Back to index

The copyright situation for this article is unclear. It does not belong to the author of this site. Please see the copyright notice. If you have information about the copyright contact me!

Beyond Player Killing

by Sayeed

This is a continuation of an article posted last month.

Advantages and disadvantage of focused automatic solutions

Automatic Location Focused PK Measures and the PK-Switch

Location Focused Measures are a current trend of various online muds. Focused Location solutions to Player Killing limit the area(s) where Player Killing is possible by coding in the game engine to either neutralize damage done, limit who can attack or be attacked, or simply disallow PvP combat in Player vs. Player Combat ("-PvP") forbidden Locations. This allows "safe-havens" to be created, where players can adventure secure in the knowledge that their surroundings will remain only as dangerous as they have mentally marked it, safe from PK's, and always suitable for particular levels of characters to fight against Monsters.

Head in some garbage, Delicatessen is a fantastic movie
Escaping! From Delicatessen the movie
The "PK-Switch", a common Player Killing measure, is essentially a Location based solution, making the entire world (the Location) +PvP if it is enabled, and -PvP if it is disabled. Currently the PK Switch is seen by many as a flawless method for limiting PvP combat, and certainly it is very effective. Unfortunately for the gaming community, however, it exhibits the problems of most Location based solutions, some of the faults of which can be understood by examining the PK Switch.

The hype for the "Perfect PK-Switch" is similar to this:

Those who do not want to be killed by players won't be, thus allowing those who want to play in a certain style of play do so without disruptions.

This argument sounds very convincing until we ask ourselves a question. Do players really want to be killed by Monsters? No? So why not a Monster Switch? This solution, however, would prevent players from having bad experiences with Player Killers AND Monsters, yet it would not set realistic limits within a fantasy world, or provide the sense of challenge which makes accomplishment exciting. It is the Possibility of death which produces this sense of challenge. So ultimately the hype for the PK-Switch is altered, now reading as:

Those who don't want the POSSIBILTY of being killed by players won't have it, and those that do will have it, therefore allowing those who want to play in a certain style do so without disruptions.

Does the PK-Switch do this? Yes. Flawless now? No. One must consider that not all Player Killers are created equal. True some Player Killers are immature and heartless irritants, yet some are also Interactively Intelligent on a Role-Playing or at least an In-context level, and contribute greatly to the excitement and immersive quality of a game. Would any classical fantasy novel be complete without the corrupt and maniacal Super-Villain? Role Played villains have the potential to be the most interesting "opponents" in Role-Playing games, and few Role-Players would wish to completely disallow ALL combat with Real Opponents, only with the few who act out-of-context and immaturely. Therefore, the final, rendition of a Perfect PK-Switch would be the following:

Those who don't want the POSSIBILTY of being killed by some players won't have it, and those that do will have it, therefore allowing those who want to play in a certain style do so without disruptions.

If the PK-Switch fulfilled this requirement, then it would be perfect. Unfortunately, with the insertion of the "some" clause, the needs of a "perfect PK-switch" go beyond what the PK-Switch actually delivers, rendering it, although an adequate alternative, imperfect for the most enjoyable Role-Playing experience. In a Role-Playing Game, this is indeed a fault.

There also exist a few other disadvantages to Location Focused PK Measures. In -PvP zones, for instance, where characters cannot attack each other, the most vile villain may taunt and verbally abuse other players, who are powerless to intervene or Role Play the "good citizen." In a PK-Switch circumstance, this becomes -PvPs taunting +PvPs, or perhaps -PvPs who are powerless to help their +PvP friends defend against other +PvPs. The existence of these zones abutting battlefields also presents the potential for player abuse, not something to be taken lightly in muds. The final issue of balance, where and how much of the world should be +PvP and how much should be -PvP, is a difficult issue which may take much balancing to decide, with proponents of both camps constantly asking for more area.

Some Examples of Location Focused PK Measures

PK Switch- +PvP may attack and aid other +PvPs yet no combat is allowed between -PvP and other players, except in a mutually agreed upon dual. Characters are -PvP by default and must perform a quest to become +PvP.

Pure Location Solution- PvP combat may only take place within defined regions or in mutually agreed upon duels.

Limited Monster Solution- Characters may play various humanoid monsters yet may only ATTACK other players within a bounded region, though they may be attacked and defend in others. Monsters must act suitably like Monsters or risk losing monster status when snooping GMs see they are not fulfilling their look.

