twitch - WoW PvP is close, but not quite).

Anyways... In an MMO, you have to have the carrot, or the thing to go for. If it isn't shinies increasing your numbers for the same 2 abilities you do all game long, then what can it be? If any attack can be potentially lethal, just make it so you have a wider variety of tactics and abilities you can perform.
Therein lies the challenge I'm proposing, though - if you get away from the numbers, you're slowly moving away from the Diku model. But, if you move away from the numbers, you have to find some other way to keep your playerbase interested in what needs to happen.

There's a start, something to chew on for now, but it's farther than you'll see games evolve away from DIKU for a good, long while.

I agree, and while I think that it's slowly moving in the right direction with faster, more PvP-centric games, we have a ways to go.

[pxib]

I definately think that "stats" should be based on body type. Larger people are stronger. Period. They may not be able to do the same "feats of strength" as the light and lean (pull-ups, sit-ups, climbing a rope) but you will always want them on your end of the rope in tug-of-war.  Taller people have a longer stride and move faster with the same amount of energy and training. Small people have more endurance and use less energy across the board. When the Warhammer zealots talk about "silhouetting" they have a good point, but they're going about it the wrong way. I can't identify somebody's "class" just by looking at them in the real world, but I can get a good idea of their strength and speed.

That's how we make Fight or Flight decisions.

The Trinity could be completely destroyed by simply elminating battle healing and taunting. Healing is something you do outside of combat, somewhere safe. If you're taking damage, get out of there or you're going down. Nobody shouts "heal me" because that's time during which they aren't either fighting or escaping. Without taunting, there is no tank. Your largest, sturdiest fighter tries to keep the attention of the most dangerous enemy. He does this by beating the hell out of it. With body-based stats, brute force DPS and damage capacity go hand in hand. Somebody has to keep that guy busy or he'll kill us.

Finesse and utility are brute's nemesis. Put an arrow in his eye and he dies like anybody else. Set a trap for him. Sweep his feet out from under him. Set him on fire. Whatever. He may be able to throw your friends at you, at a considerably less extensive range than you can David his Goliath.

Like Slayerik I don't think classes are necessary. Humans have a limited carrying capacity and relatively few limbs. There's nothing unrealistic about the limitations Planetside imposes on a character's inventory. If the game is designed with that sort of thing in mind, players won't feel straightjacketed. Rather than balancing "classes" designers balance weapons and abilities... and players are surprised by variability.

Do offer players development paths, however. The best thing about classes is that you don't have to waste time defining yourself and can get right to playing something that's tested and that works. You want to be a Brute? Here's the skills and equipment to concentrate on. Sniper? These ones. Theurgist? Thus. Ghostwalker? Here.

Perhaps even start players in a "class" and then let them mix and match once they complete certain quests and tests.

[tazelbain]

If you kill the trinity, you have to replace it something so people feel they have a way to contribute to a positive outcome to the battle.   

Personally, I'd be very interest in a game where any class could play any role in the trinity and which class is the most effective in the role would be determined by who you opponent is.  It'd make group vs group very interesting as people played a cat-and-mouse game sliding between roles trying find out what roles the other team is and hiding what roles your group is in.

[pxib]

How about Light and Heavy ranged attackers and melee...

Melee: Lots of damage, no range.
Light ranged: Relatively little damage.
Heavy ranged: Lots of damage, any attack interrupts it.

Maybe every class can do all three in different ways. People are fighting as Melees then one sneaks off and starts the Heavy shelling, so a Melee goes Light and starts interrupting, and another Melee takes the opportunity to start hitting her. Maybe she can't use her shield when she's acting Light.

Theoretically this could be spellcasting (touch, blast, ritual) or guns (shotgun, pistol, mortar) or whatever. I wouldn't be surprised to learn that this arrangement already exists in some FPS somewhere.

[bhodi]

The only thing more retarded than the trinity was the whole idea of 'pulling' and beating the shit out of a single monster while their buddies stand there, stare at the wall, and scratch their ass. WoW was the first real diku I played, and I was agast -- it looks really, really DUMB.

[Alkiera]

The only thing more retarded than the trinity was the whole idea of 'pulling' and beating the shit out of a single monster while their buddies stand there, stare at the wall, and scratch their ass. WoW was the first real diku I played, and I was agast -- it looks really, really DUMB.

