The problem of group vs group combat

[Nija]

I liked Bushido Blade more as well, though that video was cheesey sploity crap. Me and my friend went nuts in VS mode with that game. A lot of our matches either came down to who broke a standoff first and attacked, or who was the first to misjudge their range and whiff hard.

Yeah, I looked through a few other bushido blade videos, but the other ones were the intro only. (why?)

or crap like this. http://www.youtube.com/watch?v=d7d6fDzZu5o

[Hoax]

If we're talking about Massive Multiplayer anything .5 second deaths = lame as hell.  I remember playing the first Rainbow6 multiplayer and it was like 95% camping as three bullets on target were almost gaurenteed to incap somebody.

[pxib]

A lot of our matches either came down to who broke a standoff first and attacked, or who was the first to misjudge their range and whiff hard.

Go see the thread on high latency combat. As cute as it is to feel like you've got a "realistic"  combat system, if that's based on position, distance, or precise timing things get awfully "sploity" in PvP. Everybody wants to one-shot the baddy... nobody wants to be one-shot when they thought the baddy was out of range and safely to their left.

[sinij]

Also if you were a mage and you started casting (somewhere else, outside or something) and your melee buddy wasn't paying attention - if he runs in front of you right as the spell goes off, it'll hit him and blow up in your face, literally.

I think this should happen, only player should have a bit more control over casting and targeting. If you can prepare (load) spell and then target it you would see less LOS abuses and more tactics. You can also add spells that have direct line of attack, spells that have parabolic line of attack, spells that rain from directly above and spells that are on-touch only.

I'm tired of DPS vs DPS combat, more options during a fight is always a good thing. Missing or getting friendly fire due to player actions is a positive step. I think its time we move on from auto-hit auto-aim mechanics.

[tazelbain]

That all said, I think the real reason why HAM probably wasn't a good idea was that it was too complex.

Boo Yah!
Anyway, what makes them too complex is not the mechanics themselves but the inability to give back feedback so people can make meaningful decisions in combat.

[Yegolev]

I think this should happen, only player should have a bit more control over casting and targeting. If you can prepare (load) spell and then target it you would see less LOS abuses and more tactics. You can also add spells that have direct line of attack, spells that have parabolic line of attack, spells that rain from directly above and spells that are on-touch only.

You are mostly talking about Phantom Dust, except PD does not have precasting, and it's more of a CCG-style game.

Bushido Blade was a great fighting game, one of my favorites.  I'd also suggest something similar to the health system in Riddick as an option to the DPS problem.  In any case, you cannot just plop these systems into a traditional design to good effect.

Disclaimer: I never played SWG.  On the subject of too-complex combat systems, it might be worthwhile to take a harder look at that.  I am sure that complex combat systems are nichey, but how nichey?  I can imagine comparisons to the hundreds of wacky combat and advancement systems I have been exposed to over the years, but most of those were single-player games and as we all know the top 0.5% of cockholsters will ruin the game for those just trying to have fun.  In any case, I am on tazelbain's side here.  Complex combat does not scare me, but I do need adequate time to consider my options and require solid feedback.  If you cannot go turn-based, then slow things down a bit.

[ajax34i]

What about limiting "focused fire" via limitations on targetting rather than damage?  Specifically for EVE, but workable in other games too, 5 of your gangmates pinging that target with active targetting systems, the echoes totally screw up YOUR computers and you can't get a lock on it?

[Roac]

What about limiting "focused fire" via limitations on targetting rather than damage? 

Ultimately, a solution along these lines needs to result in an increase in dps per peson added, but a decrease in the dps for each participant.  If this is handled via a percentage cut in damage given, or increase in the percent misses which net the same result doesn't matter.  The thing to avoid is any ability for players to target / fire upon a friendly in order to keep them from taking damage.  That is, the lv 100 SuperKnight shouldn't be unkillable because he has 10 lv1 alts trying to hit him.

[Yegolev]

I have thought about it for a bit and have decided that I don't like the idea of diminshing returns on damage.  I prefer the notion that ten torpedos will do ten times (approx) the damage of one torpedo.  It doesn't lend itself to the current drive by CCP to slow down the pace of combat, but is more sensible to me.

[Dren]

I wouldn't like to have a cap on DPS or any other general cutoff to try and reproduce logical limitations to "zerging."  I'm not sure how any existing games could be changed, but I certainly could come up with a game that looks at this problem as a fundamental requirement from the beginning.

The types of characters would have to developed with this in mind with each having specific relationships to different damage types.  The damage types would be.

