← All articles

Why Do Random Draws Repeat in DOT Testing Pools?

Why Do Random Draws Repeat in DOT Testing Pools?

A driver was selected last month, completed the test, and now appears on the new list. It is easy to ask, “why do random draws repeat?” It can feel unfair, especially when another driver has not been selected all year. But a repeat selection is usually a sign that the process is behaving randomly, not that it is broken.

For DOT-regulated testing, the real question is not whether every driver takes turns. The question is whether every eligible driver had an equal chance of being selected for that draw, using a scientifically valid method, with records that prove what happened.

Why do random draws repeat?

Random selection does not create an orderly rotation. It does not remember who was selected last time and move them to the back of the line. If it did, the process would no longer give every eligible driver the same chance on each new selection date.

Think of each random pull as a new event. Once a selected driver completes the required test, that driver generally remains in the eligible pool for future random selections. On the next draw, their chance is the same as every other eligible driver in that pool.

That means a driver can be selected in consecutive months, multiple times in one year, or not at all during a particular period. None of those outcomes alone proves the draw was biased.

For FMCSA-regulated CDL programs, random testing rules require a scientifically valid selection method and an equal chance of selection each time a selection is made. They do not require a system to avoid drivers who were recently tested. The applicable annual random testing rate is measured by selections over the year, not by making sure a matching number of unique people are selected.

Repeats are more visible than probability suggests

A small fleet makes repeat draws feel especially conspicuous. Suppose a pool has 100 eligible drivers and five are selected each month. A given driver has a 5% chance of selection in each monthly pull. Over 12 independent draws, that driver has roughly a 12% chance of being selected two or more times.

Across the full pool, several repeat selections are not surprising. People remember the driver selected twice. They rarely notice the many drivers who were eligible in every draw and never appeared. That is normal human pattern-finding, not evidence that the randomizer favored one person.

A repeat can become even more likely when the eligible pool is small, the selection frequency is high, or drivers enter and leave the pool during the year. A 15-driver operation will see patterns that look more dramatic than a 300-driver fleet because each person represents a larger share of the pool.

What a valid DOT random draw should do

A compliant process is not defined by whether the results look evenly distributed. It is defined by the method, the eligible pool, and the records.

First, establish the pool before the draw. The roster should reflect the employees who are subject to the applicable random testing requirement on that date. If someone was not eligible, they should not be in the draw. If someone was eligible, removing them simply because they were selected recently compromises equal-chance treatment.

Next, run a documented, computer-generated selection. The process should identify the draw date, the number of eligible drivers, the number selected, and the selection method used. The output should show exactly who was selected from that locked pool.

Then, act on the result. Notify the selected employee, coordinate collection, and record each status change. A random pull is not complete because a name appeared on a list. Your file needs to show the path from selection through notification, collection status, and result entry or documented exception handling.

The random method must be defensible, but the final list should also be operationally clean. A driver should not appear twice on the same finalized selection list for the same testing event. Quality systems prevent duplicate names within one draw while still allowing the same driver to be selected again in a later, independent draw.

Equal chance does not mean equal outcomes

This distinction matters when a driver, supervisor, or auditor questions repeat selections.

Equal chance means every eligible driver had the same probability at the time of the draw. Equal outcomes would mean every driver is selected the same number of times. Random systems do not promise equal outcomes in the short term.

A non-random rotation may look fairer on paper because it spreads selections around. But a rotation also makes future selections predictable. If employees can anticipate who is due next, the program has moved away from the uncertainty random testing is designed to create.

Trying to “correct” a valid draw after seeing repeat names creates its own problem. Replacing a selected driver with someone who has not been selected recently changes the result based on a subjective preference. Unless there is a documented eligibility error, do not redraw or manually swap names because the list looks uneven.

There is one practical exception to separate from random fairness: a selected driver may become unavailable or no longer eligible before testing can occur. Your procedure should define how to handle that situation, including whether an alternate selection is required and how the reason is documented. Do not quietly substitute names without leaving a trace.

When repeat draws deserve a closer look

Repeats are normal. Missing documentation is not.

Review the process when the same names recur if you cannot answer basic audit questions: Who was eligible at the moment of selection? How many people were in the pool? What number was selected? When was the draw run? Was the roster changed afterward? Can you show the original result rather than a recreated spreadsheet?

A closer review is also appropriate if results show impossible patterns, such as the same employee listed twice in one completed draw, people selected after termination, excluded safety-sensitive employees, or a selection count that does not support the required annual rate. Those are administration issues, not ordinary random variation.

Spreadsheets create avoidable exposure here. A roster can be overwritten after the fact. A formula can be copied incorrectly. A manager can sort, filter, or replace names without preserving the original list. When an auditor asks what the pool looked like on a specific date, a current spreadsheet is not proof of the historical pool.

Records that make a repeat selection easy to defend

A repeat selection should take seconds to explain when your records are organized. Keep the selection record connected to the pool snapshot used for that draw, not just to a current employee roster.

Your documentation should show the eligible-pool snapshot, the draw date and time, the selection parameters, the documented random-number generation method, and the original selected-driver list. It should also preserve notification activity, collection status, result entry, and changes made after selection.

Timestamped history matters because it answers the questions that create audit friction. If a driver was selected again, you can show that the driver was eligible in both draws. If the pool changed, you can show when and why. If a test was delayed or canceled, you can show the event history rather than relying on someone’s memory.

A system with locked pool snapshots and write-once result records gives the selection its context. That is more useful than a polished report produced after the fact. DOT scrutiny is rarely about whether a list looks random. It is about whether your process can be traced without gaps.

Run the pull, then trust the record

The best response to a repeat selection is not to apologize for it or manually rebalance the list. Confirm the driver was eligible, confirm the draw was run from the correct pool, and follow the same notification and testing workflow you would use for any other selected employee.

Your team should be able to say: this was the eligible roster on this date, this was the documented selection method, this driver had the same chance as everyone else, and this is the timestamped record of what happened next. That is the kind of random program that holds up when DOT asks.