Case File 314 · Investigation Psychology

Why Should Evidence That Contradicts Your Theory Matter More?

Evidence that contradicts your theory should matter more because it has greater diagnostic value. Confirming evidence tells you your theory *could* be right, but disconfirming evidence tells you somet

By Murder Book Editorial TeamPublished by Gungrounds, the studio making Murder BookPublished 3 October 2026Updated 3 October 2026

Evidence that contradicts your theory should matter more because it has greater diagnostic value. Confirming evidence tells you your theory could be right, but disconfirming evidence tells you something specific is wrong with it. In investigation, a single fact that does not fit the working hypothesis can reveal a flaw in the entire case, while ten facts that do fit may simply show the theory is plausible, not that it is correct. The philosopher Karl Popper formalised this asymmetry: no amount of confirming observations can prove a theory true, but a single genuine counterexample can prove it false. In practice, investigators rarely face such clean logic, but the principle still applies. Theories that survive serious attempts at disconfirmation are stronger than theories that have only been confirmed.

Key takeaways

Disconfirming evidence is more informative than confirming evidence because it can rule out a hypothesis, while confirming evidence can only show a hypothesis is consistent with the facts.
People naturally gravitate toward confirming evidence and tend to dismiss or reinterpret evidence that contradicts their beliefs.
In criminal investigation, ignoring disconfirming evidence is a hallmark of tunnel vision and a documented contributor to wrongful convictions.
Structured techniques such as competing-hypothesis analysis and "consider the opposite" help investigators give proper weight to contradictory evidence.
Disconfirming evidence does not always mean the theory is wrong; it means the theory needs to be re-examined.

Why is disconfirming evidence more valuable?

The logic is straightforward. Imagine an investigator's theory is that Suspect A committed a burglary on Tuesday evening. Evidence placing Suspect A near the scene on Tuesday is consistent with the theory but does not prove it, because many people may have been near the scene. Evidence showing that Suspect A was in another city on Tuesday, confirmed by independent records, is inconsistent with the theory and demands a response: either the theory is wrong, or the alibi evidence is unreliable. That single piece of contradictory evidence forces a rethinking that no amount of consistent evidence does.

Wason's 1960 experiment demonstrated this principle in a controlled setting. Participants given the sequence 2-4-6 were asked to discover the underlying rule by proposing test sequences. Most participants tested only sequences that confirmed their initial guess, such as "numbers increasing by two," and never tried a sequence that might have disproved it. Those who did test disconfirming cases, such as 1-3-9 or 10-5-1, were far more likely to identify the correct, broader rule: "any three increasing numbers." The experiment showed that seeking disconfirmation is both rare and highly effective.

Klayman and Ha (1987) refined this analysis in Psychological Review. They showed that the common "positive test strategy," testing cases expected to confirm the hypothesis, is not always irrational. In many real-world situations it works well enough. But it fails precisely when the hypothesis is narrower than the truth, which is the situation that matters most in investigation. When your theory accounts for only part of the evidence, testing competing hypotheses against each other is the most reliable way to find out.

Why do people resist disconfirming evidence?

Nickerson's 1998 review catalogued the human tendency to favour confirming information across hundreds of studies. The bias operates at multiple levels:

Search. People preferentially seek information that supports their existing beliefs rather than information that might challenge them.
Interpretation. When evidence is ambiguous, people tend to interpret it as consistent with what they already believe. In investigation, this might mean reading a witness's hesitation as suspicious if the investigator already suspects them, or as understandable nervousness if the investigator considers them a witness rather than a suspect.
Memory. People remember confirming evidence better than disconfirming evidence, which means the balance of evidence in their mind shifts further toward confirmation over time.
Persistence. Even when the original basis for a belief is removed, the belief itself often persists. This is belief perseverance, and it means that discrediting a piece of evidence does not automatically undo the conclusion it supported.

In criminal investigation, these tendencies combine into what Findley and Scott (2006) called tunnel vision: a process where the investigation narrows around one suspect or theory, and all evidence is filtered through that lens. Disconfirming evidence is the first casualty. An alibi that does not quite check out gets dismissed rather than investigated further. A witness who contradicts the working theory is treated as unreliable rather than informative. Physical evidence that could support an alternative explanation is set aside.

What happens when disconfirming evidence is ignored?

The consequences in criminal investigation can be severe. Findley and Scott's 2006 analysis described how investigators can "focus on a suspect, select and filter the evidence that will 'build a case' for conviction, while ignoring or suppressing evidence that points away from guilt." The longer this continues, the harder it becomes to reverse course. Colleagues, prosecutors and witnesses who have been told the working theory reinforce it, and the difficulty of abandoning an investigation theory grows with each passing day.

The Innocence Project's data on DNA exonerations illustrates the scale of the problem. Eyewitness misidentification, a form of evidence that can appear powerfully confirming but may be wrong, has been a factor in a large proportion of overturned convictions. In many of these cases, disconfirming evidence existed but was not given adequate weight.

How should investigators handle disconfirming evidence?

Research suggests several practical approaches:

Treat it as a signal, not noise

The first instinct when evidence contradicts a theory is often to explain it away. A more productive response is to treat it as a signal that the theory may need revision. This does not mean abandoning the theory at the first contradictory fact; it means pausing to assess whether the theory can accommodate the new evidence or whether it needs to be rethought. If a suspect's alibi checks out, the investigation needs to account for that rather than dismissing it.

Use competing-hypothesis frameworks

Maintaining two or more working hypotheses and evaluating each piece of evidence against all of them makes disconfirming evidence harder to overlook. When evidence supports Hypothesis A but contradicts Hypothesis B, that distinction is informative. When evidence is equally consistent with both, it is less useful for choosing between them. This structured approach helps investigators recognise which evidence is actually diagnostic and which is not.

Ask what disconfirmation would look like

Before searching for evidence, specify what findings would weaken or disprove the current theory. Then look for those findings with the same diligence applied to confirming evidence. This is the practical application of actively open-minded thinking, and it converts an abstract principle into a concrete investigative step.

Document and revisit

Recording both supporting and contradicting evidence in writing, and reviewing the balance periodically, helps counteract the memory bias that causes disconfirming facts to fade. Ask and Granhag (2005) showed that the need for cognitive closure makes investigators less sensitive to exonerating information; written records provide a check against that tendency.

How does Murder Book reflect this principle?

Murder Book: The Art of Interrogation is an upcoming narrative interrogation game from Gungrounds, an independent studio in Zagreb, Croatia. Development screenshots show an evidence board where players connect statements to other evidence, and a case select screen that includes a "not guilty" result. The game's design means that evidence does not always point where you expect. A statement that looks damning in isolation might make sense when placed alongside another statement, and pursuing only the confirming interpretation can lead to a wrong conclusion.

The in-game casebook lists behaviours such as Stonewalling, Stress Response and Emotional Fading. The psychology research reviewed above shows why these behaviours are not proof of anything on their own. A suspect showing stress might be lying, or might be an innocent person under pressure. The game rewards players who weigh contradictions carefully rather than building a case around the first reading that seems to fit, which is precisely the lesson that disconfirming evidence teaches.

Murder Book is in development for PC (Steam) and mobile, with release currently planned for the end of Q2 2027.