The Theory Of
Here you'll find discussion of all manner of topics relating to the theory of multiplayer games. As I see it, anyway. A note to commentors: if you stray off-topic or if your reply contains ad hominem attacks, your comment will be deleted.

Synopsis: Chess and Cards are predicated in creating a fair contest of skills between players.  MMO PvP is typically predicated in encouraging players to spend a lot of time in the game to build asymmetrical advantage over others; to produce a stacked deck.  A system that draws from the lessons of Chess and Cards is presented in effort to provide for inherently-balanced MMO PvP.

I was reading vajuras's article "Static Class Based Games and Their Flaws" and found that I had a bit more to say about the topic than I cared to push into a comment.

When thinking about PvP and making it work in a fantasy setting with player avatars in combat, I've come to the realization that PvP games can work one of at least three ways.

1. Like Chess, where all players start with the exact same pieces and they then attack.  The emphasis is on the attack.

2. Like Cards, where all players start with a random set of pieces from a shared pool, spend some time jockeying to get better pieces from that same pool, then when a player believes they have a good hand, they attack.  The emphasis is on the jockeying.

3. Like MMOs, where each player spends time and energy prior to the game to build the pieces that they want to bring to the attack, then attack.

There are variations on these themes and there are others as well.  The point here is that of the three presented, I find the MMO formula the most ludicrous.  Imagine if you were to play Chess against someone and they had eight queens simply because they played more Chess.  Or if you were to play Poker against someone and they were known to play with a Royal Flush.  That they bought on a web site.  It seems rather idiotic to me, but I'm a fan of the Good Fight.

If PvP is going to take place in a fantasy setting, then I say structure it such that characters no longer have static skills and equipment.  Have every character enter each PvP encounter in vanity items only; fancy clothes, jewelry, swagger stick, whatever.  They then are either given the stock equipment and skills for the PvP encounter (as with Chess) or they are given a basic set of equipment and skills to start with, and then the game transitions into a period of rushing around trying to gain additional equipment and skills.  When one side thinks they have a winning combination, they start their attack (as with Cards).

Here's a variation on that: permit one piece of static equipment or skill on each character, but the static equipment is taken from the shared pool.  So if one person on each side of an encounter chooses the healing skill, then they each get half of the healing skill.

Something that may not be clear up to this point is that each PvP encounter starts from scratch.  You may get a great setup during one encounter, but once it is over, you are back to square one.  You may have gotten a win out of the last encounter, but it doesn't roll over into the next one.

Where's the fun in that??

It lies in playing good, fair games instead of trying to get eight queens so you can dominate your opponent in Chess.

Where does this leave the whole notion of sandboxed, free-for-all PvP entertainment?  Right where it has always been: in a niche.  Persistent gains and PvP do not match up.  This is the linkage that Jimmy_Scythe pointed out in his very good article "FFA PvP Part 3: Tracks in the Sand".  It is the reason that I believe that as soon as someone creates a PvE-only competitor, the Eve Online player base will be rather seriously depleted.

PvP must be focused on the encounter.  Players must be permitted to adapt to changing circumstances, understanding of how the encounters play out and so forth.

Note that if PvP focuses on individual encounters, then the rules of the game can be varied from encounter to encounter.  During one encounter, it may be decided to leave out all fire effects.  Players would have to adapt to that lack.  Players who favor locating fire effects and using them in their PvP would be disappointed, but if they are adamant about using fire then they just don't agree to the rule change.  But they'd do it out of preference, not out of the fact that their static character that is founded entirely on fire effects is now useless.  They have the opportunity to adapt.

Can this style of PvP have any role in a persistent world?  Sure.  The persistent world could be fought over by the various PvP factions.  Those factions would have to establish rules for encounters, however.  I've proposed a system before where a faction can declare that it wants to battle for a certain stretch of ground at a certain time with a certain number of players using certain rules.  The faction that controls the adjoining ground can either accept, reject or counter that proposal.  If they agree, they can battle it out and hope to win using fair rules.

