The following article originally appeared on an earlier iteration of See Page XX in April 2008. 

A column on roleplaying by Robin D. Laws

Give a Clue

The heart of the GUMSHOE system is its method of ensuring that players always gather the clues essential to the solution of a mystery and lots of other supplementary information as well provided that they have the right investigative abilities, and describe their characters look in the right place and/or perform the right actions. Some potential players have of the game have concluded is that the removal of random determination from the clue dispensation process must render it dry and mechanical. The reality of the play experience is that it is just as fluid as in any other mystery game.

The reason for this fluidity lies in the freedom granted the GM to dispense clues in various ways. These keep the investigative scenes spontaneous and interactive.

Until now, these methods have been implicit in the scenario text. I’m confident that GMs instinctively get them in play and in a way am reluctant to pin them down too much, for fear of overriding good on-the-spot judgment for what appears to be a heavily prescribed set of techniques.

With that caveat, here are some terms to codify the methods GMs use to provide clues in GUMSHOE:

Immediately Apparent

An immediately apparent clue is supplied to the player without action on the character’s part. All an investigator with the governing ability has to do to spot the clue is to enter the scene. Ideally, the GM scans his master list of investigative abilities, on which the ratings of the various PCs are marked, picks the most likely investigator with the ability, and announces the clue:

[indicating a particular player]: “You can tell right away that the hieroglyphics on the statue are phony modern gibberish.”

Here the GM is responding to a passage in the scenario that says:

Archaeology shows that the hieroglyphics on the statue are phony modern gibberish.

There are two reasons to treat a clue as immediately apparent: believability and playability. Believability holds that clues where anyone acquainted with the ability in question would logically spot something on a cursory inspection should be provided without prompting. On the other hand, playability dictates that essential clues which even good players are unlikely to look for should also be made immediately apparent.

Certain clues are immediately apparent without abilities. If there’s a gun hidden under a bed, and a player asks, “What’s under the bed?” they don’t need Evidence Collection to find it. All they need is a pair of functioning eyes.

Action-Dependent

Most clues are action-dependent, meaning that the players must specify that they’re doing something before the GM provides the clue. The action taken can be very basic: searching the room, looking for fingerprints, taking a closer look at that painting in the corner. Or it can be quite specific: gathering fibers, performing a centrifuge test, smelling the air for the distinctive tang of werewolf.

GM: Jenkins hands you a photograph, of what appears to be a sasquatch standing in a stand of bullrushes.

Player: As an experienced photographer, I want to know if the image of the monster has been faked.

[The GM refers to the scenario notes, which read:

A check for fakery with Photography shows that it is a composite image.]

GM: It’s a composite; the shadows in the figure don’t match the direction of light in the background.

Shifting Clue Types

The wording of GUMSHOE scenarios suggests which of these two categories the clues fall into, without being absolutely explicit about it. I toyed with the idea of making these more definite, by marking them with icons. Ultimately I decided against this, because the most important thing about clue dispensation is to pay attention to the progress players are making and adjust on the fly. Most immediate clues can be turned into action-dependent clues as needed, and vice versa.

If your players are slogging their way through a mystery whose basic backstory just isn’t registering, you may want to supply suggested actions, effectively turning an action-dependent clue into an immediate clue: “Your Forensics experience leads you to check inside her mouth, where you find a strange parasitic infestation.”

On other occasions it is more satisfying for the players if you strongly hint at a suitable action, rather than providing the clue outright:

GM: Jenkins hands you a photograph, of what appears to be a sasquatch standing in a stand of bullrushes.

It strikes you as off, somehow.

Player: I check it for signs of fakery!

Although you might expect the players to regard this as an unsubtle shove in the right direction, many players are not only content to receive hints like this, but still feel a sense of accomplishment simply for going on to fill in the obvious next action. The more frustrated a group becomes, the greater the emotional reward for pouncing on a hint.

Always allow the players plenty of time to take actual active measures before you start hinting them in a fruitful direction.

This idea can be spun in the opposite direction. If your players are especially proactive, you can reward their initiative by converting immediately apparent clues into action-dependent ones.

GM: The wall inside the burial chamber is covered in old hieroglyphics.

Player: Aha! Are they phonetic or logographic?

GM: Neither. They’re gibberish — modern forgeries.

Players are more able to show off their characters’ brilliance in areas they are themselves acquainted with.

All in all, the degree of effort players must go through to accumulate clues is a matter for constant and sensitive adjustment, based on factors including session pacing, the group’s concentration level, and players’ personal knowledge of character abilities. The defaults suggested by the scenario wording are no substitute for a GM’s judgment and attention. Knowing when to push and when to let the players push you is an essential component of the GM’s craft. You are probably already doing it, unconsciously, but by paying more active attention to it, you can further sharpen your presentation.


GUMSHOE is the groundbreaking investigative roleplaying system by Robin D. Laws that shifts the focus of play away from finding clues (or worse, not finding them), and toward interpreting clues, solving mysteries and moving the action forward. GUMSHOE powers many Pelgrane Press games, including The Yellow King Roleplaying Game, Trail of Cthulhu, Night’s Black Agents, Esoterrorists, Ashen Stars, and Mutant City Blues. Learn more about how to run GUMSHOE games, and download the GUMSHOE System Reference Document to make your own GUMSHOE products under the Creative Commons 3.0 Attribution Unported License.

This post originally appeared on DyingEarth.com between 2004 and 2007, but could prove useful for the many subsequent GUMSHOE systems.

A column about roleplaying

by Robin D. Laws

On a fundamental structural level, RPG sessions are their own beast, and are unlike movies, TV, and books. However, these related storytelling forms are always worth looking at for inspiration. Many of their surface techniques remain unplundered by GMs. Most notably, the tricks they use to compress time and make proceedings less boring demand further study, if not slavish emulation.