The whole reason for this is the issue of the PC power /NPC power ratio.  If that ratio is <= 1, then to fight any given mob, you need more than one person... Which sucks.  Most people seem to prefer a CoX-style 3/1 ratio, where a PC can beat up 3 normal 'equivilent level' guys.  There are tougher mobs(lts, bosses, etc) but most mobs are significantly weaker than the hero/villain.  There's just more of them.  This fixes the 'us 5 versus that one kobold' issue.

You still have some 'pulling' in CoX, because you don't want multiple groups.  But that's because of, as in many MMOs, there is no collision detection for NPCs, so you can have 3 dozen guys piled up on one spot, all hitting you at once.  So you need some code that says 'the space round person 'x' is full, you can't hit him with melee.  Throw a rock instead' or something.  Tactics-based games have done this for awhile, but they have a lot more time to calculate such things.  Having to figure it out on the fly for 30 mobs is more challenging, but we're getting there with games like Dawn of War.  Unfortunately, stuff like this is often dismissed out-of-hand as 'too hard' or 'too CPU intensive' by developers.  Admittedly, DoW gets sluggish when you've got more than a few dozen units on screen(orcs vs. orcs ftw!)  but how much of that is due to poly-count vs. AI I'm not sure.  I'm guessing it's more polys than AI.  It's just case of 'everyone knows', where what everyone knows is wrong.

To couple with this, you can ramp up the perception abilities of mobs, so you don't get the guy scratching himself while his buddy is being eaten by a cat on the other side of the room.  I did this just yesterday in WoW, in Desolace.  Had to rescue some chick from evil caster-types, so I killed the wandering one, and then worked my way around the edge of a 40' diameter room, killing the guys one at a time until the room was clear.  None of them noticed the cat suddenly appearing behind their buddy, or all the screaming and growling that followed.  But it had to be that way so that a solo player could have a chance at taking on the room, since they obviously can't fight all 7 guys at once... Nevermind that it happens all the time in movies and books. 

--
Alkiera

[pxib]

I think CoX is a very good example of what to work towards, even though it leans a little on the Trinity. In movies, badasses fight a whole room of guys because the guys are:

a) moving slower than the badass
b) attacking less frequently than the badass
c) easy to disorient and knock over

CoX (and a number of console games, Dynasty Warriors for example) work this way as well. The hero in all these examples uses sweeps, throws, and other "AOE" attacks to keep the thugs busy while he concentrates heavy attacks on the "boss". If there's a duo one tends to be on trash mob duty while the other tanks. The only part of the Trinity missing is (again) the healer.

[Alkiera]

I think CoX is a very good example of what to work towards, even though it leans a little on the Trinity. In movies, badasses fight a whole room of guys because the guys are:

a) moving slower than the badass
b) attacking less frequently than the badass
c) easy to disorient and knock over

CoX (and a number of console games, Dynasty Warriors for example) work this way as well. The hero in all these examples uses sweeps, throws, and other "AOE" attacks to keep the thugs busy while he concentrates heavy attacks on the "boss". If there's a duo one tends to be on trash mob duty while the other tanks. The only part of the Trinity missing is (again) the healer.

The difference, tho, is that rather than being 'tank/dps' and 'CC/dps', both group members are just 'badasses' to use your phrase.  The one guy is taking on a powerful boss, and the other is taking on the boss's group of lackeys.  Ideally they finish their respective fights at about the same time. (Dramatically appropriate)  And as for the healer... eh, toss 'em.  I like the 'defender' route, where powerful buffs/debuffs prevent damage rather than having powerful healing effects flying about during a fight.  The buff/debuff effect is also more compatible with a system without the standard HP/MANA stats.

--
Alkiera

[Glazius]

You still have some 'pulling' in CoX, because you don't want multiple groups.  But that's because of, as in many MMOs, there is no collision detection for NPCs, so you can have 3 dozen guys piled up on one spot, all hitting you at once.

Point of order:

They fixed the thing where if an enemy couldn't get where it wanted to go it would play a jump animation and then go there anyway. So you can't pile up a whole bunch of mobs at a single point anymore. In fact, mobs tend to give up pretty easily on pathing to you if the way is blocked by enough of their brethren, so you can have maybe two or three gangers in front of you and the rest _could_ flank you but instead are hanging back with pistols.

