I make apps for other people

GITS:SAC MMO design (revised)

Posted by Chris Jones
On April 6th, 2006 at 21:38

Permalink | Trackback | Links In |

Comments (7) |
Posted in Cyberpunk MMO

Ghost In The Shell:  Stand Alone Complex Gig 2My wife reviewed my Ghost In The Shell: Stand Alone Complex MMO design and had a lot of good suggestions. The next revision of this may remove the GITS:SAC theme completely in favor of a “generic” cyberpunk setting which could then be expanded in directions not present in Masamune Shirow’s creation.

The design addresses an alternative way of handling gameplay for elder players. One question that prompted it was “how can the game continue to provide interesting gameplay when end-game players are much tougher than most anything else in the game?” Part of my proposal is to extend quests and missions by providing greater variety in type and content, rather than relying exclusively on raiding or PvP content.

The anime series GITS:SAC provides an answer to continued end-game content: investigation and negotiation. Simply because the player has capped his character’s rank, body, and skills doesn’t mean that the game had ended. Investigations into crimes and terrorism are still needed, and while individual NPC opponents are outclassed, the scale, consequences for failure, and need for finesse are greater. Missions can be designed such that as a character becomes more powerful, the player needs to exert more self-discipline and control to succeed.

At all times, from the early ranks through the end game, players can choose what type of mission to perform, if it will be PvE only or PvP, and if it is intended for a solo player, duo, or group. Missions may include:

  • defending NPCs from obvious assaults while using lethal force (escort, wave attack),
  • injure, incapacitate, or capture enemies (subdual),
  • stealth and information gathering (infiltration, non-combat),
  • stealth and assassination (infiltration, combat),
  • evade enemies (evasion),
  • chasing enemies (evasion, timed mission, exploration),
  • interrogation (subdual, negotiation, trickery),
  • investigation (infiltration, exploration),
  • intimidation (combat, trickery),
  • disguise (infiltration, trickery),
  • item recovery (capture the flag),
  • bomb defusal (timed mission, exploration),
  • guarding an area/door/item (combat, intimidation),
  • and just for fun, heavy combat.

(Other ideas: fed-ex, since all games need delivery quests, and healing/first aid, essentially the same mission as item recovery.)

Mission success is based on achieving primary objectives. Additional objectives may be part of the mission (either advertised or not, a secret bonus). Mission rewards can be increased by working effectively as a team, having no characters killed or injured, or other efficiency and social triggers.

Mission rewards may take the form of increases in rank, cash awards, item awards, skill awards, and faction adjustments. Player can abandon missions at any time: missions track progress and certain points indicate partial success or partial failure (with commensurate rewards). Faction adjustments indicate how NPC members of various factions regard a player. Player characters belonging to a particular faction also get an indicator of the other player’s relative standing with the faction.

Would it be better to hand-craft threats and investigations as most quests are done in MMOs today or should a procedural process be defined to let the server generate them? At the least, details would need to change (NPC names, ages, sexes, locations, and roles) with each assignment so that players don’t walk through the rank-appropriate material. How should characters receive changes in rank (number of successful missions, the degree of success in a given mission, time spent in game, completing optional objectives in a mission)?

Depending on the type and scope of failed mission, the impact on the world could be greater or lesser. A government player character attempting to limit criminal access to military equipment who fails a mission could find not only criminal NPCs equipped with better gear, but also non-governmental player characters could find more available and less expensive military gear available from their fences. A high rank government player character who fails to prevent a building from being destroyed may face not rank gain penalties (rank debt), but a change to the game environment (replacement of the building model with ruins, later replaced with new construction and eventually a different building model or texture).

Death penalties are:

  • the time required for the character’s ghost to adapt to the new shell (generally 30 to 60 seconds, or longer if it takes longer to get into combat or the next mission, subject to balancing and playtesting), and
  • repair and replacement costs for bodily damage.

Failed mission penalties are:

  • first, the failure of a mission unless other agents or officers were sharing the mission and completed it (a psychological penalty the player may impose on himself),
  • second, monetary impact, including the loss of mission rewards,
  • third, slowing down rank gain (rank debt–capped at the equivalent of three failed missions), and
  • finally, limiting access to certain items for purchase from NPC sources while the player has recently failed a mission (the most prestigious items, limited for 24 hours).

The risk of failing missions may be enough to encourage players to group. Death itself will become increasingly rare for players as they gain ranks and tougher cyber-bodies.