For example, the Lean Machine faction borders three other factions.  One of them, the Knights Tumbler faction controls an NPC town that makes a certain style of armor.  The Lean Machine guys get armor according to the normal rules of encounter, but the way it looks is determined from the armor-crafting NPCs that they control.  And this particular NPC town makes very cool looking armor (that has absolutely no role in character effectiveness).  So they challenge Knights Tumbler for the town, placing a chunk of their own territory on the block in balance.  Maybe that land has some resources that the Knights Tumbler would like to get.

So one way or another, the two factions agree that it will be a 10v10 encounter using the standard rules, starting in 15 minutes.  The time arrives, and the two groups enter the contested area.  They play to completion, and the winner takes all.  Knights Tumbler loses the town but calls foul because one of their guys had to deal with an emergency at home.  They want a rematch.  Lean Machine ignores the call for a rematch and looks forward to wearing their new armor.

Consider too that a faction may build its homes (full of trophies and vanity stuff) in a particular area that it captured from a neighboring faction.  If that area becomes contested, they might lose all that stuff.  They could move it out before the fight, of course, but whatever is permanently fixed would be lost.

Why would anyone enter battle?  Because of someone betting something that you want - or your battle timer running down.  A faction is obligated to enter battle once a week (or some amount of land per week, whatever).  If they do not, then they sacrifice some standard chunk of land to each neighboring faction.  That continues until the faction is swallowed up by its neighbors.

My prior article in this series, "Making PvP Work: Ranking Players" still holds; when players vary in skill, the players of widely disparate skills should be isolated from one another.  In a faction system like the above, players who are bumped up to a higher rank would have to find a faction to join.  Factions would always be looking for good players so that they can hold onto their gains.

I could go farther into this whole system, but the important point here is that I believe that PvP should start with either a level playing field, or one that is randomly canted per encounter.  That makes it fair for all players who are in the game.  It also means that player skill and chance become the keys to success.  Not stacking the deck.

Synopsis: MMOs are designed to be intertwined single player experiences.  A means of restructuring game play to make large group interactions the norm is presented.

For some screwy reason, MMOs are designed to be intertwined single-player games.  Each player enters any given session of a game with a unique agenda born of that player character's current level, set of quests, location in the world, current faction with NPC groups and other various and sundry factors.  That combination of factors constitutes the game state for a given player.

In order to play with other players, those factors must be aligned.  The players must have the same geographic location, be at about the same level, have the appropriate faction, be on the same quest and so on.  If they want to experience the same content, they must delicately converge their single player game experiences.

This is not conducive to experiencing the game with other players.  Being on the other side of the world impairs their ability to play together.  So games invent teleportation schemes.  Being on different quests impairs their ability to play together.  So games permit sharing of quests.  Still other factors, such as level disparities, insufficient faction, lack of qualifications and so on can all impair the ability of players to play together.  There's a whole bunch of factors involved.  Those factors are part and parcel of the achievement gameplay ethic, and they get in the way.

With so many factors shaping the way the game can be accessed by a given character, players end up pursuing highly-individual goals.  They want to advance a level, get certain faction, obtain the next upgrade in equipment, and so on.  Even when the players find a common ground, they may quickly diverge because of the great number of personal goals that players can pursue.  It simply seems odd to me to have attainment of personal goals as the core of a multiplayer game.  It may be nice to see other people playing the game at the same time that you are, but I know that I would enjoy playing the game with other people.

I can get that in current games, but I have to work for it.  I'd really rather have a game that naturally brought players together on common tasks.

This takes me back to the notion that I brought up in "Raids for Everybody"; when players enter the game, they can choose between the various global tasks that players are tackling.  A game might offer a few combat-related tasks, a few crafting tasks and a few exploration tasks.  But the goal of all those tasks is to have the players work together.  That is, all the tasks are larger than any given player.  To join in doesn't require an invitation.  You simply move your character into the activity in question.

For example, the classic military conquest scenario: the players are up against a fortress or a warren of caves with lots of monsters.  It is up to the players to collectively clear it out.  A player interested in combat joins in on that.  There might be four or five such challenges going on simultaneously, but players don't group.  They don't raid.  They just go where there are other players tackling the same problem.