Then when you've gotten used to the pattern 15 Sky Raiders teleport to your back ranks and swing their machetes in unison. Brr.

--GF

[Samwise]

But that's because of, as in many MMOs, there is no collision detection for NPCs, so you can have 3 dozen guys piled up on one spot, all hitting you at once.  So you need some code that says 'the space round person 'x' is full, you can't hit him with melee.  Throw a rock instead' or something.  Tactics-based games have done this for awhile, but they have a lot more time to calculate such things.  Having to figure it out on the fly for 30 mobs is more challenging, but we're getting there with games like Dawn of War.  Unfortunately, stuff like this is often dismissed out-of-hand as 'too hard' or 'too CPU intensive' by developers.  Admittedly, DoW gets sluggish when you've got more than a few dozen units on screen(orcs vs. orcs ftw!)  but how much of that is due to poly-count vs. AI I'm not sure.  I'm guessing it's more polys than AI.

In Starcraft you could have a good 600 units in the world at once, each with serviceable AI that did reasonably good pathing, and each one "colliding" with its neighbors and the world so that you didn't have them piled up on top of each other (unless they were flying, in which case piling was permitted).  Starcraft ran on my Pentium I.  I should certainly hope that if Dawn of War is having trouble doing the same exact thing with 30 mobs, it's because of rendering and not AI.  This shit has been SOLVED for a long time now.

At GDC there was an AI middleware company showing off highly intelligent pathing (better than anything I've seen in any game to date; stuff like being able to plot optimal walking routes across 3D piles of randomly falling debris, identifying tactical choke points dynamically, that sort of thing) that could account for thousands of dynamic objects, with thousands of NPCs all runnning at the same time.  Clicky.  So scale or complexity aren't really very good excuses either at this point.

[bhodi]

If you push enemies to extremes, I really only see only 3 types of enemies:

Bosses -- Single huge monster with complicated scripted events and attacks
Doubles -- Enemies similar to the character with similar abilities, with some AI and logic scripting
Peons -- Enemies that throw themselves at you by the dozens but no real threat except in numbers

It's pretty easy to match them with the games, so I'd like to suggest a good game needs to incorporate a combat system that involves all 3. People like beating down huge bosses, showing superior tactical ability versus doubles, and carving paths through peons.

No more pulling back to a camp spot; At most, you'd separate them by room, and you need to give a sense of urgency and scripted events to force people not to be able to stay in one place... or move them through the environment in such a way that camping becomes pointless except in very small bursts. People would get angry, but you could have alarms sound if they path too far; they'd run back and get friends and then come hunting for you.

A combat system that is robust enough to be able to take large amounts of enemies, which means a flexible damage system and possibly AoE attacks, and at least some sort of collision detection is needed.

Absolutely no taunt and no "aggro management" game. That gets rid of the holy trinity right there. Enemies will go after the closest person or some other sort of scripted criteria. A skill that will force a monster to suddenly ignore their target over and over again is retarded unless it's used very very sparingly. A panic button, not the norm.

[Nonentity]

A combat system that is robust enough to be able to take large amounts of enemies, which means a flexible damage system and possibly AoE attacks, and at least some sort of collision detection is needed.

Absolutely no taunt and no "aggro management" game. That gets rid of the holy trinity right there. Enemies will go after the closest person or some other sort of scripted criteria. A skill that will force a monster to suddenly ignore their target over and over again is retarded unless it's used very very sparingly. A panic button, not the norm.

I completely agree. I want a game where the environments are more then just glorified boxes with pillars here and there. If the environments were actually detailed areas, you could completely do collision detection as an 'aggro' feature. Give the big meaty guys hooks and stuff that will physically pull people towards them, knock them down, and so on. Give the gunners or ranged guys kicks and stuff that will knock them back.

I know that collision detection is possible in an MMO, but the griefing features are what keep developers from implementing friendly collision detection. But, I personally think that inside your warehouse/office building instance or whatever, if you couldn't get out of the way of your teammates, they'd have to get a way around you. It would add that element of needing to worry about spatial positioning.