Close Melee - physical force applied by touch, character to character.
      - Defensive stance - medium influence from LOS, medium damage, high accuracy, fast reload
         typically.  Backside defended well while stationary.
      - Offensive stance - highly influenced by LOS, high damage, high accuracy, reload dependent
          on weapon type.  Backside very vulnerable.
Ranged melee - physical force applied by use of a tool that is propelled towards another character.  Always vulnerable to attack.
       - Lofted Ranged Melee - avoids any issues with vertical LOS, low damage and low accuracy rating.  Medium reload speed.
       - Direct Ranged Melee - can hit objects within verical LOS, high damage and medium accuracy. Slow reload speed.
Artillery (magic or conventional) - objects or energy propelled towards another character often creating a
          secondary impact within its own energy. (fireball or bomb)  Always vulnerable to attack.
       - Lofted Artillery - avoids vertical LOS, low accuracy and low direct damage or medium damage over
           time. Secondary effect (if present) very much influenced by LOS of surrounding characters,
           frendly or not. Medium reload speed.
       - Direct Artillery - greatly affected by LOS, high accuracy, medium direct damage or high damage over time. 
           Secondary effect (if present) affected by LOS but not as much as lofted. Slow reload speed.
Indirect Influence - methods used to reduce or boost abilities, position, influence, etc. (buffs, debuffs, etc.)
        - All ignore LOS, no damage, perfect accuracy although resistances could be applied.
        - This could be used either in a magical game as standard, or as music (think battle chants, bugler, etc.)
           or as shouts.  You get the idea.

LOS and collison detection would play a huge role for this all to work.  This would all lead up to a limited number of close melee character able to be in proximity to a target thus limiting their high damage attacks to 4 or 5 of them.  Ranged attacks could be made, but if an open shot is not available, lofted shots would have to be made with the limitation of lower damage and accuracy.  Buffs and debuffs are pretty much open with restances being their only obstacle.

The calculations for a succesful attack would:

First:  Check for LOS.  If LOS does not satisfy conditions, the attack fails and is not fired.  Friendly fire is not possible.
         (This is just making the judgment call of a shot or no shot for the player since it isn't always obvious and time doesn't really permit it.)
Second:  Once LOS is obtained, the attack will only be directed at the target.  Further calculations will only involve that target.
Third:  Accuracy is taken into account and used to calculate whether the attack has actually hit the target.
Fourth:  If the target is hit, the damage is calculated and submitted.  Otherwise, no damage is taken.  Resistances can affect damage taken based on armor or other special reasons.

The only part I'm a bit lost on is the splash damage from AE spells/artillery that have a secondary affect.  I'm half inclined to just leave them out as possiblities, but that seems wrong.  Including them would mean providing chances for friendly fire, which always seem to kill the use of the attacks for all BUT the griefers (a la UO.)  Perhaps, the secondary affect does not fire if friendly chars are included in LOS.  The secondary affect would just fizzle in that case.  Dunno, seems like a sticky one.

For those wanting to avoid an attack by ducking around corners, once LOS has been made for you, the attack will come regardless.  Ducking is still viable as it would help keep LOS broken if you time it right.  However, once you've been "made," the attack will start its graphical show and calculate a hit or a miss.  This might have the affect UO had with archery with an arrow following you around a corner, but that's just the animations catching up with the attack.

Calling targets should still be a very good strategy, but with a bit more intelligence than "Everyone get XXX."  Now, there would be a lot more orchestration that would probably include the use of smaller cells dedicated to certain jobs like breaking down the front lines, or infiltrating to the backline of ranged attacks, or influencing the combat through indirect means, etc.

From an individual perspective, this also helps with the ranged fighter vulnerability syndrome.  While they will still be somewhat more vulnerable from physical attack as compared to a front line fighter, they will be by no means weak.  While they will only be able to do limited damage if the foe is surrounded by close range fighters (dropping arrows on them from above,) if a fighter were rushing towards an archer/mage/etc. without anything in between, the direct attacks would be far more damaging and hard to ignore.  Other effects could be included to slow the opponent, etc.
 

[Typhon]

My suggestion is to make defenders aware of concentrated fire, and make defensive actions be typically quicker (or much quicker) and scale more then multiplicatively.  (what the hell did that mean?)