Automatic Loot Focused PK Measures

If one of the greatest problems of Player Killing is the material wealth lost by victims, then the answer would seem simple; limit looting. This requires that either the contents of a victim's body at death contain only a few, valuable items, with the rest remaining with the "spirit," or that a Player Killer is allowed to loot only a few items from a corpse (though the latter implementation produces recovery problems). The Loot Focused solution seems straightforward and simple, with minimum interference, yet still has many problems. Some would claim that Loot Focused measures alone are not strong enough, for unless other controls are put in place, Player Killers who are discouraged by the small loot of PKing may be spurred by this to Kill MORE instead of less, to make up lost wealth. This also does not affect the Interactive Intelligence of PKs, only their incentive to PK, so while it MAY reduce the number of Player Killers, it will not affect how those that do Player Kill act. Also, much of a Player Killer's impetus is the pure joy of Player Killing rather than the wealth gained from it. Therefore, simple loot focused Player Killer solutions are extremely lacking.

Some Loot Focused PK Measures

Monster Loot Limitations- Add the same limitations which Monsters loot under to Player Killers, so they can only deprive their victims of a FEW treasures. (As most solutions, this can be implemented in several ways) Possibly putting no limitation on the looting of Evil characters. Another implementation of this would be to provide incentives for characters to stay in Groups (or code the engine to differentiate between groups) and allow the attacking group/individual to only loot a number of items while the defender has the option to loot without limitation.

Level Loot Limitations- Prevent Player Killers from looting characters of lower levels than themselves.

Loot 'Fence' Solution- Mark PvP looted items as "stolen" unless taken in an engine supported war/duel, and have only a few NPC buyers in the mud world who will buy Stolen Items, and then only in specific amounts.

Automatic Strategic Intelligence Focused PK Measures

Similar to Loot Focused solutions, Strategic Intelligence Focused solutions may be isolated in that they have little effect on the other three problem areas except by lowering the overall Player Killer population. By limiting the Strategic Strength of Player Killers, other players can be given a fighting chance, and if it is much harder to successfully Player Kill, then perhaps less Players will take that route, and the problem of "PK Mobs" will be solved (as there will be less PKs to form mobs). An example of a Strategic Intelligence Focused solution might be limiting the number of times a character can die in PvP combat per day. While this would not erase the Player Killer problem, it may be more realistic or easily Role Played than location based solutions, where Players cannot die at ALL in some areas. It would also prevent players from being continually killed by other players, somewhat reducing the Mass Murder problem.

Strategic Intelligence Focused PK Measures

Death Limitation- Limit the amount of times a character can be killed in PvP combat, unless in mutually agreed upon duels.

Player Death and Unconsciousness- Institute Player Death and Unconsciousness. Player death being meted out only by very high level monsters. Give Player Killers the option of causing death or unconsciousness, though provide severe penalty if she/he chooses death. (ie- Evil for life, no shop will deal with him, gets no experience and not allowed to loot, an in-context-chance to die himself, guards will search for him and KILL him on sight, his location is tracked and published frequently on boards to aspiring heros, etc.)

Surprise Prevention- If a character with a "Bad Reputation" (in whichever method of reputation tracking the mud uses) is near to a player character, provide an in-context alert which prevents ALL encounters with PKs from being "over in a second" ambushes. IE- Your spine tingles and you know great evil is nearby

Automatic Interactive Intelligence Focused PK Measures

Inherent in Role-Playing an evil character are realistic limits on how much murder takes place. There are only so many Serial Killers in any world, and so if Players decide to Role Play evil characters then in addition to making encounters more enjoyable for the player, this type of play will also curb all the other 3 Player Killing problems. In other words, Role-Playing Villains might choose challenging fights and Role Play before they begin rather than going on Mass Ambush Serial Killing Sprees. Villains might also might be picky about what loot they take, and they will likely stick to their own "home" region. Thus, a solution which affects the Interactive Intelligence problem area has the potential to be a highly effective general solution. Unfortunately, though this type of solution is seemingly the most efficient, the problem here lies with Implementation.

Interactive Intelligence includes a wide range of Player Behavior, and not only language (which can be censored through filters), and therefore a pure and comprehensive Automatic Interactive Intelligence Focused PK Measure is very difficult to implement except through human intervention. As the issue of bias arises if this solution is Player-Effected, the only recourse is generally Staff-Effected Interactive Intelligence Solutions. (though Player-Effected is possible) Here is where the problems of Interactive Intelligence solutions rise to the surface.

The most publicly disputed problem with this method is the problem of Manpower. With thousands of players online, the idea of differentiating between In-context and Out-of-context players to enact a Player Killing solution may be time consuming and expensive. Even if such a solution is limited (for example staff members only make judgements on whether players can attack other players once the players reach higher levels) then with multitude of players the Manpower requirements are decreased but still significant. Combine this with the regular expense of hiring regular "GM's" to maintain, upgrade, and oversee the game, and we discover that only an extremely "tight" and considered solution will be economical for mud companies.

