Trivia Hub
Quiz yourself across eight topics — geography, space, animals, science, history, maths, food and film — with every question checked against a source.
In the raw question banks the answer is option two 93.8% of the time
The shuffle in the source is documented as fixing “a heavy positional bias”, which turns out to be an understatement. Counting where the correct answer sits across all 1,653 questions in the 8 banks:
| Slot | Questions | Share | Uniform would be |
|---|---|---|---|
| Option 1 | 67 | 4.1% | 25.0% |
| Option 2 | 1,550 | 93.8% | 25.0% |
| Option 3 | 28 | 1.7% | 25.0% |
| Option 4 | 8 | 0.5% | 25.0% |
194 times more answers in the second slot than the fourth. Chi-squared against a uniform distribution is 4174 on three degrees of freedom, where 16.27 would already be conclusive.
| Topic | Questions | Answer positions | Commonest slot |
|---|---|---|---|
| World Geography | 207 | 13 / 180 / 11 / 3 | 87.0% |
| Space & Astronomy | 225 | 25 / 190 / 9 / 1 | 84.4% |
| Animal Kingdom | 207 | 6 / 194 / 5 / 2 | 93.7% |
| Science Basics | 205 | 10 / 195 / 0 / 0 | 95.1% |
| World History | 201 | 7 / 193 / 0 / 1 | 96.0% |
| Math & Numbers | 203 | 4 / 196 / 3 / 0 | 96.6% |
| Food & Drink | 203 | 1 / 201 / 0 / 1 | 99.0% |
| Movies & Entertainment | 202 | 1 / 201 / 0 / 0 | 99.5% |
It is not one careless bank. All 8 lean the same way, from Space & Astronomy at 84.4% to Movies & Entertainment at 99.5% — 201 of 202 questions with the answer in the same place. 3 banks never once put it last. The cause is authoring rather than carelessness: you write the real answer down first and invent distractors around it, and “one obviously wrong, then the answer, then two plausible” is a natural rhythm. Repeat it fifteen hundred times without checking and the pattern is total.
Which is why the options are shuffled every time one is served
Without it, always picking the second option would score 11.3 out of 12 — against 3 by guessing, 3.75 times better. That is not a quiz. Hand-balancing 1,653 questions would mean rewriting most of them; shuffling costs four lines and removes the bias completely.
| Slot | Before | After one shuffle each |
|---|---|---|
| Option 1 | 4.1% | 24.3% |
| Option 2 | 93.8% | 26.5% |
| Option 3 | 1.7% | 25.3% |
| Option 4 | 0.5% | 23.8% |
| Chi-squared | 4174 | 2.8 |
The shuffle itself is a correct Fisher-Yates: 200,000 runs on a single question put the answer in each slot 24.88 to 25.17 per cent of the time. And it preserves what it must — across every question in every bank the shuffled options are the same four strings and the moved answer index still points at the same one. A shuffle that lost track of the answer would be worse than the bias it was fixing.
And the banks are otherwise clean
Every question has exactly 4 options, an answer index inside range and a follow-up fact. The smallest bank holds 201 questions against a run of 12, so nothing repeats within a session and the bank is nowhere near the limiting factor. One question appears in two banks — “What is the speed of light in a vacuum, roughly?” is in both Space & Astronomy and Science Basics. A run draws from a single topic, so nobody sees it twice in one game, and it is a reasonable thing for those two topics to share.
How to use
- Pick a topic and how many questions you want.
- Answer without looking anything up — the point is retrieval.
- Read the fact attached to each answer, including the ones you got right.
- Try a topic you know least about rather than one you already do well in.
- Play again for a different set of questions from the bank.
Frequently asked questions
Where do the questions come from?
They are written for this site and checked against reference sources before being included. Facts that turn out to be contested, or that have since changed, get corrected — a question carrying a stale answer is worse than having no question at all.
Why does every question have a fact attached?
Because getting an answer wrong and being told only that you were wrong teaches you nothing at all. The explanation is the part actually worth reading, and quizzes that leave it out are testing you rather than teaching you anything.
Does quizzing actually help you remember?
Yes, and the effect is one of the better established findings in learning research. Retrieving an answer from memory strengthens it considerably more than rereading the same material does — which is why self-quizzing beats highlighting, despite highlighting feeling more productive.
Are the questions the same every time?
No. They are drawn from a larger bank and shuffled, so repeated rounds give you different questions. The order of the answer options is shuffled as well, which stops you learning the position of the right answer rather than the answer itself.
How is the difficulty balanced?
By mixing questions most people will know with some that few will, within each topic. A round where you get everything right teaches nothing, and one where you get nothing right is merely discouraging — the useful range sits between those two, and that is what the bank aims at.
What should I do about a question I think is wrong?
Say so through the contact page. Trivia goes stale in specific and predictable ways — records get broken, scientific classifications change, and disputed facts eventually get settled — and a correction that comes with a source attached gets acted on quickly.
Is there a way to make this harder?
Answer before reading the options, then check whether your answer appears. Multiple choice gives away far more than it seems, because recognising the right answer is much easier than recalling it, and covering the options turns an easy round into a genuinely testing one.
🔒 This tool runs entirely in your browser. Nothing you enter is uploaded, logged, or stored.