Have offensive abilities take (in general) a couple seconds to pull off.  Have some graphical representation of when an offensive ability is being brought to bear on someone.  Defenders now have some (short) period of time to stack defenses on the target, defenses stack more then multiplicatively (e.g. 3 shield actions results in 5 shields defense, which would more then balance out the combined attack strength of 4 bolts).  the system shouldn't penalize 2 on 2 combat (i.e. 2 shield actions results in 2 shield defense), but should make a quick acting defensive effort more then able to completely counteract a focused 5 or 6 man (or larger) effort.

This would require most/all classes being able to bring to bear both offensive and defensive abilities.  But I like hybrid classes, I think pure classes result in boring gameplay.

[Rhonstet]

Concentrated fire is only an issue in games that support known hit points.  When you know for a fact that you can survive a minimum of X number of hits from an enemy regardless of what you do, you're going to get target-calling.

One way to solve target-calling are creating defensive abilties that are similar to the now-standard hokey attacks.  An example is the Paladin Bubble from WoW: a short term defensive abilty that provides total protection but limits attack ability.  Temporarily negating the effect of that focused fire is one way to counter target-calling.

Another possibility is to figure out a way to handle threat in PvP.  Special hotkey attacks or buffs that are applied to attackers who are generating the most threat if their attacks go unnoticed is an example.

A way that games like EVE could handle the target-calling problem is defensive ECM.  Making ECM that prevents anything from establishing a lock on anything within X km, or that allows a consumable to be expended to break all targeting locks on a ship, or just turning a defensive ECM into a percent chance to miss, might change fleet battles for (what is usually considered) the better.

[Hoax]

Defensive skills usually suck because they are designed with PvE in mind, and mobs have nothing on players when it comes to wanting to unleash pain in large doses.

Good point though.

[eldaec]

Surely collision detection takes care of this problem for melee, and for ranged, just drop 3% off of to-hit for everyone you are shooting through.

This way focussed fire is still possible, but you actually have to plan around it - rather than just making an assist macro.

It would also give much more incentive to thinking about positioning in general.

MMOG combat needs to make formations more important.


Another solution is for someone to hurry the fuck up and make total war medieval, the mmog.

Total war combat tactics are entirely about focussing fire, but they make it non-trivial to do, because you have to think about size of units, friendly missile fire, making units arrive at the right place at the right moment etc. And while you can run total war units through each other, you really don't want to.

[Lightstalker]

Greeting your foes with Massed Blaster Fire is exactly what we train our players to do in the PvE portions of many MMOGs, and (was) the only way to play Shadowbane.  What we've turned off with the lack of Collision Detection is maneuver.  Maneuver is a big deal in squad and larger unit conflicts, which often mitigates the frequency of Massed Blaster Fire in real life.  We're all familiar with the "cloud of ninjas assaulting the lone hero" battle, where a thousand enemies are defeated a handful at a time.  That has to do with packing real objects into the same space - only so many people can be in reach of our hero at once.  Not so in most MMOGs.  Maneuver in many MMOGs is all about creating a "cannot attack that" situation for the other guy by running away while you kill your enemy (if you aren't just outright exploiting some sync issue).

Many (board, war) games overlay on a hex grid, which provides 'better' freedom of mobility than square systems while keeping the number of nearest neighbors reasonable and remaining a regular close packed array.  Some of these games also incorporate damage patterns - specific attacks effect certain arcs to specified ranges over specified spreads.  These damage patterns can natually create interesting situations (friendly fire, widowed armor, etc.).  This is a problem that really has already been solved in other genres.  Magic and ranged attack could similarly be addressed by causing X damage to spread over a widening cone of area the greater the range.  This would naturally mitigate the ability to focus fire overmuch and produce a reasonable abstraction (not everyone ought to have sniper accuracy every time they pull the trigger - unless they are willing to sacrifice mobility and rate of fire to do it). 

The example of a SuperWarrior with his points of attack filled with allies, thus rendering the Super Warrior invulnerable, deals with itself.  Said warrior cannot reach anything of interest; boxed in as he is by allies he'd have to fight through.  The exploit here is if SuperWarrior can perform some action that violates the rules on nearest neighbors (like a management role or "Magic" ability - bardbot springs to mind).  The SuperWarrior surrounded by 'ablative' manpower can also have that manpower chipped away, and chipped away unexectedly leaving an unanticipated vulnerability for real damage.  If a patterned damage is used, that pattern can incorporate both width and depth, basically trample damage through the ablative/friendly ring and into the real target.  Not an altogether unrealistic scenario to attack one's foes through one's less valuable troops that were set out solely to pin down valuable enemy units. 