For example, let’s look at the differences between a story of investigation as it plays out in a TV cop show as opposed to the way they usually unfold in an RPG.

In a cop show, each encounter or interrogation generally a few important points of information. Then the script quickly moves onto a new scene in which another character provides more information.

Often, though not always, the investigators must score a win by overcoming the informant’s reluctance to spill the crucial beans. The informants’ reasons for reluctance, and the means necessary to overcome them, will vary enough to disguise the formula and keep the proceedings entertaining.

You can’t break it down to a formula, but often the informant:

A) provides one clue

B) rules out one possibility

and concludes by

C) supplying a third nugget of information pointing the investigators to the next encounter.

RPG interrogations tend to unfold in actual time. In that, they’re like real police interviews: given the chance, the PCs will ask every question under the sun, looping around, repeating themselves, and amassing great reams of information from each informant, which they’ll then try to sift for the crucial point.

This poses a challenge to you as GM, because you want a sense of forward movement, to build excitement and stave off boredom and paralysis. Players become easily confused in investigative scenarios. Unlike real cops, they’re picturing their nonexistent people talking to your nonexistent people. As they go, they’re filling in the imaginative blanks, often mistakenly. The more editing and pre-sifting of information you can do for them, the happier they’ll be, and the more satisfying the episode’s pacing will seem.

By imitating a cop show trick, you can keep each interview quick and to the point. No one in a cop show has time to talk to the cops. (Maybe this is why most of the best cop shows are set in New York City, where no one has time for anyone.) The random group of eccentrics and semi-outlaws who compose the average adventuring group will earn even less time from the basic NPC civilian.

Here’s a form you can use for each interviewee in an investigative adventure:

  • Reason for Reluctance:
  • Overcoming Reluctance:
  • Clue supplied:
  • Possibility Ruled Out:
  • Next contact:
  • Cut-Off:

Just like a cop show screenwriter, you’ll want to create as many different reasons for brushing off the PCs as possible, for variety’s sake. Informants crucial to your storyline will require reasons directly related to the motivations you’ve preset for them. For walk-on characters, you can choose reasons at random — or start with the reason and build the character from that starting point.

Examples can include:

Complicity: Informant peripherally involved in the crime.

Confusion: Informant is cooperative, but his perceptions are muddled.

Greed: Informant seeks payoff before talking, and drives a hard bargain.

Guilt: Informant has done something bad, but unrelated to the mystery, and fears that this is what the PCs are investigating.

Hostility: Informant has good reason to hate adventurers as a group.

Ideology: Informant belongs to a group or class politically opposed to the PCs or their patrons.

Loyalty: Informant wants to protect someone she (rightly or wrongly) assumes to be the target of their investigation.

Paranoia: Informant assumes PCs are his (real or imagined) enemies.

Preoccupation: Informant more concerned with his own pressing business or agenda than with helping the PCs.

Snobbery: Informant considers himself social better of PCs; recoils at the thought of associating with them.

The manner in which the PCs must overcome the informant’s reluctance arises from the nature of that reluctance.

Complicity: PCs must convince informant they know what he did and can arrange for worse treatment if he doesn’t talk.

Confusion: PCs must sort through informant’s scattered recollections for the important fact.

Greed: PCs must pay him off, or convince him he’ll be worse off if he doesn’t talk.

Guilt: Must assure informant that her particular misdeeds are not their concern.

Hostility: PCs must mollify the informant, or use leverage his grudge against him with intimidation tactics.

Ideology: Informant must be shown how cooperation benefits his faction.

Loyalty: Convince informant cooperation will lead to a better outcome for the person she’s protecting.

Paranoia: Either reassure or terrify the informant.

Preoccupation: Show how lack of cooperation will hurt the informant’s business or cause.

Snobbery: Show how cooperation will lead to the PCs’ speedy departure.

Alternate methods of persuasion should always be possible. Otherwise you risk falling into a variant of the classic plot bottleneck, in which there’s only one way to get a particular piece of information on which all forward development depends. PCs should be able to intimidate snobs or bribe paranoids. For variety’s sake, ensure that no single tactic works on all informants.

Structurally, any investigative adventure consists of a trail of clues leading like bread crumbs from one encounter to the next, so the nature of the clue is up to you.

The next contact positions the encounter within that structure, telling you which new scene the character will point the PCs toward. In a cop show, the leads find the clues in a particular order. If you can prepare several different orders in which the clues can be assembled, you face less chance that a dead end point will arise in mid-scenario. (Putting the encounters on index cards helps if you intend to shuffle them as you go.)

Finally, under the entry labeled cut-off, slot in the reason for the NPC to conclude the encounter after the PCs have squeezed it for all of its information and entertainment value. It’s easier to get NPCs out of scenes in a modern setting with busy schedules and ringing cellphones, but self-respecting supporting characters in any era or genre should be anxious to get on with their own lives as your sense of expediency dictates. Cut-offs may refer back to the character’s original reluctance to talk. A snob wants to shoo uncouth PCs out of his manor as quickly as possible. A paranoid wishes to escape an imagined threat. If the PCs haven’t slapped the cuffs on a complicit character, he will want to leave the jurisdiction as soon as possible.

Unrelated cut-offs work just as well, and provide an added sense of reality to your world. Mundane details like crying babies, overflowing sinks, cookpots in need of tending, escaping horses, or goods in need of protection from the rain all provide otherwise helpful NPCs excuses to bring their discussions with the heroes to an end.

I’d stick around and elaborate, but you have all the clues to piece it together. I have an owl to feed. Or something! Good luck with that investigation, now!