Three experts, the assassin, spy, and sage, require special treatment. Each of these, unlike other hirelings, can affect the direction and content of an on-going adventure. Used carefully and sparingly, these three are valuable DM tools to create and shape stories in a role-playing campaign.
Assassins[]
Assassination is not a discreet occupation per se, but a reprehensible mind-set. The assassin requires no special skills, though fighting, stealth, and even magic are useful. All that is really needed to be an assassin is the desire and the opportunity.
Hiring an Assassin: When a player character hires an assassin (which is not a good or lawful act), he is taking a chance. There is virtually no way to assure oneself of the reliability and dependability of such a person. Anyone willing to make a business out of murder is not likely to have a high degree of morals of any type. Clearly, this is a case of "let the buyer beware!"
Once a character has hired an assassin, it is up to the DM to determine the success of the deed. There are no simple tables or formulae to be followed.
Consider the intended victim: Assassination attempts by one player character against another should not be allowed. This type of behavior only leads to bitterness, bickering, and anger among the players. NPC-sponsored assassination attempts against player characters should be used sparingly, and then only as plot motivators, not as punishment or player controls. Any time a player character is targeted, role-play the encounter fairlyógive the PC a chance.
If the intended victim is an NPC, the DM should decide the effect of the assassination on his game. Sometimes, player characters do these things out of spite. At other times the deed may be motivated by simple greed. Neither of these is a particularly good motive to encourage in a campaign.
If the death of the NPC would result in a major reworking of the campaign for no good reason, consider seriously the idea of making the attempt fail. If the death of the NPC would allow the player characters to by-pass or breeze through an adventure you have planned, then it's not a good idea. Don't just tell the players, "Oh, that'd be bad for the game so you can't even try to knock that guy off." Work the attemptóand its failureóinto the storyline.
Precautions: If you decide the attempt is legitimate, consider the precautions the intended NPC victim normally takes. These may make the job particularly difficult or easy. Kings, emperors, high priests, and other important officials tend to be very cautious and well-protected. Wizards, with wise magical precautions, can be virtually impossible to assassinate! Devise specific NPC precautions before you know the assassin's plans.
Wizards make use of magic mouth, alarm, explosive runes, and other trap spells. Priests often rely on divination-oriented items to foresee the intentions of others. Both could have extra-dimensional or other-planar servants and guards. They may also have precautions to foil common spells such as ESP, clairvoyance, and detect magic. Kings, princes, and other nobles have the benefit of both magical and clerical protection in addition to a host of possibly fanatically loyal bodyguards. If the victim has advance warning or suspects an attempt, further precautions may be taken, and the job can become even more difficult.
The Plan: After you have decided (secretly) what precautions are reasonable, have the player describe the plan he thinks would work best. This can be simple or involved, depending on the cunning of the player. This is the plan the assassin, not the player character, will use, therefore the player can presume some resources not available to the player character. However, you must decide if these resources are reasonable and truly exist.
For example, if the player says the assassin has a map of the castle, you must tell him if this is reasonable (and, unless the victim is extremely secretive and paranoid, it is). A plan involving a thousand men or an 18th-level thief is not reasonable. The player character hasn't hired an entire arsenal!
Finally, compare what you know of the precautions to the plan and the success or failure will usually become clear. Ultimately, the DM should not allow assassinations to succeed if he doesn't want them to succeed!
In general, allowing player characters to hire assassins should not be encouraged. Hiring an NPC to kill even a horrible villain defeats the purpose of heroic role-playing. If the player characters can't accomplish the deed, why should they be allowed to hire NPCs to do the same thing?
Overuse of assassins can often result in bitter feelings and outright feudingóplayer vs. player or player vs DM. Neither of these is fun or healthy for a game. Finally, it is a very risky business. Assassins do get caught and generally have no compunctions about confessing who their employer is. Once the target learns this, the player character will have a very dangerous life. Then the player character can discover the joy and excitement of having assassins looking for him!
Spies[]
While less reprehensible (perhaps) than assassins, spies involve many of the same risks and problems. First and foremost, a spy, even more than an assassin, is inherently untrustworthy. Spying involves breaking a trust.
A spy, unlike a scout, actively joins a group in order to betray it. A person who can so glibly betray one group could quite easily betray another, his employer perhaps. While some spies may be nobly motivated, these fellows are few and far between. Furthermore, there is no way to be sure of the trustworthiness of the spy. It is a paradox that the better the spy is, the less he can be trusted. Good spies are master liars and deceivers even less trustworthy than bad spies (who tend to get caught any way).
In role-playing, spies create many of the same problems as assassins. First, in allowing player characters to hire spies, the DM is throwing away a perfectly good role-playing adventure! Having the characters do their own spying can lead to all manner of interesting possibilities.