One of the things that makes Massed Blaster Fire so damning is the stickyness of target selection.  Instead of attacking into whatever arc we've got ourselves facing in most MMOGs we attack the selected target and only the selected target.  Very few abilities are strictly tied to our facing, most taking anything in the front 180 as close enough to launch an attack with pinpoint accuracy.  Even if we can't get the shot off, we don't "lose the turn" or have to find another target.  We keep running in a circle until our target is in a valid arc then we fire off an attack, maybe losing some fractions of a second.  I'm not entirely opposed to sticky target selection as a simplification so long it doesn't simplify too far, which Massed Blaster Fire is the result of.  If one was limited to sticky target selection I'd like to see patterned damage centered on that target.  If one allows pinpoint accuracy, one should have to aim the shot or risk hitting whatever happened to be in the sights at the time you pulled the trigger (dissallow sticky targeting if accuracy is absolute).

In short:

• Imposing a nearest neighbor grid on the combat can address Massed Blaster Fire for melee units
  or a patterned damage for melee attacks (effectively limiting attack density - unless you do something stupid like allow typed damage that one can anticipate...).
  
• Imposing patterened damage can address Massed Blaster Fire for ranged units.
  

Cheers,

[Evangolis]

Bit of a necropost, but Design is always a slow discussion.  Interesting points.  I'd throw Line of Sight in there as well, to minimize clumping by exposing combatants to possible friendly fire.

On the other hand, can an internet game really handle all the computation this idea implies?  With collision detection, what happens when all of the players are lagging by different amounts?  I could see calculations becoming deadlocked due to interdependent positional corrections, and/or possible exploits based on controlling lag to exploit collision detection.

[Hoax]

It better be a computation problem, because otherwise current game devs are even more lame/retarded then I think they are. 

[eldaec]

Bit of a necropost, but Design is always a slow discussion.  Interesting points.  I'd throw Line of Sight in there as well, to minimize clumping by exposing combatants to possible friendly fire.

On the other hand, can an internet game really handle all the computation this idea implies?  With collision detection, what happens when all of the players are lagging by different amounts?  I could see calculations becoming deadlocked due to interdependent positional corrections, and/or possible exploits based on controlling lag to exploit collision detection.

Computation isn't nearly a problem.

Lag and potential abuse outside of pvp is what has worried devs with regard to collision detection.

DAoC is probably the single game most in need of collision detection on this earth. And it's available in the engine.

The reason we are told it was never switched on is network traffic and lag related exploits (I guess without collision detection the client can determine player location on its own. With it, the server and clients need to negotiate)).

Personally I think collision detection between enemies is simply a must. Someone has to find a way to make it work for large scale pvp to stay interesting.

[eldaec]

Just rereading some of this thread - I'm still not clear why focusing fire and target calling is 'a bad thing'. I agree that most games make it too easy - but that's really about the lack of things like collision detection or meaningful ways for players to shield teammates using their positioning and abilities.

But prioritising targets just seems like common sense to me.

The tactics that require organisation + common sense should probably be rewarded?


The problems with hit points are probably unrelated to this as well. Any fun model is surely going to require multiple attacks to defeat a PC, and the quicker multiple attacks are made surely the quicker a PC is going to die under any sensible model.

[Lightstalker]

Focussed Fire is a bad thing when the expected combat lifetime of a character is significantly less than the expected latency on the connection.  Under that condition the player just dies, with no opportunity to make a decision - like moving into position, coordinating, or cleverly targeting a critical component of the opposition force.  We'd like to see clever player choice rewarded and brute overwhelming force mitigated, because it is more "fun" for the user to go down fighting than to "random" acts of overwhelming force.

Prioritizing targets is common sense, but in the real world there are practical limits to how many people can punch you in the nose at once.  You have one nose.  It doesn't matter how many fists the other team has, there just isn't enough space for every fist to hit your nose at once.  There is, in a physical world, a limit to how quickly you can take damage; which provides some measure of decision making for both parties to the conflict.  Focussed Fire is dropping a 16 ton weight on the other guy before he can do it to you.  That isn't clever gamesmanship winning the day, that's a game of chicken.

One could just put collision detection on the client, and trust that exploiters typically are not subtle and will clearly self select themselves into a flaggable and disciplinable pool of cheaters.   :-D   Of course, I'd keep patterened damage in that sort of world so all those players illegally piled into the same space kill each other as much as they kill their intended target.  Damage patterns for every damage source and full on friendly fire can introduce a cost to the present no-cost focused fire. 


