Daily Inspiration
Design excellence, every day.
Jury-selected work from the A' Design Award, presented fresh each morning.
Each of the 2,118,760 possible combinations is equally probable.
Selecting 5 unique numbers from a pool of 50 is a fundamental operation in combinatorics. The result is a combination: a set where order does not matter and repetition is forbidden. The number of possible outcomes follows the binomial coefficient C(50, 5) = 2,118,760. Every single one of those combinations has exactly equal probability of appearing. This is the mathematical guarantee behind fair sampling.
In 1938, Ronald Fisher and Frank Yates published a method for generating random permutations in their book Statistical Tables for Biological, Agricultural and Medical Research. The algorithm works by walking through the pool and swapping each position with a randomly chosen position from the remaining elements. For drawing a subset, only the first k swaps are needed. This partial shuffle runs in O(k) time and guarantees that every combination of k items is equally likely, provided the underlying random source is uniform.
The tool on this page implements a partial Fisher-Yates shuffle powered by crypto.getRandomValues(), the Web Cryptography API built into your browser. This is the same entropy source that secures TLS connections and encrypted messaging. Hardware-level thermal noise and electrical jitter produce randomness that quantum mechanics proves is fundamentally unpredictable. The algorithm runs entirely on your device. The server delivers this page and does nothing more.
Intuition fails spectacularly when estimating overlap probability. If two people independently draw 5 numbers from 1 to 50, what is the chance they share at least one number? For 5 from 50, the answer is approximately 42%. Nearly half the time, two independent draws will collide on at least one value. This is a variation of the birthday problem, one of probability theory's most celebrated counterintuitive results. The original birthday formulation shows that in a group of just 23 people, the probability of a shared birthday exceeds 50%. The mechanism is the same: there are far more pairs to compare than people expect.
Random sampling without replacement is the backbone of survey methodology, clinical trials, and quality control. Jerzy Neyman formalized the mathematical theory of sampling in 1934, establishing that a properly drawn random sample allows valid inference about an entire population. When a pharmaceutical company tests a new treatment on 500 patients drawn from a population of millions, the fairness of the selection process is what makes the results generalizable. Each patient must have an equal probability of inclusion, and no patient can appear twice. This is precisely what "unique random numbers from a range" accomplishes in abstract form.
Random number sets are a versatile teaching tool. Have each student visit /numbers/6/1/49 and draw a set. Ask the class to predict how many students will share at least one number with a classmate, then verify. The gap between prediction and reality is a productive entry point for discussing conditional probability and the birthday problem. For a more advanced lesson, have students repeatedly draw 5 from 50 and track the running average. The convergence chart on this page shows the class average approaching 25.5 in real time.
The tool adapts to any lesson plan through the URL. Try /numbers/3/1/10 for younger students exploring small sets, or /numbers/10/1/100 for larger exploration. No accounts, no sign-ups, no data collection. Students use the tool and leave no trace.
Every number set on this page is generated entirely inside your browser. The server delivers the page with its parameters baked into the URL. Your device creates the random outcome using hardware-seeded randomness from the Web Cryptography API. Your draw history lives in your browser's localStorage, under your control alone. Sharing the URL sends the tool configuration to a friend, and their device generates completely independent results. Two people visiting the same link will almost certainly see different number sets. The URL carries the tool. Your device carries the randomness.
The URL defines the tool completely. Adjust count and range directly in the address bar:
Pick a preset or enter your own count and range. The URL updates, the tool reloads.
Send this link. They get the same pool, their own draw. Compare sets and count overlaps.
Daily Inspiration
Jury-selected work from the A' Design Award, presented fresh each morning.