When that problem is solved, the game advances to the next combat problem as part of an overarching story line.  Perhaps there are roving bands of monsters in the plains, and the players must bring their baggage train through the plains.  As the train slowly moves along, the players are obligated to defend it.  If nobody defends it, it doesn't move.  In fact, the group as a whole may begin to suffer because the train is being attacked.

Those are the few open quests that all players interested in combat are involved with.  There are no levels.  The equipment permits modest variation in character effectiveness.  Modest wealth can be accumulated for vanity items.  The point of the game is to have the players interacting, fighting side-by-side, alternately healing each other, getting into sword fights and firing their bows as circumstances dictate.

Players don't have personal agendas that can conflict.  The range of agendas is artificially limited such that the players stay near each other in all the ways that standard MMOs permit them to diverge.  They are kept near each other, both figuratively and literally.  They don't diverge in personal power.  They don't diverge in geographic location.  They don't diverge in faction gain.  They don't diverge in equipment.  If divergence is permitted, then the high end players will want high end challenges, and the low end players will be unable to participate.  So the point is to keep the players bunched up.

If they are bunched up, then they will tend to compete for whatever is available, right?  So don't give them much to compete over.  Keep the rewards global to the group.  When a fortress is captured, the magic dingus in the fortress becomes available to everyone.  The magic dingus grants a +1 on any weapon that you put on it, and that +1 stays with the weapon for a week of real time.  Now the players have a nice reward for having captured the fortress, and everyone gets it.  Or perhaps everyone is simply paid the victory money of 1 gold piece that they can spend any way they want.  Of course, their choices are mostly limited to vanity items and a few equipment upgrade choices.

This technique applies to the crafters, explorers and every other activity that an MMO can support.  Whatever the activity is, it must be scaled up until it can accomodate as many players as care to join in on it.  The crafters can build whole towns according to some master plan formed by the developers.  It's up to the crafters to do the building.  The explorers would not scatter to the four winds to do their exploration, but would group together into a few caravans in an effort to find certain far-off cities, lost civilizations and such.  They have to solve vast puzzles with many small pieces in order to do that exploration.  What they find may enable the other large groups to change the way they experience their content.  But no single player gains by the exploration.  Everything is a group effort.

To put all this another way, there is a modest set of grand quests that all players join in on.  It might be that players have a character in each grand quest, and can hop between them as the mood strikes them.  That way, whatever configuration changes can take place on a character need not be compromised.  The explorer character can be pure explorer while the combat character can be pure combat.

The net result is not an MMO that is not a spaghetti ball of intertwined single-player game experiences but rather an MMO that is a set of large group game experiences.  This is where I believe MMOs must go.

As I mentioned in the comments for "Raids for Everybody", Warhammer Online begins to approach what I'm talking about with their common/simultaneous quest mechanism.  Unfortunately, it falls short because the game retains a focus on individual accomplishments.  I'm advocating grand quests, which relegate individual rewards to a secondary role; it's nice to get a temporary bonus for my weapon.  The real entertainment of the grand quest is being part of a vast undertaking that many players are taking part in.

It's just fun to storm the castle.

Synopsis: Player characters can possess special senses that permit them to see the game environment in a way that is not based on visible light.  Examples are given to illustrate the idea.

Fantasy MMOs usually just present a fantasy environment as someone walking off the street would see it.  There are trees, rocks, grass, buildings, etc.  To be able to see a 3D environment in great detail in real time is pretty amazing stuff, given where we were just a couple decades ago, but games have the opportunity to go beyond that.  Fantasy MMOs already show things that we don't see in real life, such as sparkling and glowing effects on various and sundry things.  Magic.  There is also the obligatory translucent character ability.  Stealth.

It occurred to me years ago that games have an interesting opportunity to make a game out of what a player can see through a character's eyes.  Suppose warriors could see other characters in terms of their fighting skill.  One orc looks pretty much the same as another, but a warrior would be able to sort out that this orc is more skilled than that orc.  Or a warrior can appraise an object as a potential weapon.  That chair would make a 40-or-so-DPS mace.  Or consider a mage who is trying to locate items of magic in a room.  Surely a mage should be able to sense which had magical abilities and which did not.