The server doesn't need to care how much lag the client has, if the client steps out of bounds (via delayed information transfer) the server (should) just ignores the client's wishes.  For the player side that can quickly become an intollerable user experience, which is probably why client side delays are coded in so the user doesn't see "That spell isn't ready yet" when the client side timer runs out.   The server needs to care about "Is this move legal now" and "which character resolves its movement action first" neither of which necessarily needs to wait on additional input from the client to resolve the last input from the client.  High latency users will be seen as static obstacles to well connected users, just as it is in normal practice.  Dealing with gridlock is another issue entirely, which still hasn't been worked out reliably on our city streets yet.  ;-)

[eldaec]

Focussed Fire is a bad thing when the expected combat lifetime of a character is significantly less than the expected latency on the connection.  Under that condition the player just dies, with no opportunity to make a decision - like moving into position, coordinating, or cleverly targeting a critical component of the opposition force.  We'd like to see clever player choice rewarded and brute overwhelming force mitigated, because it is more "fun" for the user to go down fighting than to "random" acts of overwhelming force.

Yeah great - what post UO MMOG is that?

Seriously, every dev group knows that people dying before they can respond is bad.

They usually solve the problem by having PCs unlikely to die without being smacked several times, and they usually give support characters skills to react to it.



If the real problem here is people just getting sniped, then fix the real cause of front loaded damage  - which is mana and endurance pools.

Having your primary resource for attacks start out as high as it will ever get is a recipe for battles with all the dramatic spells and high damage attacks front loaded into the first few seconds.

Basic rules of drama tell us that this is stupid; never mind balance issues.

Combat should get more dramatic as is goes on; not peter out into a petty cat fight because I used all my mana shooting fireballs out of my ass in the first volley.

Mana and endurance in mmogs needs to work at least like rage meters, and preferably more like M:tG mana.


EDIT: edited for readability because I was typing like ass yesterday.

[Typhon]

[ability to inflict damage] in mmogs needs to work [in a way where your ability to inflict damage grows overtime, if unchecked]

That's how I read it, and it's a very cool idea.  What I immediately though of were two examples: a pyromancer type class that gained access to more power/more powerful spells based upon how much was burning in the combat zone and a whirling dervish class that gained similarly based upon how many fast he was spinning (and spinning faster was only achievable by hitting something).

Of course enemies of those classes would be trying hard to kill them (duh), and/or put out fires/slow the spinning.  Very cool premise.  Deciding when to attack and when to counter-attack (i.e. put out fires or slow down the spinning) adds a good amount of complexity/skill.

[Calantus]

Eldaec's post immediately made me think of a rage bar that does not empty until out of combat for a certain period of time. Damage you deal/suffer boosts the bar and you have a bunch of skills that have "requires X <rage-like resource>" before they can be used, and the size of the bar could boost power by itself as well as possible affecting things like cooldowns. That way you'd have a situation where the beginning of the fight is more to get into a good position for when the rage is flowing. It would also mean that someone being focussed is going to be juiced up pretty quickly. A few fighting games work a little similar to this, you could be putting the beatdown on someone pretty hard but then you get whalloped because their guage filled up and they caught you with a special.

[eldaec]

It also makes the burst damage vs sustained damage distinction meaningful.

And shamelessly borrowing from M:tG again, would allow at least three distinct approaches to winning....

Aggro mages : Who attempt to win by loading up with as many small spells as possible, and casting them as quickly as possible.

Combo mages : Who do not very much to other guy and hope to stay alive while desperately building up resources as quickly as possible for a massive combo attack.

Control mages : Who seek to prevent enemy win conditions (possibly through healing, counter magic, whatever) to exhaust enemy resources and extend the battle into an end game, and then, having planned for a late game, have a few of the most powerful spells available to make use of all the mana swilling about.

To me at least, this is a lot more interesting then endless red fireballs vs infinite blue iceballs.


Feel free to switch mage for warrior, spell for ability - it's all the same.

[Hoax]

I agree with you, in fact I meant to post earlier that artificial limitations on how much dmg a person can take per second or how many people can attack a given person at once are dumb.  I forget if I did or not.

There are plenty more things that can be done by a clever game dev imho, we've come up with plenty of ideas that seem to get around the current set of tactical stagnation.

Mostly though I think it usually comes back to the problem is having your pvp in a pve world.  There are times where I feel that every skill would need to have two sets of effects one versus mobs and one versus pc's to actually get any kind of good gameplay in such an environment.