[sinij]

Why class? Let people specialize in what they like AC/UO style and let them define classes as they chose. This way tank will be one with most damage mitigation skills and so on.

[ajax34i]

If you make the AI be more intelligent, closer in ability to what players can do, how exactly is the player supposed to win consistently?  A lot of the fun in a game depends on that.

The AI level is a direct buff to the difficulty of the NPCs, which requires a proportional buffing of the player's character in order to maintain fun.  But, the devs cannot control or predict how "smart" the player is, so the only way they can buff the player character is just health, mana, and built-in damage mitigation.  Taken to its limit, this results in a game where the AI is genius, but player characters are boss-like in their HP's, mana, and AC, which is pretty much a reversal of the scene that we have now.

There's also a huge problem with "balancing" force multipliers.  It's easy to buff or nerf a warrior, priest, mage, and rogue, so they're "just right", but put them in a group together, and it's difficult to balance the power of that group as a whole.  Doing this for groups of NPC's, with "intelligent" AI, is probably very difficult, simply because of the large number of possible combinations that you can have.

Anyway, I'm not a designer, maybe these issues have been solved already, and I don't know about it.

[bhodi]

If we continue and accept the 3 different classes, the only thing you have to worry about is the middle class of beastie.

Boss battles take more teamwork, coordination, timing, and positioning to win; I think of them as a virtual trap they will fail over and over again until they find the right combination. It's not about raw output in terms of damage or soaking, if you stand in the death-beam as it sweeps across the floor, you will die.

Peons are simply meant to hamper and wear down people, but they pretty much die on contact. They don't need to be strong and they are easily balanced -- a "perfect" group won't be able to kill significantly more of them faster, since you're either going to be limited by scripted events that 'unleash the horde' or you're going to be limited by the move/walk/travel speed of the party.

That leaves the doubles. It could be tricky, but let's be honest -- a lot of the efficient combinations are going to come out during testing and you can balance against them. Make sure you design in such a way that characters only stack so much, and the problem can be mitigated.

[Samwise]

If you make the AI be more intelligent, closer in ability to what players can do, how exactly is the player supposed to win consistently?  A lot of the fun in a game depends on that.

Maybe you should solve it by just having a "win" button on the interface.  That way no matter how much you suck at the game, you can hit that button and everything will be better.

This is why MMOG design discussions always piss me off.  Things that are taken for granted in every other type of game, like "it should be fun" and "victory should require some sort of skill on the part of the player", go right out the fucking window when people are talking about MMOGs, and I don't see why that should be.  Something about mumble mumble subscription mumble mumble.  FUCK.  THAT.  SHIT.

(edit) And you know what really pisses me off?  Game developers wave off lots of genuinely good and thoughtful feedback from players because, in their view, all the players are vapid dumbasses who just want an I Win button.  And that perception is YOUR FAULT.  Yours and other people's who can't handle the concept of PLAYING a GAME being an interactive, dynamic, challenging activity.  No wonder we have treadmills instead of gameplay and comparison operators instead of combat.  FUCK.

[Alkiera]

At GDC there was an AI middleware company showing off highly intelligent pathing (better than anything I've seen in any game to date; stuff like being able to plot optimal walking routes across 3D piles of randomly falling debris, identifying tactical choke points dynamically, that sort of thing) that could account for thousands of dynamic objects, with thousands of NPCs all runnning at the same time.  Clicky.  So scale or complexity aren't really very good excuses either at this point.

Dang.  I watched a few of those videos... and you're right.  There are no more excuses.  I now wanna be one guy at Helm's Deep, animated in realtime.  I think this stuff could do it.  I'd almost settle for that level of art detail, too, if the game was fun.

--
Alkiera

[bhodi]

I'd also like to take another second to say "fuck tagging/spawn camping/kill stealing". It's retarded to the max. Since this is MMO-land you are trying to foster teamwork and grouping with random strangers. Even Disney got it right on the first try. Do what Toontown does: you join in, you get full experience. The game's about fun and spawn camping and kill stealing don't add to that. You could limit it at say 10 people per mob if you are REALLY worried about huge groups gang-banging everything in sight in an effort to power level their way to the top, but I'm pretty much of the opinion that if they want to get there THAT badly, than you might as well let them -- you've already failed somewhere at that point anyway.