The knee-jerk implementation that I'm sure we'd see from a game would be that the player would click on an orc to see the numbers associated with its fighting ability, or click on an item in a room to see the numbers associated with its properties.  That is, unfortunately, a result of network bandwidth problems.  A player can only ask for only so much information from the server.

Suppose that wasn't a problem.  What if a mage character could alter its vision to see the world in magical terms?  The entire world would turn translucent, except for objects that were magical.  They would remain rather opaque, perhaps glowing in different colors to indicate what sort of magical potential they held.

The same treatment applies to the warrior.  If the player asks for it, the warrior character's senses can be enabled to instantly appraise the 30 orcs in the enemy group.  That one is bright with skill and the rest are pretty dim.

The player isn't obligated to click on everything in sight.  He simply says that he wants to know what's significant around him and he sees it.  I know that I'd love to get that from a fantasy MMO when I was trying to figure out what objects are going to give me some kind of useful behavior.

And sure enough, Blizzard is savvy to this treatment.  When I went back to World of Warcraft after a long hiatus, I noticed that they had added sparklies to objects that were pertinent to a quest.  The game called the player's attention to the object because (in theory) the character was looking for it.

But I'm actually looking for something rather more visually dramatic.  Again, Blizzard has the best implementation that I've seen.  When your character dies, it is dropped off back at the nearest graveyard.  What does the world look like?  It's different.  It's the world as seen through the eyes of a dead guy.  You can't see anything living until you get near your body, but you can see dead people.  This is precisely the sort of visual alteration that I'm talking about.

Suppose you had Thief Vision.  The world turns ghostly white, but small objects with great wealth associated with them sparkled and drew your attention.

Suppose you had Priest Vision.  The world turns ghostly white, but the alignment of characters relative to your priestly order is indicated in sparkles and glows.  You can spot a dirty, rotten paladin who is opposed to your wonderful and marvelous black arts order.

Those who have always wanted to wear something small to indicate their membership in an organization could have that.  Use Organization Vision to spot people wearing your organization ring, pin or brooch.  Or to spot someone in an enemy organization.  Or someone that you're trying to meet.

How about the archer who is playing an MMO that relies on player skill to place bow shots?  He's going to need some kind of vision to give him accurate distances to targets, wind conditions, perhaps even some other factors.  He gets all that by dropping into Archer Vision.  The quality of that vision could be made dependent on the specialization of the character, making the player's success a combination of his own skills and his character's.

This vision system places emphasis on the character skills, or class, or profession.  Whatever specialization a character has, it would gain the appropriate visions.  If the character can have greater and lesser degrees of specialization, then the vision that is gained can vary in accuracy, quality, etc.

This is all just an adaptation of real world physics.  Not all energies are in the visible spectrum of light, and being able to see in other parts of the spectrum can grant a significant edge to the viewer.  Think about seeing in the infrared.  Military night vision and seeing through fog.  Or seeing in the ultraviolet.  This is all about sensors.  It seems more naturally part of a science fiction MMO, except that the visuals are far more entertaining in a fantasy MMO.  Space is mostly empty, while a fantasy setting is quite rich.  There's lots of potential information to collect.

That is, of course, the great problem: how to get all that information about all those objects and characters in the field of view from the server to the client.  It can be done in basic ways today, and I'm sure a sufficiently clever engineer will figure out how to adapt many cross-sections of world data to this technique.

I, for one, would very much enjoy being able to see a bit more information about the world that is specific to the role I've chosen for my character.  That additional information could make my membership in a group all the more valuable.

Synopsis: Current games throw all players into a single arena and let the strong prevail.  By ranking players, games ensure that they find more balanced competition.

Hrothmund prodded me about returning to my blog, and I'll use the opportunity to comment on a point of design that one of his recent articles suggested to me.

There are a variety of motivations for players who like to experience PvP systems.  They vary from domination of other players to personal challenges to just being part of a good scrum.  Being the Christian soul that I am, I'll focus on the latter two because I believe they are a sign of, and permit reinforcement of, maturity in players.