I've found a majority of defensive skills suck balls in pvp because if they didn't you would be invulnerable in pve.  Mobs will never throw out the kind of damage (not to mention the cc's and burst nature of it) that players will.  No matter how you design players will find a way to do a large amount of damage, as burst as possible, to a single target.  Like eldaec says that is just the smart way to go.

Perhaps if someone designed the pvp first, then created npc statlines and attacks to counter the strength of the pvp skills...

To my knowledge that has never been done (see Prot Warriors in WoW for one glaring example) in any game.  Instead the game is released with certain skills being stupid in pvp and slowly they have to re-balance classes for pvp as a result.

I think that is the major flaw in the current system you have to account for first.

[Lightstalker]

Yeah great - what post UO MMOG is that?

Seriously, every dev group knows that people dying before they can respond is bad.

They usually solve the problem by having PCs unlikely to die without being smacked several times, and they usually give support characters skills to react to it.

I was going to let this go but then some other server in my WoW battlegroup came onto our forums to complain about one of our 10 Hunter WSG teams.  Every 3 seconds they can remove a player with an instant, coordinated, defense ignoring strike.  If their target is a warrior or a druid that target might get lucky and survive the initial burst, but surely won't last another 3 seconds of regular focussed fire (and certainly wasn't targeted before the healers were).  That is an example of Focussed Fire being a very poor game experience for the receiving side.  There was also a video running around where one player controlled 4 Warlocks from the same keyboard.  This was a similarly disturbing instance of focussed fire removing the opportunity to react from the opposition.  Taking it to an absurd level, watch a Hogger Raid video with 40 level 1 gnomes taking out a level 11 elite NPC.  Any trivial amount of damage in WoW can be lethal if applied with sufficient focus.  While Hogger raids are not instant kills, it should be pretty straightforward to extrapolate to an arbitrarily large amount of non-trivial damage applied in an instant.

This (focussed fire problem) is a relevant issue and accurate description of a problem facing contemporary MMOGs, including the current subscription king.  Specific to this thread's context: Shadowbane is an excellent example of Massed Blaster Fire changing in a fundamental way how large scale conflict is carried out.  While we may not always agree on what "every dev group knows" we can make the observation and independently verify that regardless of dev team knowledge this focussed fire issue exists and persists despite efforts (if any) to date intended to mitigate it. 



While many of the ideas for what to do during combat are interesting - they are all vulnerable to focussed fire making them utterly irrelevant (or other common game elements like crowd control).  I'd like to see decisions made during combat matter in large scale combat - rather than strategic decisions like group composition and firing order rule the field.  I too would like to see alternative victory conditions (alternative to health -> 0 due to incoming damage) in a large scale conflict.  The problem is not limited to sniping, or frontloaded damage.  The problem is one of scale, in that without collision detection and/or line of sight any number of opponents can strike any one target at the same time.  So long as you can deal 'damage' you can eliminate an opponent in an instant by massing enough sources of damage under the current MMO models.  That's the nut to crack; while diversifying routes to victory and making combat more dramatic are fine objectives they do not crack the nut.

If focussed Fire cannot be avoided through clever design, then perhaps there are ways to mitigate its impact? 

One such example comes from WoW's instances.  In AQ40 (and to a lesser extent MajorDomo in MC) there are groups of mobs that gain the powers of their fallen comrades.  This makes individual pieces of the Environment Side harder to deal with as the Environment side takes losses.  That's the kind of solution that would work in a world where focussed fire is a reasonable expectation.  One of the breaking issues with Shadowbane GvG conflict was how after a battle the losing alliance was weakened while the winner was strengthened.  There was no (little) practical cost for defending a wider perimiter or advantage to having ones force concentrated in a single geographic location for defense. Attempts made to limit guild power via code changes were circumvented by guilds reaching out of game (around the rules) and forming several in game shell guilds to manage assets.  Designing with awareness of focussed fire is neither easy nor straightforward.

I'm about to lose power, so that means I'm done.

[eldaec]

Yeah great - what post UO MMOG is that?

Seriously, every dev group knows that people dying before they can respond is bad.

They usually solve the problem by having PCs unlikely to die without being smacked several times, and they usually give support characters skills to react to it.

WoW

I take your word for it - I haven't got into WoW so I can't comment.

