The sheer scale of global online fraud is mindblowing. Take Facebook, for instance. According to Statista, Facebook closed down a staggering 1.6 million fake accounts in the first quarter of 2022 alone. But, of course, Facebook isn’t alone in its fight. Every social media platform has to contend with the same relentless flow of fraudsters. Google probably has the most brutal online battle, if only because of the huge numbers involved. Consequently, Google reported that 3.4 billion fraudulent ads were rejected in 2021, nearly twice as many as the year before. Fake ads ranged from phishing scams and bogus products to pop-ups with malware embedded in them and weight loss hoaxes. Of course, online fraud detection requires a multifaceted approach. But one way to weed out the fakes quickly is to use the remarkable Benford’s Law.
Benford’s Law Detects Dishonesty
What is Benford’s Law? Benford’s Law is a way of analysing data. It’s a probability distribution. If the analysis doesn’t demonstrate the expected pattern, the data are probably fake, fraudulent, or mistaken. So, Benford’s Law (also known as the Law of First Digits) states that the leading digits in a collection of data sets will most likely be small. In other words, if a large group of naturally occurring numbers are analysed, such as figures in a tax return, then number 1 should appear more frequently, number 2 less so, and so on down to 9.
This distribution is unexpected but universal across large, ‘real world’ data sets. For instance, this surprising result has been proven to apply to a wide range of data sets, including tax returns, social media accounts, websites, stock market values, street addresses, electricity bills, stock prices, population numbers, house prices, death rates, birth rates, river lengths, country area, and even lottery numbers.