Unique PvP
Player versus Player activities can take place in the context of players belonging to one of multiple organizations:

  • Section 9 (Public Security),
  • Section 6 (Ministry of Foreign Affairs),
  • various crime syndicates and terrorist groups (in-line with themes from GITS:SAC),
  • refugee or aid groups, or
  • foreign intelligence agents.

Players can be put against each other in missions (which will always be consensual):

  • different ministries or investigation teams attempting to complete a mission to get the credit and discredit other ministries),
  • law enforcement PCs attempting to stop criminal PCs,
  • terrorist PCs given orders to prepare for an activity, and law enforcement PCs are sent to stop it, or
  • foreign intelligence agents seeking entry into government ministries, where PCs are to stop and capture them using non-lethal means.

Some missions may require player characters from different organizations to work with each other against similar teams of player characters from multiple organizations.

A player economy
Higher rank player characters may have access to virtually unlimited funds: cyber bodies can be replaced, whatever weapons or special equipment is needed can be obtained, and as long as the character has a record of success in missions, there won’t be any real problems with unavailability. Lower rank characters may have to work within a budget.

Methods for obtaining funds and items should be established through missions for most characters. Each faction should have access to relatively identical (in effect) items, skills, etc., per rank.

As part of the mission rewards, players will receive cash payouts for successful missions. The money can be used to purchase off-duty lifestyle items (available to all players): housing, civillian clothes, and status gadgets. All factions need to purchase weapons, equipment, skills, etc., at least when they are not provided as part of the mission rewards.

Government player characters may wish to sell government equipment to non-goverment player characters. If discovered (and they will eventually be discovered–lost equipment may cause them to gain rank debt or be investigated, possibly by players from another section), they will be given a basic cyber body (no better than a normal human) and discharged from government service–at that point, the player will have to start over in rank in a new non-government organization, likely without their cash. In essence, this is a way to reroll a character.

Player characters can perform certain missions to betray from one faction to another. This should be limited in accesibility (a character may only be able to betray to a faction once).

Player characters can sell goods directly to any other player character, through fences to pawn items for cash, or through brokers for off-line selling.

Grouping and guilds
Characters working together in a group or team have a higher chance to successfully complete most missions. They will share a common chat channel and will appear to each other on any maps displayed. When a mission is successfully completed, the reward will be evenly split among the players on the team with a bonus based on the number of characters in the team ((reward / characters) + (5% * characters)). Teams with many characters are more likely to complete bonus objectives and will tend to receive higher rewards for that as well. Teams should be between two and five players for most missions, although solo missions must also be made available.

Characters belong to a meta-guild through their organization, sharing a chat channel.

Finally, characters may create two different kinds of guilds: families and clubs.


  • are the smallest kind of guild,
  • require at least two player characters to start,
  • cost no money to create,
  • have only two officers at any time,
  • have the least amount of control over the shared communication channel,
  • have the least number of control mechanisms over how resources are shared (everyone can deposit and withdraw from the family funds),
  • share a last name, and
  • can belong to different organizations.


  • are larger and require at least five people to start,
  • require a modest amount of cash to start,
  • provide tools to allow better officer management (with multiple levels and roles),
  • provide tools for chat channel management (mute, for instance),
  • have more control over sharing resources (tied to levels and roles),
  • provide fun benefits (such as unique outfits and sequences of animations for secret handshakes),
  • have names that are visible as a tag under the player’s name, and
  • can belong to different organizations.

Players may belong to one family and one club simultaneously.

Other social tools
The net in GITS differs from the Internet of today: it’s primarily a way for various machines and ghosts to communicate with each other and query for specific data. Websites and entertainment are secondary. In game email is provided, as are limited message boards (which may be posted to be subscribers of the game from a website forum). Players may elect to allow pluggable support for Teamspeak or Ventrillo in the game client.

To get a feel for the city of Newport in Ghost in the Shell, download this clip from the film. (26.8 MB, mp4 Quicktime)

Characters with cyber bodies can remotely communicate over chat channels or private tells either through thoughts (cyber brains) or subvocalization (organic brains). Neither has any animation associated with it. Characters with normal bodies can communicate through phones or walkie-talkies, which do have associated animations.

Private vehicles are an important way to get around town to various missions. Characters who can’t afford a van, car, or moped will have to make due with taxis, trains, busses, people movers, and boats. In order to speed transportation between different districts of the city, private vehicles should be able to enter an intra-city highway and be instantly deposited in the destination zone or section of town. NPC road traffic should adjust to player activity: if PCs are driving, traffic should be made relatively light. If no PCs are driving, traffic should become heavier.