Another problem with Interactive Intelligence solutions has to do with human judgment management and perception of it. In Online Role-Playing Games, response from consumers is fast and sometimes violent. Judgement Calls by mud Staff, therefore, will quickly produce an intense response from those of the Gaming Community who feel slighted by GM decisions in a Staff-Effected Interactive Intelligence solution. For example, the player who is not allowed to attack other players may go to a message board and accuse staff members of favoritism or misjudgment. This effect multiplied a number of times may snowball and create frustration and disappointment in the gaming community. This resentment has the possibility of driving players to the mud's competition, something which no developer would appreciate. Finally, the management of flexible human judgment is a difficult concept. Who is to say that one GM's decision is not completely arbitrary and actually IS based on favoritism? Is it possible to create general standards to judge aberrant behavior by? While this can be done, it is again something which developers may have to take great care in doing properly.

Player-Effected solutions, such as allowing characters to flag themselves Role Player, non-Role Player, or both, and only showing them the text of those who are similarly flagged, may be another Interactive Intelligence Measure. This measure, however, deals with the Perception of the language of Player Killers (by possibly filtering out out-of-context speech) rather than any other aspect. Still, though a crude solution, there may be benefits in something like this.

Some Examples of Interactive Intelligence Solutions

Monster Filtering- Allowing high/mid level players the privilege of eventually playing monsters or Player Killers, limited to a certain location, but having Staff-GMs check on those players now and then and turn them -PvP if they do not properly play Monsters. (In this solution playing Monsters or +PvP characters is a privilege for those who can enhance the game experience, not a right, and carries with it certain responsibilities)

Name Filtering- Have GMs delete any names of characters they see which are blatantly out-of-context.

Text Filtering- Allow characters to tag themselves Role Player, Non-Role Player, or Both and only allow their character to see text from other players similarly flagged.

Discretionary solutions

A less "brute force" method of limiting Player Killing in muds is by Discretionary Solutions. Such solutions are not necessarily large additions to the game code which forbid or limit Player Killing, but are instead methods of working in-context with a game engine to encourage would-be player killers not to take that road. This is done by providing clear incentives to stay non-PK and grave consequences if they choose to Player Kill.

A common example of such a method is town guards. Town guards may be programmed to swarm any evil ranked (by the mud reputation system) player who comes into or close to the towns. Thus Player Killers must learn to live without the amenities of a good-aligned town, and such players may be encouraged not to Player Kill because of the difficulties inherent in doing so.

Discretionary solutions are effective in the way that they are generally clearly in-context, without using any wildly unbelievable magic or method, but also may be quite weak. Most player killers, using mules, "good-aligned" friends, or a variety of other methods, are enabled to avoid the penalties of town guards, and using various other techniques enable Player Killers to avoid many different discretionary methods or even abuse them to provide themselves with advantages. Realism vs. Weakness is the issue. Discretionary solutions alone may not produce the desired situation, yet game developers often use them in combination with Automatic Solutions to solve the problem of Player Killing.

Examples of Discretionary Solutions

Player Policing- Make it easy for Heroes to find players who go to excess in Player Killing, (ie- a reliable reports put Sir Jurnat the Evil, charged with 50 murders, in the Red Wood forest) thus having the Players participating in solving the Player Killing problem.

Limiting Mass Transit- Simply hindering the escape means of Player Killers, by limiting Mass Transit methods by which they may easily escape justice may also help solve Player Killer problems.

Guard Forays- Have powerful guard patrols, along with NPC mage guards, travel the routes between towns, attacking evil players they find along the way.


To conclude, from the four major problems caused by Player Killing, to the myriad of possible Automatic and Discretionary solutions to these problems, Player Killing and the solution to its disadvantages is an extremely complex issue. It involves looking at how it will affect not only the actions of Player Killers, but also how it will affect the environment, thoughts, and actions of other players, and what other problems may be encountered when implementing it. The assertions that the PK-Switch is a "perfect" solution and that no solution is needed are false. The perfect solution, I am sure, is still possible. This solution will give muds realism and balance, and will make the mud experience extremely enjoyable for Players and Player Killers alike. It was the intent of this article to instigate deep thought on this issue, so that, at length, some aspiring Player or Developer will find the Ultimate Solution. When this happens, it is my sincere dream that Role-Players will appreciate Player Killers for the benefits they have the potential to give, and enjoy our experiences with them as much as we enjoy facing the evil NPC denizens of Online Worlds.


