Using a random tool to make decisions sounds like abdication — giving up your judgment in favour of dice. But researchers studying decision-making have consistently found that introducing randomness into certain categories of choice leads to better outcomes, less regret, and healthier group dynamics. The science is more interesting than you'd expect.
Roy Baumeister's research on ego depletion established that decision-making is a finite cognitive resource. The more decisions you make, the worse you become at making them. This isn't a personality flaw — it's a feature of how human cognition works. Executive function draws on the same pool of mental energy as self-control, concentration, and emotional regulation.
The practical consequence is that the quality of decisions degrades across a day. A senior judge who reviews parole applications makes more generous decisions in the morning than in the afternoon — not because they're crueller after lunch, but because the cognitive cost of deliberate reasoning has mounted. Salespeople close fewer deals late in the afternoon. Doctors prescribe more unnecessary antibiotics at the end of long clinic shifts.
For low-stakes decisions — what to eat, what to watch, which task to tackle next — using a random picker doesn't just save time. It preserves cognitive resources for decisions that actually matter. Offloading trivial choices to chance is a genuine productivity and well-being strategy.
Barry Schwartz's research on what he called the "paradox of choice" demonstrated that more options don't make people happier — they make them more anxious and less satisfied with whatever they choose. When you have three options, any of them might be good. When you have thirty, you're certain there's a correct answer and increasingly convinced you'll pick the wrong one.
This analysis paralysis is particularly acute for decisions involving multiple equivalent options — which restaurant, which film, which project to start with. When options are roughly equivalent in value, more deliberation doesn't improve the outcome. It just costs time and creates anxiety.
A random picker cuts this knot cleanly. You've already decided that all the options on the list are acceptable (otherwise you wouldn't have put them there). Letting randomness choose between them doesn't degrade the outcome — it just stops you from spending 20 minutes on a decision that has the same expected value regardless of which option you choose.
When you make a deliberate choice and it goes badly, the regret mechanism kicks in hard. You had agency; you could have chosen differently; you made the wrong call. This regret is psychologically costly and can linger far longer than the decision was worth.
When a random mechanism makes the decision and it goes badly, the regret calculus is different. You didn't choose badly — you chose to use a fair process, and the outcome was unlucky. That distinction is psychologically meaningful. Research on procedural fairness consistently shows that people evaluate outcomes more positively when the process that produced them was perceived as fair, even when the outcome itself is worse than they'd hoped.
This is why "we drew straws" is a psychologically healthy way to handle certain group decisions. Nobody can regret their choice if there was no choice to make — only a fair process to observe.
Random decision tools are particularly powerful in group contexts. Any time a human authority figure makes a selection — who presents first, who gets the good assignment, who goes on the business trip — there's an implicit evaluation happening. Even when the decision is genuinely fair, it doesn't necessarily feel that way to the people who didn't get picked.
This is especially true in hierarchical settings like classrooms and workplaces. Students are acutely aware of teacher favouritism. Employees track who gets chosen for high-visibility projects. Even when the authority figure has good intentions and makes fair choices, the optics of selection create friction.
A visible random mechanism short-circuits this entirely. When a ball bounces through a physics board and lands on a name, there's no favouritism to allege. The teacher didn't pick their favourite student; physics did. The manager didn't pick the employee they like best; the drop did. That visibility of randomness — the fact that you watched it happen in real time — is what makes the outcome feel genuinely fair rather than just claimed to be fair.
None of this suggests that random decision-making is appropriate for all decisions. There are categories of choice where judgment, expertise, and deliberation are essential and cannot be outsourced to chance.
Career decisions, financial decisions, medical decisions, ethical decisions — these are cases where the outcome varies enormously based on the quality of the choice, and where your accumulated knowledge and values should drive the result. A random tool has no role here.
The right frame for using a random picker is: have I already decided that all the options on the list are acceptable? If yes, randomness can choose between them. If you haven't already pre-screened the options to ensure they're all genuinely fine outcomes, you need to do that work first before the random picker becomes useful.
There's an important distinction between randomness that's hidden (a server generates a number and produces a winner) and randomness that's observable (you watch the ball drop in real time). Both might be equally random in a statistical sense, but they produce different psychological effects.
Observable randomness creates trust. When an audience watches a ball bounce through pegs and land on a name, they experienced the randomness directly. There's nothing to allege, nothing to dispute, no hidden process to worry about. This is why a physics-based drop creates a qualitatively different experience than a random number generator — even if both are mathematically fair.
For group decisions, that visible process matters enormously. It's the difference between "trust me, it was random" and "you were all watching when it happened." The second version is the one that actually resolves disputes and builds confidence in the outcome.