Environments should be moody: between missions, characters may experience bright sunlight and crowded, colorful NPCs. During missions, characters have dark clouds and heavy rain, or the neon city lit under a black sky.

Signage should be polyglot, depending on the part of the world or city players are in. Japanese, Chinese, Korean, and English share many signs, although Japanese and Chinese tend to edge out the other two languages. A little German, French, and Spanish may make its way into the game, but not as prominently.

Background music should vary based on the mission and situation. Between missions, players will hear upbeat pop-like music in public places, soulful downbeat music in bars, and busy music while travelling. On missions, driving techno, metal, and suspense filled themes can be used to drive player reactions (a suspense theme while a character is using thermoptic camoflauge to sneak into a building; metal for combat; techno for a chase through a crowded street or from rooftop to rooftop). Likewise, background sounds should be appropriate: quiet rooms with a quiet air conditioner, industrial areas with heavy machine sounds, crowds with a confusing babble of voices, etc.

Players should be encouraged to spend their money on apartments. These provide a way for players to show off character accomplishments and spend excess cash. An in-game directory should allow player characters to find other player apartments. Player housing should always be near transit hubs (such as where trains and busses intersect at a common stop). In general, player housing should be evenly distributed throughout the city and highly accessible–it should take no more than five minutes of real time to get from one apartment to another.

Non-combat pets vary from traditional dogs and cats through cyber varieties of normal pets. Combat, medical, surveillance, and stealth pets may be purchased and can be commanded by the player to assist in missions. The forms of combat pets may range from the functional (boxes on wheels), to those that mimic natural forms (cyber wolves), to the fantasic (miniature dragon spy drones).

Cyber bodies do not necessarily need to be humanoid. They may take the form of a vehicle (limiting access to missions in buildings, underground, in vehicles, etc.), humanoid animals, chromed bodies, transforming animals (humanoid/animal-like forms), etc.

Downsides and risks
This game could be removed from the GITS IP, although names and some concepts would need to be changed. I don’t know what the cost of licensing a GITS:SAC MMO would be, or if the current rightsholder has already sold the rights (possibly as part of the FPS game).

In terms of a second generation MMO, this would have several features:

  • world deformation,
  • player impact on the world and economy,
  • avatar customization (gadgets and clothing),
  • multiple social channels, including ones that lead out of game,
  • player housing and housing customization (as in EQ2),
  • in-game vehicles (it sounds so easy. . . ;)), and
  • alternative gameplay (non-combat or non-lethal missions)

It misses in terms of:

  • player world customization (no player towns),
  • limitless world (a big city, but it may still be considered to have limits), and
  • non-repetitive gameplay (many varieties, many changes, but the core is still mission/quest based).

Finally, the biggest risk (non-technical) is the setting. Most players seem to prefer a medieval fantasy setting, at least based on market figures. The audience for anime is growing, but if this is not or cannot be marketed as an anime game or can’t tie into the GITS:SAC name, it will be more difficult to introduce the game to otaku to generate buzz. Mitigate this risk through aggressive marketing, with the identification of key opinion makers in both the on-line gaming and sci-fi/anime communities–consider crafting any NDA with exceptions for screenshots, chat logs, etc., while protecting mechanics and game- or genre-unique features. Encourage the opinion makers to produce good press for the game (which means that as soon as this is made available to them, that is be solid while not feature complete). Publicize moving into different stages of alpha and beta leading to a public release. See also Massively Multiplayer Game Development 2, edited by Thor Alexander, “Small Fish In a Big Pond…” by Mike Wallis, pp. 409-424.

7 Responses to “GITS:SAC MMO design (revised)”

  1. Technocrat Says:

    I’m lovin’ the heavy emphasis on missions! I’ve always felt that a really good mission building AI would be well worth the effort; no matter what the cost, no matter how many man hours, no matter what it took it’d be worth it. A really good mission building AI would be able to build thousands of unique multiple-objective, PC specific missions. Each mission would be like playing through a level or two of a really good SP game. You know what I’m talkin’ about.

    concerning penalties; I’ve always been a big fan of fairly harsh punishments–nothing massive, just enough to give ya that “white knuckle ride” experience! 😉

    What kinda skill system would you have in this game? How many skills? Personally, I’ve always wanted to see a scheme whereby the PC would advance in several different ways simultaneously, such as: using a particular skill and gradually improving it with practice (a la Project Entropia). This would give the player a steady stream of those yummy “ding” moments; skills gained from books or holodiscs (a la EVE Online);
    learning from expert training (a la SWG); etc…

    I gotta get going…

  2. Chris Says:

    Many revisions were based on her principles (and drilling them into my skull):

    • Gamers are looking for something fun
    • Gamers are looking for something fair

    Post-WoW, fun and fair have come to take different meanings than they with UO or EQ. Some of the changes recognize that (such as making all organizations equal in rewards, rank, and abilities, although those all take different forms, or how missions are something the player selects and can abandon). In general, she made the game design less hard core but more accessible. By the end of the second draft, she said, “Well, I played City of Heroes. I guess I could play this.” High praise!

