Costly Success and Information Gathering

A column on roleplaying

by Robin D. Laws


GUMSHOE bases itself on the observation that failing to get information is never as interesting as getting it. It gives the characters the clues they seek, provided they have the right skill and look in the right place, with no die roll required.

Since its advent, some designers and game groups have sought a middle way, that of the costly success. When presented with an opportunity to gather information, the player still rolls an ability die (or whatever.) The outcome of the roll determines not whether the character receives the information that advances or complicates the developing story, but whether ancillary consequences arise from the effort to gather it. Depending on the game’s narrative control assumptions, the blowback from a failed roll might be detailed by the GM, the player or, in some cases, the scenario.

I’m hardly going to write a column condemning the costly success. This mechanism appears in several of my designs. You’ll find it in HeroQuest 2, most notably.

Costly successes, though without the die roll, show up in GUMSHOE as well. They most often occur when the most logical ability to gain a particular bit of information is a general one. Say you need to break into a lunar outpost to acquire a recorded surveillance hologram. Here you assume that the character gets the item he’s looking for. His Infiltration test determines whether he’s pursued or captured on the way out.

What GUMSHOE avoids is making every information-gathering moment a costly success. This technique is strongest when used sparingly.

By making the provision of information to the characters the default choice, GUMSHOE is able to construct more complicated mysteries. The action flows more smoothly, even as players sort through a larger set of plot points.

In simpler mysteries dependent on successes to move forward, the number of clues drops. Each becomes a pathway to the next scene or necessary event.

If, as in GUMSHOE, you construct a mystery from a large number of clues, dreaming up an interesting cost for each potential failure becomes increasingly difficult.

The entertainment value of a character failure hangs on two elements, one positive and one negative. It must be novel, and it mustn’t stop the story.

Original failure descriptions add flavor and variety to an adventure.

Failures can certainly forward a story. They can send the characters on a fun diversion into sub-plot territory. As they colorfully unfold, possibly underlining or developing key PC character traits, they could well prove more memorable than the solving of the main mystery. A free-flowing, collaborative RPG narrative must always remain open to this possibility.

At the same time, setbacks can advance the primary plotline:

  • Your capture by the Bleedist rebels leads you to a key witness.
  • As the suspect flees from you, you see that his limping stride rules him out as the perp who left the footprints in the victim’s back garden.
  • When the door slams shut to trap you in the cyclopean tomb, hieroglyphics on its previously unseen surface grant insight into the Yithian extinction.

But many failures, especially when thought up on the spot, under time pressure, simply reverse the characters’ previous successes, or place them in a cul-de-sac, with no clear action at hand.

Thinking up a couple of novel, plot-driving failures is way easier than thinking up dozens of them.

An issue of traffic direction comes up as well. In a scene where there might be twelve or twenty pieces of information to glean, as there are in many GUMSHOE scenarios, what happens when the players fail six or eight times? Does the first failure send them off in a new plot direction, requiring them to come back and repeat this scene a whole bunch of times before they move forward? A structure this circular tangles your group in guaranteed frustration. The economy of plot thread deployment calls for at most a couple of possible failures in a given scene, no matter how potentially fascinating each individual one might be.

To make the costly success principle work for an investigative game requires that you revert to the simpler model, where each scene has one or two gatekeeper points, and progressing from one scene to the next is less a matter of puzzling out the situation than in rolling well.

Taking the costly success approach to every piece of information also fails the CSI test. If you don’t see the characters fail in a particular way in CSI (or the mystery-solving source material of your choice, from Sherlock Holmes to Star Trek to House), having them fail that way in a roleplaying context constitutes an error of emulation. You don’t see the forensic scientists on CSI constantly messing up as they operate their lab equipment, getting information from them anyway, and then enduring some ancillary penalty that takes them away from the current plot thread. The show, like most procedurals, functions as a romance of competence. Its characters confront plenty of obstacles to maintain suspense and narrative rhythm, but not at the cost of its central fantasy: that they are experts in their fields, delivering justice through superior know-how, discipline, and observational power.

Any scene in a properly collaborative roleplaying game has to present several potential lead-outs, so that the choices made by player characters remain meaningful. But a lead-out based on a failed roll is not a true player choice. It avoids the dread specter of the railroading GM by handing that authority over to a set of railroading dice.

A few scenes triggered by setbacks certainly add variety and suspense to the proceedings. Players have to feel that failure is to be feared and avoided.

In gaming, though, we have traditionally overestimated the number of actual failures needed to foster that impression. We’ve internalized this, fetishizing failure in a way that often undercuts the stories we set out to make together.

I used to commit this mistake, too, until I embarked on the analysis that led to Hamlet’s Hit Points. Downbeats of fear and suspense, when varied in an unpredictable pattern with compensating upbeats, keep a narrative compelling. But down moments need not be outright failures on the protagonists’ part. The mere introduction of a fresh obstacle acts as a down beat. Any unanswered question does the same. It introduces anxiety in the audience (or in RPG terms, the players), which is then countered by an upbeat when the character uncovers its corresponding answer.

Moments when the players are confused or frustrated, even when they haven’t missed a roll, also function as down beats. You’ll see plenty enough of these in any investigative scenario without having to burden every essential clue with the possibility of costly success.



This site uses cookies to offer you a better browsing experience. By browsing this website, you agree to our use of cookies.