A high-profile murder case. A suspect claiming innocence. And an investigating agency reaching for one of the most misunderstood tools in modern forensics: the polygraph.
The Siya-Ketan case — which has gripped national attention over the past several weeks — has put lie detector tests back in the public conversation. But most of what people believe about polygraphs is wrong. And understanding how they actually work matters, especially when courts and investigators are deciding how much weight to give them.
The Siya-Ketan Case: What We Know
Without going into details that are still sub-judice, the broad outline is this: Siya, a college student, was found dead under circumstances that investigators say were not accidental. Ketan, her partner, has been the primary suspect from early in the investigation. He has denied any involvement.
The Central Bureau of Investigation (CBI), which took over the case after state police faced criticism over the handling of early evidence, sought and received court permission to conduct a polygraph examination on Ketan. He reportedly agreed to the test.
The results — or rather, the fact that a test was conducted — immediately became a major news story. Which brings us to what most people got wrong about it.
What a Polygraph Actually Measures
A polygraph doesn’t detect lies. This is the most important thing to understand, and it’s the thing that almost every popular depiction of the technology gets wrong.
What a polygraph measures are physiological responses: blood pressure, respiratory rate, skin conductance (essentially, how much you sweat), and sometimes pulse rate. The assumption underlying the test is that when a person lies, the stress of doing so produces measurable physiological changes — elevated heart rate, increased perspiration, altered breathing patterns.
An examiner asks a series of questions — some control questions (designed to produce a baseline stress response from most people), some relevant questions (directly about the matter being investigated), and some irrelevant questions (neutral, designed to establish a resting baseline). The physiological responses are recorded and compared across question types.
If your response to relevant questions is significantly more pronounced than your response to control questions, the examiner may conclude that deception is indicated. If responses are similar, the conclusion is “no deception indicated.”
Notice what’s missing: an actual measurement of whether you told the truth.
The Problem With Accuracy
The scientific community has been sceptical of polygraphs for decades, and for good reason. Studies on polygraph accuracy have produced highly variable results, with reliability estimates ranging from 60 percent to 90 percent depending on the methodology, the examiner’s skill, and the specific population being tested.
A 60 percent accuracy rate means you’d do better with a coin flip than you’d realise. Even at 90 percent — which represents optimistic conditions — in a group of 100 innocent people, 10 would be flagged as deceptive. That’s not a margin acceptable in criminal justice.
There’s also the problem of countermeasures. Research has repeatedly shown that people can be trained — in as little than a few hours — to manipulate their polygraph responses. By artificially inflating responses to control questions (through mental imagery, physical tension, or other techniques), the comparison between relevant and control responses can be flattened.
And then there’s the opposite problem: people with certain psychological profiles — sociopaths, highly experienced liars, people on certain medications — may not show the expected stress responses when lying, producing false “no deception” results.
Why India’s Courts Treat Them With Caution
Indian courts do not admit polygraph results as evidence. The Supreme Court has been explicit about this: in Selvi vs State of Karnataka (2010), the Court ruled that administering narco-analysis, brain mapping, and polygraph tests without the subject’s consent violates the right against self-incrimination under Article 20(3) of the Constitution.
Even with consent, the results are not admissible as direct evidence. What they can do, legally, is provide investigative leads — point investigators toward areas of further inquiry, or generate information that, if independently corroborated, can then be used in court.
This is an important distinction. The CBI conducting a polygraph on Ketan doesn’t mean the results will appear before a judge. It means investigators may use whatever comes out of it to decide where to look next.
So What’s the Point?
Fair question, given the limitations.
Investigators often use polygraphs not primarily for the physiological data, but for the interrogation process itself. The extended session, the questions, the controlled environment — these can sometimes produce admissions, inconsistencies, or new information through the psychological pressure of the process rather than the instrument itself.
A suspect who knows they’re being physiologically monitored while answering questions sometimes behaves differently than in a standard interview. Whether that difference is useful depends enormously on the specific case and the specific individual.
In the Siya-Ketan case, investigators have not publicly confirmed what the polygraph revealed — and legally, they may never have to. What happens next will depend on what physical evidence the CBI has assembled, and whether the test pointed them toward anything independently verifiable.
The lie detector makes for good TV. In real criminal investigations, it’s a much messier, more limited tool than the cameras would have you believe.