Even if NPC spies are allowed, there is still the problem of success. Many variables should be considered: What precautions against spies have been taken? How rare or secret is the information the character is trying to learn? How talented is the NPC spy? How formidable is the NPC being spied upon?
In the end, the rule to use when judging a spy's success is that of dramatic effect. If the spy's information will create an exciting adventure for the player characters without destroying the work the DM has put into the campaign world, it is best for the spy to succeed.
If the spy's information will short-circuit a well-prepared adventure or force the DM to rework vast sections of the campaign world, the spy should not succeed. Finally, the spy can appear to succeed while, actually, failing—even if he does return with information, it may not be wholly accurate. It may be slightly off or wildly inaccurate. The final decision about the accuracy of a spy's information should be based on what will make for the best adventure for the player characters.
Sages[]
Unlike other expert hirelings, sages are experts in a single field of academic study. They are most useful to player characters in answering specific questions, solving riddles, or deciphering ancient lore. They are normally hired on a one-shot basis, to answer a single question or provide guidance for a specific problem. A sage's knowledge can be in any area that fits within the limits of the campaign. Typical sage areas are listed on Table 61.
Frequency is the chance of finding a sage with that particular skill in a large city—a university town of provincial capital, at least, Normally, sages do not reside in small villages or well away from population centers. They require contact with travelers and access to libraries in order to gain their information. Roll for frequency only when you can't decide if such a sage is present. As always, consider the dramatic effect. Will the services of a sage further the story in some exciting way?
Abilities and limitations define specific limitations or rules effects. If this column is blank, the sage's knowledge is generally thorough on all aspects of the topic. One race only means the sage can answer questions that deal with a particular race. One region only limits his knowledge to a specific area—a kingdom or province. The size of the area depends on the campaign. One plane limits the sage to the study of creatures, conditions, and workings of a single extra-dimensional plane. Where no limitations are given, the sage is only limited by the current state of that science or art in your campaign.
What does a sage know? A sage's ability can be handled in one of two ways. First, since the DM must answer the question any way, he can simply decide if the sage knows the answer. As usual, the consideration of what is best for the story must be borne in mind.
If the player characters simply can't proceed with the adventure without this answer, then the sage knows the answer. If the answer will reward clever players (for thinking to hire a sage, for example) and will not destroy the adventure, then the sage may know all or part of the answer. If answering the question will completely unbalance the adventure, the sage doesn't know the answer.
Of course, there are times it is impossible to tell the effect of knowing or not knowing something. In this case, the sage's answer can be determined by a proficiency check, modified by the nature of the question. The DM can decide the sage's ability or use the following standard: Sage ability is equal to 14 plus 1d6 (this factors in his proficiency and normal ability scores).
If the proficiency check is passed (the number required, or less, on 1d20), the sage provides an answer. If a die roll of 20 is made, the sage comes up with an incorrect answer. The DM should create an incorrect answer that will be believable and consistent with what the players already know about the adventure.
Questions should be categorized as general ("What types of beasts live in the Valley of Terror?"), specific ("Do medusae live in the Valley of Terror?"), or exacting ("Does the medusa Erinxyes live in the Valley of Terror?"). The precision of the question modifies the chance of receiving an accurate answer. Precision modifiers are listed on Table 62.
If a question is particularly complex, the DM can divide it into several parts, each requiring a separate roll. Thus, a sage may only know part of the information needed. This can be very good for the story, especially if some key piece of information is left out.
The resources required by a sage can be formidable. At the very least, a sage must have access to a library of considerable size to complete his work. He is not a walking encyclopedia, able to spout facts on command. A sage answers questions by having the right resources at hand and knowing how to use them. The size and quality of the sage's library affect his chance of giving a correct answer.
This library can belong to the sage or can be part of an institution. Monasteries and universities typically maintained libraries in medieval times. If a personal library, it must be at least 200 square feet of rare and exotic manuscripts, generally no less than 1,000 gp per book. If the library is connected with an institution, the sage (or his employer) will be expected to make appropriate payments or tithes for its use. Expenses in the range of 1,000 gp a day could be levied against the character. Of course, a sage can attempt to answer a question with little or no library, but his chances of getting the right answer will be reduced as given on Table 62.
Sages need time to find answers, sometimes more time than a player character can afford. Player characters can attempt to rush a sage in his work, but only at the risk of a wrong answer. The normal length of time depends on the nature of the question and is listed on Table 63. Player characters can reduce the sage's time by one category on this table, but the chance that the sage's answer will be incorrect or not available grows. These modifiers are also listed on Table 62.