It surprises me though, because everyone else fixed this just by reducing player v player damage in general and giving support characters quicker action times in order to compensate for being unable to take initiative. It's a kludge, and it doesn't help with other symptoms of the underlying problem, but it does get rid of sniping in every other game. /shrug


I still maintain that if the underlying problem is that a player can be doomed by an opening move alpha strike, then preventing focused fire probably isn't the way to fix it. You need to find ways to restrict opening move alpha strikes. My own opinion is that late game alpha strikes are a legitmate strategy; the 'philosophical' reason opening move alphas are bad and will always create one problem or another is that they externalise the (high) cost of the alpha from the battle at hand - the mana you use was saved up prior to the battle, and the reuse timer 'cost' was paid prior to the battle, typical MMOG combat always gives everyone their first move for free - so it shouldn't surprise anyone that players will seek to get maximum value from this free pass.

The cost of anything you do in a battle needs to be fully paid for within that battle.

Another good example of the problems that arise from externalised costs is Barrier of Faith - a daoc realm ability unique to clerics in the first version of daoc realm abilities.

BoF allowed a cleric to increase the resistance of his group to all melee damage by 50 points for 30 seconds. In practice this would be around a 70% reduction in incoming damage. ie. this would win you an 8v8 almost every time (unless the other team had an equally broken ability). Costs of the spell were as follows....

instant cast
0 mana
0 endurance
30 minute reuse clock
10 realm ability points (one off cost to buy the ability)

...in other words for the first use in a battle every cost was externalised - and you almost never need a second use.

From the perspective of the team using the ability these are genuine costs - if I use BoF now, I probably can't use it in the next few fights. Also, saving up 10 realm points was a non-trivial about of realm point farming.

BUT from the perspective of the team facing the ability, these costs are all free. The other team gives up no time or resources from the pool available for this battle in order to cast BoF. The battle is being decided (or at least dramatically affected) by an ability they can't counter, can't predict, and which did not cost the opposition anything - so as well as BoF they still have just as much healing, nuking, and tanking as they would normally have. QQ much?


Free abilities suck - always. Frontloaded damage is just as free as BoF was.

[pxib]

This is long and rambly. I apologize.

My favorite of the twitch console fighters is a bizzare little game called Guilty Gear. Its designers took a surprising but effective route towards balancing their game. First they created an enormously robust fight framework which every character shares, described below... then they let the artists create whatever sort of fighter they though would be fun. Each character plays entirely differently and has a barrage of attacks completely unique to them. In other words, the moves were designed not because they make intuitive sense or because "everybody ought to have something fast and short range" but because they look impressive. In any other game certain characters would have huge advantages against certain others... many oughtn't exist in any fighting game ever... yet here they are all playable, and all beatable, because the basic framework is so solid.

To summarize:
Attack beats Throw beats Block beats Attack. There are high and low attacks and blocks.
Percentage damage done exists on a continuum (expressed as a sliding scale near your health bar)
If you haven't been blocked, your damage ticks slightly lower. Long combos stop doing appreciable damage.
Every time you are blocked, the scale ticks the other way.
Advancing towards the enemy and attacking her (whether she blocks or not) adds to your "super meter".
If you play defensively for long enough, your super meter depeltes entirely.
Any given super move requires half of your meter, but
You can also rapidly deplete your meter for "faultless defense", which blocks anything COMPLETELY. Your super meter drops over time, not by amount of damage prevented.
You also have a "BURST" meter, which starts full. It can be used at any time to stop an enemy's attack, throw him back a small distance, and stun him for a moment. Once used it must be replenished by taking damage.

Fighting games have at their core what I will refer to as "initiative". Being attacked makes it difficult to attack, so whomever has initiative has a temporary advantage and the abused opponent must play entirely defensively while looking for an "opening" through which he can make a counter-attack and grab initiative for himself. If no opening presents itself, if defense is ineffective, or if the particular defenses and openings available to your character are completely bypassed by the opponent's character... you feel powerless. Worse, you are powerless. Once initiative is lost it can never be regained. Guilty Gear steps around this by providing every character with an easy, universal defense and an easy, universal opening. Faultless defense and BURST.

...from the perspective of the team facing the ability, these costs are all free. The other team gives up no time or resources from the pool available for this battle in order to cast BoF. The battle is being decided (or at least dramatically affected) by an ability they can't counter, can't predict, and which did not cost the opposition anything - so as well as BoF they still have just as much healing, nuking, and tanking as they would normally have.

So what does a 1v1 twitch duel have to do with focused fire and group combat in the latency drenched world of MMOGs? That bit I highlighted from eldaec. The problem isn't the ability, it's the framework. When defense and offense happen at the same time, the group with the most of either gains an unimpeachable advantage. In short, there's no initative. Rather than defense being reactive, an effort to passively shield the group from assault while they look for an opening, it becomes proactive: a cloak the group wears while they actively  beat their opponents black and blue. This will not do.

