Why Do Three Buses Arrive at Once? (The Science of Random Clumping)
By Numeric Forest Team | Published on 08 May 2026
It is a familiar scenario: standing at a stop for twenty minutes, only for three vehicles to appear around the corner at almost the exact same moment. While it may feel like a coordinated inconvenience, this clumping behaviour can often be explored through the lens of mathematical probability.
When events occur at random, our intuition often expects them to be spaced out in a perfectly uniform pattern. However, statistical models show that random events naturally tend to cluster together. One way to model this behaviour-whether analysing transport intervals, incoming messages, or delivery queues-is by using the Poisson Distribution.
The Mechanics of Vehicle Clumping
In a theoretical transport schedule, vehicles might be timed to arrive at fixed intervals. However, real-world variables can quickly alter these timelines, leading to a cascading effect often referred to as "bunching".
If an initial vehicle experiences a brief delay, more passengers accumulate at the subsequent stops. Boarding a larger crowd requires more time, causing the vehicle to drop further behind schedule. Conversely, the following vehicle has fewer passengers to collect, allowing it to move faster and close the gap. This interaction often results in multiple vehicles arriving simultaneously.
The Probability of Clusters
Even in a completely independent system where occurrences do not influence one another, true randomness is inherently less orderly than expected. For instance, if an automated system receives an average of five notifications per hour, they rarely arrive precisely every twelve minutes. Instead, long quiet spells are frequently followed by sudden bursts.
The Poisson Distribution Calculator provides a method to model these distribution patterns. By entering a baseline average rate, the tool calculates the probability of experiencing specific counts within a fixed window.
Analysing the Numbers: The "Three in a Row" Probability
Consider a simplified model of an information desk where individuals arrive independently at an average rate of five per hour. The tool can be used to determine the likelihood of observing exactly three arrivals within that timeframe.
Plugging these specific parameters into the statistical framework yields the following breakdown:
Average Rate (λ): 5 occurrences per interval
Target Event Count (x): Exactly 3 occurrences
Calculated Probability: Approximately 14.04%
A probability of 14.04% demonstrates that this specific count is a relatively common statistical variation. While it might appear unusual, it is a standard outcome within a random distribution model.
The Broader Distribution Profile
Looking at the wider distribution table for an average rate of five shows how a random system is spread across various counts:
| Event Count (x) | Probability | Statistical Context |
|---|---|---|
| 0 | 0.67% | Low probability of a completely clear interval. |
| 1 | 3.37% | Low probability of a highly quiet interval. |
| 2 | 8.42% | Moderate probability of below-average occurrences. |
| 3 | 14.04% | The specific cluster count modelled above. |
| 4 | 17.55% | High probability cluster near the baseline average. |
| 5 | 17.55% | Equally high probability matching the baseline average. |
This table demonstrates that higher groupings (such as four or five occurrences) carry a greater mathematical probability than lower counts under these specific assumptions. Random distribution models naturally allow for clustering and variance.
Modelling Varied Scenarios
This statistical approach can be applied to a variety of everyday counting scenarios where events occur independently at a steady average rate. By adjusting the baseline parameters within the Poisson model, it is possible to observe how probabilities shift:
- The probability of receiving a specific volume of emails during a work hour.
- The likelihood of multiple deliveries arriving on the same day.
- The frequency of customer queries entering a queueing system during peak times.
Disclaimer: This article serves as a general, simplified educational guide to theoretical mathematical models. Real-world logistics, transport operations, and scheduling frameworks involve complex, interdependent variables and operational constraints that differ from independent probability distributions.