I hope you enjoyed or learned something from this article, now let me tell you more of my opinions on Player Killing and the PK-Switch, which you have probably already guessed. Keep in mind that this part of the editorial may go off on a tangent, and may not be quite as relevant to the topic as the rest of the article.

The intention of this article was to provoke thought. It has been rather long and relatively academic, and for any who have struggled through to the end, I admire your tenacity. As I stated before, I didn't set out to give a definite solution, just to show you the many problems and possible solutions, and hope that someone starts picking from all of them, inventing his own, and finally synthesizes the Perfect Solution. Maybe it will be you, maybe it will be me, but only with hard work and conscientious testing will it be found. For us, the RPG Community, I hope it is found soon.

You may find it surprising, but personally I dislike intrusive solutions to Player Killing. True, I dislike even more a world where you cannot adventure without being killed, but I am fairly sure that well thought-out and powerful Discretionary measures, and perhaps a simple Automatic measure or two will get the job done. Now all I have to do is invent it, and prove it, and I pray that you beat me to it! For any measure that is implemented, whether Automatic or Discretionary, I feel it is extremely important to involve the mud storyline in justifying the measure. I think the PK-Switch, if justified like this, should be farther along its way towards being the Perfect Solution. I feel the PK-switch is a commendable solution, yet not perfect. Whether modifying or replacing the PK-Switch, or replacing whichever solution precedes it, the Perfect Solution will come, and we will find it.

As for In-context GM intrusion, I am ALL FOR IT. In muds, in pen'n'paper RPGs (what else?!), in focused Chat Rooms and Message Boards, in PBEM games, in muds, in everything RPG related. However, this is contingent on the fact that the GMs can be trusted. And, no matter how cynical I am and have been convinced to be (yes, I know of foul mud GMs), I am overall a trusting person, are you? I trust game companies, because if they intrude in their game, then they know that it is their financial future that may be affected by it. I think that GM Interactive Intelligence solutions in addition to Discretionary solutions, if implemented very "tightly" and efficiently, have the potential to make our mud experience a great one. Come to think of it, if the obscure deadfalls associated with most Focused PK Measures are avoided, they all have potential. Personally though, I'm going to look for the Perfect Solution in the area of Discretionary Measures. (and perhaps a little Staff- Effected Automatic I.I.)

What is the most powerful Discretionary measure? There are a countless number of them, but the most "important" is one which deals with a variety of issues and a very important equation.

More Unique Challenge = Less Player Killers

Head on a plate, Delicatessen is a fantastic movie
Clowning around, from Delicessan the movie.
This is practically self-explanatory. "Unique" challenges refers to a variety of quests/monsters in a variety of different environments, with at the monsters having at least acceptable AIs which cannot be exploited. Unique implies that a later upgrade to an mud intended to provide more challenges, yet which simply raises the Health of various monsters, or creates similar one with few alterations, will not be sufficient and will not provide a good Discretionary solution for Player Killing. Unique implies that if there is a method of game engine manipulation which allows players to deal with all challenges similarly, thus negating the "Unique" aspect of challenges, then players may become frustrated with a lack of challenge and begin to Player Kill. This of course only deals with the group of players which turn to Player Killing in response to a lack of unique challenge in the mud world, yet that is an important group. That group may or may not choose to Role-Play if they turn to PvP to provide themselves with challenge, and if not, the effect will be serious.

I like Role-Playing interactions with mature Evil Characters. I like the thrill of vying with them for victory. I like the excitement of dueling to the death with them. But I don't like being ambushed 10 times in a row and killed by 5 simultaneous spells every time I enter a dungeon. If that happens I like there to be in-context methods of retaliation, with the chance to succeed. I think that limiting Mass Transit Systems in muds will go far towards this goal. In-context solutions to the problems of Player Killers must be developed. Solutions which have realistic limits, and allow those encounters and ambushes to not be eliminated, but to be limited, not be punished, but to be put in-context and to good use, not be flamed by Victims, but to be Role Played by Villains.

In the end, the enjoyment we get out of a Role-Playing Game will be decided by how good are its Role-Playing Gamers, so let's ROLE PLAY! (I mean PKs too.)

PS- Especially in this topic, comments, ideas for the Perfect Solution, experiences with the PK-Switch, and other constructive criticisms and observations are much appreciated.


Special Thanks to Dirge, for his support and insightful comments which guided me away from the major pitfalls with associated with this article.

I'd like to thank that guy who posted that Quake II should be bundled with muds for people to relieve certain suppressed emotions, this all started as a response to that post.

I'd like to thank the mud companies who are going through the arduous task of testing Pkilling Solutions.

'Sayeed from somewhere I am not certain of.