The Stadium Seat Lottery

‍

‍

The 2026 FIFA World Cup ticket ballot opened this year to absolute, predictable chaos.

‍

You likely sat at your kitchen table, aggressively refreshing a browser window, hoping to put your name in the hat for a single, modest general admission seat. Somewhere across town, a major corporate sponsor quietly submitted an automated application for 10,000 premium suite seats. When the digital draw finally happened, the sponsor secured their massive allocation, and you received a politely worded "Better luck next time" email.

‍

It feels entirely lopsided. But the algorithm didn't look at the number of people entering the draw; it looked at the number of entries. The entity holding the most units naturally took up the most space in the statistical hat.

‍

It is a system built entirely on probability, heavily weighted toward whoever holds the most volume. That exact same probability engine is what keeps the world's financial statements from collapsing.

‍

Here is the uncomfortable truth about data analysis: treating everything equally sounds great in theory, but it is a massive vulnerability in practice. In traditional sampling, teams might just pick 50 random invoices from a pile. Every invoice gets the exact same chance, whether it is a receipt for a $10 sandwich or a massive contract for $10,000.

‍

When the primary concern is material misstatement, if we are concerned about revenue being reported as higher than it truly is, treating every piece of paper equally means the biggest risks might quietly slip right through the net.

‍

This is why Monetary Unit Sampling (MUS) is so widely relied upon. Instead of looking at a $100,000 ledger and seeing 50 individual customers, MUS completely changes the lens. It treats every individual dollar as its own "sampling unit." The system simply sees 100,000 individual units of $1.

Because a $10,000 invoice contains significantly more individual dollars than a $100 invoice, it has a much higher chance of being selected. This core concept is known as "Probability Proportional to Size" (PPS). Simply put, the bigger the transaction, the more likely it is to get audited.

‍

Under ISA 530, the goal is to design a sample that provides a reasonable basis for conclusions. MUS achieves this beautifully because the math automatically hunts down the larger items. It focuses valuable time on the transactions where a mistake would actually hurt the financial statements, and it often requires a smaller sample size to get there. As a bonus, teams don’t have to manually stratify "big" and "small" items, the math simply does it for them.

‍

‍

The mechanics of the draw are brilliant in their simplicity. First, teams determine a sampling interval (the "jump") based on Materiality and risk. After picking a random starting number, they systematically count through the ledger and pick every nth dollar. The absolute beauty of this logic is that if a client's balance is so exceptionally large that it exceeds the sampling interval, it is mathematically guaranteed to be picked.

‍

Of course, if a mistake is found, ISA 530 dictates that the error must be projected to estimate how much the entire population might be off. If that projection breaches Materiality, a very serious conversation is waiting to happen.

‍

But it isn't a flawless net. While excellent for overstatements, MUS is terrible at finding things that simply aren't there. If a massive invoice is accidentally recorded as $1 instead of $10,000, it has almost zero chance of being picked. Furthermore, customers with a $0 balance or a negative credit balance will never be picked by the algorithm and must be tested separately.

‍

Auditing isn't about giving every transaction a fair and equal chance to be seen. It is about understanding who bought the most tickets to the lottery. When teams properly document the source of the population, the sampling interval, and the random start, they are proving exactly how that statistical lottery was run.

‍

P.S Nailing the sampling draw is only one piece of the puzzle. Once the selection is made, teams still have to clean the messy ledgers, tick and bash the PDFs, and risk-score the journals without draining their cognitive battery. We are hosting a live webinar walking through how the entire Excel experience can be optimised. You don't need to enter a blind ballot for a seat to this one. Grab a ticket to the webinar right here.

‍

‍