Consider PvP in an MMO.  There are thousands of players interacting.  But the best players are interacting with the worst players.  The worst players are getting pounded on a regular basis.  Is that helping the worst players to get better?  For some, it's the perfect motivation to get serious about their PvP.  For others, it's a perfect opportunity to quit the game and find something that they can enjoy.  However, in the interest of entertainment, we want both players to stay.  How can we do that?

We get those groups of players away from each other, in the spirit of design that I advocate in "When Flocks Collide".  In this case, the flocks colliding are the players of differing skill levels.  The solution is to create leagues.

Leagues are a time-honored way of organizing players of disparate skill levels so that they can always find a good game.  After a fashion, traditional FPS games structure into ad hoc leagues.  Visit one of the hundreds of servers out there and see if you can get a good game out of it.  If the opposition is too weak or too strong, try another server.  The general tendency is towards the formation of leagues according to player skill.

But clearly not always.  They are only ad hoc leagues, and one or two skilled players who are more interested in selfish domination rather than taking on their skilled peers can destroy the skill balance in any given server's games.  They simply enter a server that is frequented by less-skilled players and start lopping heads off.

Unreal Tournament Onslaught has been my poster child for PvP because I enjoyed it so much.  I could usually find a good fight on particular servers.  Yet we had the occasional marauding elite players who would come in and stomp everyone else, rendering the game pointless for the rest of us.  The one thing that Unreal Tournament could have used was a player ranking system.  Rank all players into leagues from 1 to 10, then designate each server as accepting players from a specific league.

So how can leagues be created in an MMO?

At one level it's straightforward.  Rank players, thereby organizing them into leagues.  Rankings are assigned according to a player's ability in the game, with rewards going to the more highly ranked players in ways that will motivate them to seek out high ranks.  World of Warcraft does this with titles and special items that rank can bring.  Those motivations exist partly to discourage players from trying to become ringers in lower level leagues.

Once leagues exist, have the teams in those leagues face off in the MMO environment.  They can challenge each other, have little wars, races, build-offs, whatever the competition might be in the game.

At another level, however, it's rather more complicated.  How can all these leagues coexist in a reasonable way within an MMO whose fiction is supposed to be maintained?  Why doesn't the rank 10 league team known as "The Elite Troopers" never declare war on the rank 1 league team known as "Bob's Pickup Team"?  What if a member of "Bob's Pickup Team" insults one of "The Elite Troopers"?

Instead of introducing game mechanics to permit the game to recognize insults, separate the players of different ranks.  There could be whole servers devoted to supporting leagues.  There would be the rank 1 league server.  It is where all the rank 1 clans play.  If you get promoted to rank 2 because of your successes, you'll have to move on to the rank 2 server.  That continues until you reach the rank that matches your potential.  That means that "The Elite Troopers" aren't in the same universe as "Bob's Pickup Team".  If they want to insult each other, they'll have to do it on the web.

Some players could certainly play intensely for months and maintain a high rank, then back off on the game, lose some skills and find that they have to play a rank or two below their peak.  It would mean no longer playing with the same group of people, but that's the nature of league play.  After a fashion, the people who play at your level tend to be more or less like you with regards to the game.

What about that pernicious evil known as "alts"?  Those skilled players who insist on being ringers for their friends can simply fire up an alternate character and reenter the MMO at rank 1 again so that he can ensure that his friends' team always wins.  In brief, there's no way to address this until the game can successfully identify the player.  Players will be ranked, not characters, so player anonymity rears its ugly head again.

Now I understand that this is a dramatic departure from the stock treatment of PvP in MMOs.  It would require at least one server for each ranking, and as the player demonstrates more skill, they would move from server to server.  That's a pretty dramatic departure.  However, players should tend to stabilize at a certain rank.  A given game may require only a few ranks, making movement between servers infrequent.  In a game like Eve Online, it would mean dividing the player base into multiple instances.  I would view that as a good thing, given that it would mean that the more casual players would be able to fight more casually while the elite players could lock horns as rough and tumble as they like without the casual players complaining about their behavior.