Failed Investigation Check in a TTRPG: Keep the Mystery Moving

A failed Investigation check does not have to mean the characters learn nothing. When an RPG mystery depends on one successful roll, a bad result can stall an entire session even if everyone is making sensible decisions. The fix is not to hand out every answer. It is to distinguish finding the essential lead from extracting extra detail, speed, or safety.

This article focuses on searching scenes, examining objects, and piecing together evidence. For a closely related situation involving what characters notice at a glance, see our guide to failed Perception checks. For planning the overall case, see running mysteries without hiding the only clue.

First decide what failure can actually change

Before a roll, ask: what question is the character trying to answer, and what would an unsuccessful attempt change? If a bloodstained receipt lies openly on the desk and the character searches the desk, discovering the receipt does not need to depend on a check. Interpreting its unusual handwriting might. Finding a hidden compartment while avoiding a noisy trap might.

If you cannot imagine an interesting failure beyond “nothing happens,” consider giving the basic information automatically. You can still roll when time pressure, rival investigators, consequences, or degrees of certainty make the result matter.

Use three layers of information

Divide clues into three levels. Essential information tells the players where they can act next. Interpretation explains how evidence fits together. Bonus information reveals a shortcut, weakness, or hidden connection. A check might improve the latter two without blocking the first.

For example, everyone searching the archive finds that the ledger was moved to the old mill. A success reveals the clerk’s stamp was forged. A particularly strong success identifies the workshop that made the counterfeit stamp. The party can move forward regardless, but better investigation produces leverage.

Five consequences that keep the case alive

• Time: the clue is found, but the search takes long enough for the suspect to leave.
• Exposure: the investigators uncover the evidence and accidentally alert someone.
• Cost: recovering the document damages an item, spends a favor, or requires payment.
• Incomplete context: the lead is real, but its significance remains uncertain.
• Choice: two useful trails remain and the group must decide which to pursue first.

Pick the consequence that follows logically from the action. Do not automatically turn every failure into an ambush; predictable punishment teaches players to avoid investigating.

Worked example: the missing magistrate’s seal

The characters search the magistrate’s records room after a forged travel order caused a wagon to disappear. The GM wants them to learn that the order passed through the riverside relay station. That information is necessary to continue and should have several possible sources: a dispatch register, a witness, and a copied receipt.

The rogue searches the desk and rolls poorly. The GM says, “You find the dispatch register, but some entries were scraped away. One surviving note mentions the river relay. While you work, the evening clerk returns and asks why you are handling sealed papers.” Now the party has a lead and a social problem. The failure changed the scene without deleting the adventure.

When should a false lead appear?

A mistaken conclusion can be dramatic, but a purely invented false clue is often unfair if it arises only because of a low roll. Give the party a reason to suspect uncertainty. Perhaps a witness truly remembers a red coat but misidentifies who wore it. The players can compare accounts; they are not punished for trusting the GM’s description of physical evidence.

Never require players to guess that a successful-looking clue is secretly worthless because of a hidden die result. Make uncertainty legible in the fiction.

Let different abilities uncover the same route

Investigation is not the only way to progress. A historian may recognize a seal, a courier may know the route, and a persuasive character may convince a witness to talk. Multiple methods reward creative play and stop one character’s modifier from controlling the whole mystery.

For notes the party can actually understand, use a simple three-clue mystery recap or make a brief list of confirmed facts and open questions.

One-minute pre-session checklist

• What is the essential next lead?
• How many plausible ways can the party reach it?
• What extra interpretation can a strong roll reveal?
• Which cost or complication follows a failed roll?
• How can players check an uncertain conclusion?
• What changes in the world if the group takes too long?

Is giving a clue after failure railroading?

No, provided the players still decide what to do with it. Railroading removes meaningful choices; accessible information enables choices. A discovered address does not force a visit. The party may bargain with a witness, ignore the trail, warn the suspect, or follow a different lead entirely.

A failed investigation can change who knows about the inquiry, how much time remains, or how confidently the characters understand the evidence. Let those consequences become the story, rather than ending the story at a locked door.