Method
How We Build Our Lists
Last updated
Every generator on this site draws from one curated dataset of 1,207 objects. This page explains where it comes from, the rules each object must meet, how it is checked for safety, and how we make sure every pick is fair.
1. Dataset Size and Categories
The list currently holds 1,207 objects in 20 categories. Counts below come directly from the dataset.
| Category | Objects |
|---|---|
| Food | 111 |
| Kitchen | 85 |
| Toys & games | 75 |
| Nature | 69 |
| Tools & hardware | 68 |
| Sports gear | 65 |
| Electronics | 62 |
| Clothing | 60 |
| Vehicles | 58 |
| Garden & yard | 57 |
| Home decor | 55 |
| Office & stationery | 55 |
| Art & craft supplies | 54 |
| Accessories | 52 |
| Travel & camping | 51 |
| School supplies | 50 |
| Bathroom | 50 |
| Musical instruments | 45 |
| Furniture | 45 |
| Cleaning & laundry | 40 |
2. Where the Objects Come From
The list was written by hand, category by category, from the everyday contents of homes, classrooms, offices, gardens and shops. It was not copied or scraped from other generators or word lists. Each category was drafted as a whole and then checked for gaps (is something common missing?) and duplicates (are two entries really the same object?). Near-duplicates such as “lamp” and “table lamp” were merged or replaced.
3. Rules Every Object Must Meet
- Concrete: something you can see or touch. Abstract ideas, people and animals are excluded.
- Recognisable: most readers should picture the same thing from the name.
- Generic: no brand or trademark names — “flying disc”, not a product name.
- Suitable for all ages: see the safety section below.
- Findable or drawable: 9 pieces of large outdoor scenery, such as a mountain, are included for drawing but tagged so hunts and charades can leave them out.
5. Safety Filtering
Before the list is published, an automated check rejects any object whose name contains a blocked term. The blocklist covers weapons, alcohol, tobacco, drugs and medicines, adult and shocking items, and common trademarks used as generic names. The build fails if a blocked term appears, so it can’t slip through. Separately, 31 objects that are sharp, hot or otherwise hazardous to handle — graters, irons, sewing needles — are marked, and the Kid-safe only filter hides them.
6. Localization
Objects have a stable ID and a separate name for each language, so translations don’t change which object is which. Lists are adapted rather than translated word for word: a language may hide objects that are unfamiliar locally and add ones that are common there. Each language goes live only after a fluent speaker has reviewed every page and its object names; until then it isn’t indexed by search engines.
7. How Picks Are Made
All picks happen in your browser using crypto.getRandomValues(), the browser’s cryptographically secure random number generator. Random numbers are turned into choices with rejection sampling, which avoids the small bias of the common “remainder” shortcut, and objects are chosen with a partial Fisher–Yates shuffle, so every matching object has the same probability. With No repeats on, the generator works through every matching object before any comes back.
8. Fairness Test Results
On 2026-11-03 we ran the generator’s own code 100,000 times, picking one object at a time from all 1,207. Each object was expected about 82.85 times; the actual range was 58 to 111, with a standard deviation of 8.85 against 9.1 expected by chance. A chi-square test gave p = 0.905, which shows no evidence of bias. In 20,000 lists of 5, 0 contained a duplicate, and No repeats showed every object in the minimum 242 rolls.
- Observed
- Expected by chance
Each bar counts the objects that were picked a given number of times. The dashed line is the distribution pure chance predicts; the closer the bars follow it, the fairer the generator.
Show the data as a table
| Times picked | Objects (observed) | Objects (expected) |
|---|---|---|
| 55–59 | 4 | 3.8 |
| 60–64 | 20 | 18.3 |
| 65–69 | 56 | 59.5 |
| 70–74 | 121 | 135.2 |
| 75–79 | 244 | 220 |
| 80–84 | 253 | 261.3 |
| 85–89 | 225 | 230.9 |
| 90–94 | 170 | 154.2 |
| 95–99 | 81 | 79 |
| 100–104 | 21 | 31.4 |
| 105–109 | 10 | 9.8 |
| 110–114 | 2 | 2.4 |
9. Updates and Feedback
The list changes when readers suggest objects, report mistakes or point out an object that doesn’t suit its tags. Every change goes through the same automated checks, and we rerun the fairness test whenever the list or the generator code changes. Suggestions are welcome on the contact page.