Benford’s Law, explained.
Benford's Law describes a leading-digit pattern in some datasets: smaller first digits occur more often, with 1 appearing about 30% of the time.
Why it happens
Processes spanning multiple orders of magnitude can distribute values approximately evenly on a logarithmic scale. Leading digits then occupy unequal portions of each logarithmic decade. This is different from selecting single digits uniformly.
In many scale-invariant collections, smaller first digits occupy a larger fraction of logarithmic space.
Read the result
Compare observed leading digits with the reference pattern and inspect how the data was generated. A close or poor fit is informative only when the dataset is suitable for this comparison.
A worked example
Numbers across scales
Imagine measurements spread across many scales, from tens to millions.
On the Benford reference curve, a leading 1 has probability log10(2), about 30.1%; a leading 9 has about 4.6%.
The imbalance comes from logarithmic intervals, not from the digit 1 having a special causal influence.
OPTIONAL DEEPER DETAILGo deeper: inside the model
Inside this model
The theoretical first-digit probability is log10(1+1/d). The chart shows digits one through nine and highlights the chosen digit.
Where this idea is useful
A practical use
Use first-digit patterns as one screening clue when auditing suitable numeric data.
A common misconception
“Any non-Benford dataset contains fraud.”
Assigned IDs, narrow ranges, minimums, and rounding can break the pattern naturally. A mismatch is not evidence of fraud by itself.
What this explanation leaves out
- This is not a fraud detector by itself; assigned numbers, narrow ranges and many other data sets do not follow this law.
Should telephone numbers follow Benford's Law?
Usually no. Their digits are assigned under numbering rules rather than generated by a scale-spanning measurement process.
What generates these numbers, and does that process justify the reference pattern?
Associated thinkers
Further reading
Explore the original research or the teaching reference behind this experiment.