[Samwise]

I'm pretty much of the opinion that if they want to get there THAT badly, than you might as well let them -- you've already failed somewhere at that point anyway.

YES.  Creating a series of cockblocks to keep people from rushing through your game is treating a symptom, not the disease.

MMOGs by and large aren't actually fun to play for the period of time that they're intended to be played for.  That's the real problem.

[Typhon]

If you make the AI be more intelligent, closer in ability to what players can do, how exactly is the player supposed to win consistently?  A lot of the fun in a game depends on that.

Maybe you should solve it by just having a "win" button on the interface.

I think you can have a game where you don't win consistently if developers give up on the death penalty concept.  Not winning is enough of a penalty already.  If I'm going to lose 1/3 fights because the AI is smarter, I don't need a death penalty, in fact, I should still be gaining something even for the loses.

On another topic, I'm really excited to see that pathing engine as well.  IMO what made CoH combat so much more fun then any other MMO is that you could use the world geometry during combat.  Shoot/duck, knock-back juggling melee attackers, etc.  THAT is what is needed to make these games more exciting - you win against multiple opponents not because they do/take significantly less damage then you, but because they lack the special abilities (faster run speed, knockback/knockdown, daze, stun, fear, etc) and the intelligence to use the world geometry to their advantage.

"Pulling" becomes "moving", where you move around enough so that you are only fighting one enemy at a time (even though they are all trying to close with you).  I see it something like this (going to use the super hero setting).

As part of you current "quest" you find a group of thugs in a back alley.  You drop off a roof onto one of the thugs, knocking him flat.  He turns a deep red, which will slowly fade as he recovers from the blow, after which he will get up to come for you.  Upon knocking him flat, you do a foot-sweep to the second villain, also knocking him flat.  You quickly duck around a corner, causing the third thug to follow.  As he rounds the corner you perform a body-throw, tossing him into the wall behind you, and you set in to beating on him.  The footsweep victim should be rounding the corner about now, so you have to decide whether the third thug is nearly KO'd, or do you need to use more specials to keep them off you.

The concept is like pulling, but it's more dynamic.  THAT sounds like a lot of fun to play to me.

[Samwise]

I think you can have a game where you don't win consistently if developers give up on the death penalty concept.  Not winning is enough of a penalty already.  If I'm going to lose 1/3 fights because the AI is smarter, I don't need a death penalty, in fact, I should still be gaining something even for the loses.

In a "real" game (one where the outcome is determine by how good the player is at playing the game), you learn from your losses, and through practice eventually gain the ability to defeat more difficult opponents.  The game itself doesn't need to reward you, because you generate your own reward simply by playing.

Again, this tends to go right out the window with MMOGs because there's no skill to speak of involved in playing the game.  And there's no skill involved because if playing was dependent on skill, people wouldn't always win.  And then they wouldn't have fun because they don't gain anything from losing.  And they don't gain anything from losing because there's nothing they can learn from their losses.  See?  Vicious circle.  Also, panties.

[Typhon]

Samwise I understand what you are getting at (players play chess for the pleasure of playing chess, part of which comes from getting better at playing chess), but my fear is that solely skill-based progression in a PvP game short-circuits the retention that progression-based systems bring (progression-based systems include UO, any Diku game, EVE, ATitD (learning skills is progression), Planetside, GuildWars (unlocking skills is progression), etc, etc, etc).

Why I think there would be no retention with solely skill-based character progression is because:  Based entirely upon progression, the bottom 10% (from a player skills perspective) of the playerbase will progress so slowly that there is no hook to keep them playing.  This is made worse by the fact that these folks loose in a majority of the encounters in which they participate.  They'd have to be very stupid to pay to play a game where they were losing all the time and also not progressing in any measureable way.   Once those folks are gone, you have a new "bottom 10%" that also doesn't progress and eventually leaves.  You will bleed players until you are left with the top X% of players where X is the percentage where difference of player skill allow for an "any given Sunday" mechanism to kick in.  My thinking is that if X is larger then 50% you are talking about a game that many people would not feel is skill-based.

