This is the right question to ask, and the answer deserves more than a number on a banner.
The figures you will see quoted
The most commonly cited research is the American Polygraph Association’s 2011
Meta-Analytic Survey of Criterion Accuracy of Validated Polygraph Techniques
(Polygraph, 40, 194–305). Across 38 studies it reports roughly 89%
accuracy for a properly conducted single-issue examination and about 85% for
broader multi-issue screening.
Two caveats belong next to those numbers, and most sites leave them out. The first is that the
research is the polygraph profession’s own rather than independent. The second is that the US
National Academy of Sciences, reviewing the same field, put the figure lower and warned against
treating a polygraph result as proof of anything on its own.
We quote the source because, under the CAP Code and the DMCC Act 2024, an objective claim in
marketing has to be evidenced before it is published — and “a number everyone in the industry
uses” is not evidence. If a competitor’s website promises you 99% accuracy with no source, that
tells you something about the competitor.
What the figures depend on
Accuracy is not a property of the machine. It depends on:
- The examiner. Training, current practice and how carefully they work matter more
than the equipment. - The questions. A narrow, single-issue examination about one specific event is
far more reliable than a broad sweep across years of someone’s life. - The examinee’s fitness to be tested. Certain medical conditions, some
medications, alcohol, drugs and extreme distress on the day can all make a valid examination
impossible. - The conditions. This is one reason we test at a office rather than in someone’s
front room.
Inconclusive results are real
Sometimes the charts do not support a clear opinion either way. That is an inconclusive result,
and it is a legitimate outcome rather than a failure or a fudge. We will tell you if that is what
happened.
The honest summary
A well-conducted polygraph examination is a useful, structured piece of evidence about how
someone responded to specific questions. It is not a certainty machine, it is not infallible, and it
is not accepted by the UK courts. Anyone who tells you their test is 100% accurate is telling you
something they cannot possibly evidence.
[OWNER TO REVIEW and supply a cited accuracy figure before publishing.]