I propose the following:

Defense and offense must be mutually exclusive, but defense is the default... the first attacks against an unaware player (automatically defensive) must always be the game equivalent of a "throw" (picky and/or short range) or they'll block it. Every unblockable "super move" must require a period of effective combat before it can be used. Every character needs a limited "OH SHIT" button to universally counter any given "I WIN" button, and thus any I WIN must require enough concentrated time to allow the opponent to OH SHIT. Damage is best done incrementally rather than in instant chunks. If there is a chunk it should be at the end of a long incremental, and defense ought to be able to begin anywhere within that increment.

So one player can defend effectively and take minimal damage from a great number of attackers if she decides not to attack. Sometimes she can even make the decision to go defensive during an attack because most heavy attacks will start with a light damage warning sign. If she remains defensive, the enemy players will gain combat points towards their unblockable super attacks which, once unleashed, will offer the defender no choice but to hit the OH SHIT button and make a break for it... and if that proves ineffective she's done for.

This process still sucks if you're all alone, but it ought to take enough time that your friends on the field can see what's going on and come to your aid. Every enemy concentrating on attacking you cannot also be defending, so they'll be vulnerable to anything that comes their way.

Err on the side of defense, and the game stays fun. Attacking a defending player is a little irritating, but not nearly so irritating as being defenseless.

[Krakrok]

What if by joining a group your gave up your own ability to move and the group leader would move the entire group in a formation. Only people in groups could attack (solo players SOL). So you have your Roman legion formation and the leader of the group moves the entire formation.

Each player still has to attack individually. Each player can break out of the formation and run but can no longer attack if they do. Each player at the front of a formation also has the option to fall back to the back (only the front players can do this) and stay in the formation. The person behind them takes their place at the front. So when the person at the front gets beat to crap they move to the back (via a button push) and the fresh person takes their place. If the fresh person had already fought they'd have the same HPs as when they left the front before. Repeat. Formations could get whittled down this way.

Possibly if your formation destroys enough people your group leader would get a berserk option which would allow the formation to continue to exist (and therefore do damage) but allow individual player movement for X period of time. So if your formation has significantly routed your enemies your leader can berserk and everyone can give chase. The group leader could also have different formations which the players would automatically shift to when he choses that option like wedge, line, square, etc.

Single 'scout' positions away from the main formation with limited player roaming might also be an option.

Maybe not everyone in the formation has to be a PC. Or PCs would always man the front of the formation and only the characters/avatars would cycle to the back.

---

Which leads to my second idea which is a futuristic version of the formation idea. Everyone is in a tank the size of a turret. If a player joins a group with another tank the group leader becomes a 'host' tank (think treads only) and the second person becomes only a turret (think gun only) on the 'host' tank. Each new tank which joins the group becomes a new turret on the 'host' tank. The turrets are spaced out so the more you add the more field of fire the host tank has (each turret has a different field of fire). Each turret on the host tank is player controlled. The group leader moves the host tank and everyone else can only fire their turret (with it's limited field of fire). Each new turret makes the host tank move slower. Individual turrets can be blown off the host tank.

Throw everyone in the battlefield. Let them form up or group up however they want to win.

Crude example of the tanks:

@=
---

@=
---

@@=
-----

@@=@=
----------

=@@@=@=
  -------------

[Trippy]

What's the advantage for linking up? If you get slower as you add more I don't see how that's an advantage over separate speedy individual tanks, especially since the turrets aren't linked up so you aren't automatically getting focused fire.

[bhodi]

Invincibility for the primary tank bulk; you have to blow the pieces off while the whole can fire. At a specific size, it could be an unstoppable juggernaut as people may be able to respawn and connect at a faster rate than they can be blown off. You'd want this for your vanguard to break a defensive line, to absorb some amount of punishment during the close-distance period, whereas small individual tanks would never make it.

Also, http://en.wikipedia.org/wiki/Ogre_%28game%29

[Pococurante]

Still not following - seems a swarm of one/twos and a few threes would overwhelm the ogre build using the repleneshing benefit of raid and retreat.

[sinij]

How about shared pool of hit points during group fights with only portion of damage dealt to individual target and rest getting distributed to other participants?

[eldaec]

How about shared pool of hit points during group fights with only portion of damage dealt to individual target and rest getting distributed to other participants?

What problem is this intended to solve?

[Trippy]

We're trying to figure out how to make it so that with Krakrok's link up idea people would actually want link up rather than move and shoot on their own.