Thus my comment that "they need something for losing", specifically I'm referring to progress in whatever avatar-based progression metric is included to lock-in player retention.  [Also note: I assume that we aren't talking about Diku level progression here.  An example of non-level based progression: take the CoH character customization screens (the "Tailor") and have "progression" give more character customization options.]

[Hoax]

I still think Xilren solved a fuckton of MMOG problems when he said that devs need to diversify the type of gameplay.

If the melee is button mashing, make the crafting puzzle solving, have a stealth/hacking system that requires platformer timing and make the casting system be all timing, anticipation and strategic coordination.

Now nobody should end up in the bottom 10% of all aspects of the game.  Toss in some social tools, players as content and call it a day.

Seems easy to theorycraft, but the problem is building multiple gameplay systems and not breaking any of them is sure to be a costly bitch.

[Merusk]

If you make the AI better to defeat "skill" then what skill level are you aiming at?  If you go for the median of the entire game, then most players here will STILL bitch, because they are above the median.

If you aim for a skill-level that is on-par with most players here, then you alienate a large segment of players, and still have bored "omg so lame" whines from the top-tier.

Aim for that top-tier and you may as well never open your doors, because nobody's paying to get curbstomped by an NPC all the time.

That's the shorthand of what Typhon just said.  It's also why you have Handicaps in Golf, Bowling and other scoring-based games of skill. Are we proposing targeting handicaps in skill-based shooters?  Or perhaps being spotted extra HPs/ Armor/ MPs in RPG-based games? 

Even diversifying the gameplay isn't a really good solution.  Maybe I'm the best puzzle-solver EVER.. but all I want to do is bash some PC/ NPCs.  I don't get to, ever, because I'm bad at button mashing?  Yeah, great.. I can be the top in that one segment of gameplay, but if it's not what I'm after I don't give a flying fuck how good I am at it.

Most folks aren't looking for validation that they roxxorz, they want to play a GAME and be entertained.  (The most important point, yet it keeps getting lost.)

In the end designers just need to make the game what they want it to be.  Nothing will ever attract or satisfy all audiences and all players.  There is no "one solution" to gameplay.

[Samwise]

Wow.  Does anyone who's arguing in favor of the "treadmill to victory" mechanism play any games other than MMOGs, ever? 

Seriously.

[Typhon]

I was thinking of the issues with a PvP game but I agree that there are many similarities between the two (PvP handicapping versus PvE handicapping).

I think that with PvE the developers have, at least in theory, a way to give everyone what they want - Auto-handicap an encounter based upon the ranking of the players by increasing/decreasing NPC difficulty (I would suggest that decreasing stats rather then AI would lead to more satisfying encounters for the gimps, er, skill-challenged).  I know it hasn't been done well to date, but how many game developers think about how to rank players after the fact rather then as part of the game design (ok, has any developer ever created an internal mechanism to rank player ability? probably not, but they'll have to start at some point).

Also, believe it or not I rewrote the above probably half a dozen times before posting.  I'm not very concise.  I realize that long-winded posters are boring.  can't seem to fix it though.

[Typhon]

Wow.  Does anyone who's arguing in favor of the "treadmill to victory" mechanism play any games other than MMOGs, ever? 

My introduction to online gaming was via Quake II.  I'm eagerly awaiting UT3.  I also play alot of MMO.  I'm actually basing my analysis theorectical-mental-masturbation on what happens in the FPS market - overtime the playerbase drops off until it's servers with people who are really good at it (that eventually go password-only), or servers with people who cheat.  I'm not basing my analysis on because I'm a slave to a treadmill.

I think that the fact that UT is adding a significant amount of effort to a singleplayer offline mode with the next version backs up this perspective.  I think it comes down to the fact that people aren't all that wild about getting their ass kicked when they just trying to play a game and blow off steam (although a fair amount is undoubtedly due to people not wanting to be bothered by the bottom-of-the-barrel slice of humanity that hangs out in open server games) - which is what Merusk is saying.

If you want to tag on progression to a skill based game, but you give limited opportunities to a percentage of your playerbase to enjoy that progression (along with the development of skills), you are artificially limiting your audience.  Again, I'm basing this entirely upon my experience with what happens to FPS game communities over time.  Also, most FPS games are purely skill based, and they are not pay-to-play.