    I’m still considering skills. They’re tied to rank, but I haven’t decided if skills act more like abilities or talents (total count tied to rank, or total level tied to rank, or skill level capped by rank), or if they’re used based, and if use based, if they’re capped or uncapped based on rank, time played, etc. I also haven’t formalized a mechanism for gaining rank. As I have time, I’ll think about those and post my conclusions here. I only have a starting point, the raw recruit who barely shoots a pistol straight and the Major, who can shoot a machete blade in half in mid stroke, or fire her gun so accurately and quickly that her bullet can intercept another in mid-flight.

  3. Mischiefblog » Blog Archive » A little horn-tootin’ about PvP Says:

    […] WoW, EQ, SB, and AC depend entirely upon PvE for advancement. EQ2 and DAOC allow experience and advancement through PvP, but that is a pathological play style for levelling. Arguably, removing the PvE content and replacing it with purchaseable items, ala Guild Wars, is an alternative that could work for EQ2 and DAOC. Guild Wars is a special case, with both PvE and PvP playstyles, where an exclusively PvP playstyle lives in arenas, while a PvE character can participate in both games. I would hope that my partial design for a cyber-punk GITS:SAC-influenced MMO (GITS:SAC MMO design (revised), Cyberpunk MMO rules (first draft)) would provide a game that lets both PvP and PvE players participate and advance on an equal footing. In particular: […]

  4. Psychochild Says:

    I love GitS. I think this is an interesting idea.

    One major problem with your design: there’s a negative feedback loop with failure. If failure lowers a player’s available options and resources, this encourages future failure from the player. If he or she couldn’t accomplish a mission with top-of-the-line items, what chance does the person have with lesser items? At the very least, it will encourage a bit of bottom feeding for people to work off the “debt”.

    Some thoughts,

  5. Chris Says:

    You’re right. One thing to note is that I viewed missions as expendable, so if one was failed, you could try for an easier one or a group mission the next time (although that introduces the risk of belonging to a really bad group).

    I’ve been drifting toward player characters purchasing everything that isn’t in-mission loot or a mission reward. Rank restrictions may be tied to the purchase and/or use of some items, but otherwise, the players could buy the best gear that they could afford. This allows some twinking.

    The intent of the limited, prestigious items was to be the equivalent of the Really Big Gun or the Laser of Molten Faces. Rather than being restricted to the player’s mission success (or mission debt), they may instead be limited to faction–that encourages particular play styles (straight cop, dirty cop, viscious thug, kidnapper–whatever the sort of mission may grant for faction) to gain access to the item.

    What kind of items do characters get for free? If a character is broke and failed missions (so no rewards were given), how can the character break the cycle? Do surveillance missions? Do missions that don’t involve pricey gear? Or always make a certain level of gear available to the character for free?

  6. Bastian Says:

    Try not to concentrate too hard on char classes, ranks and missions. Players want fun and fair, alright. IMO that means as much player freedom as is possible. I know it’s a lot of work already to think up stuff, but instead of a certain class let the players choose an education, which enables them to take on jobs/ missions that suit their field, f.e. IT, mechanics, business, military background etc.
    Forget levels as in “LVL 20 Warrior-Monk”. It’s no good. Qualification might be another way. Players could have a few slots to train additional skills and have to exercise them, in order not to lose or “forget” them.
    Missions are important but giving players professions might be good, too. You can’t plausibly send everyone off to extremely risky and important missions that help saving the city. Some players are happy running a back alley shop where they can cook and sell drugs (strongly against them, but there’s no serious cyberpunk universe without them) or knit carpets, build vehicles etc.
    So far your ideas sound pretty impressive, actually that would be the kind of MMO I’m looking for but doesn’t exist yet.

  7. Mischiefblog » Mission or quest goals as graphs Says:

    […] Many missions or tasks should be able to be completed by a single player or (faster) with a team. Several roles or skills are used (such as unlocking doors, defeating security systems, watching for guards, cracking safes, etc.) that fulfill tasks in the mission (in order and in arbitrary order). With a team, these may be completed faster. […]