Random Yes or No Generator
Press the button and get a yes or a no. Each press is an independent fifty-fifty — the button does not remember what it said last time, does not try to even the score, and is not influenced by your question. The running count under it lets you check that for yourself.
Press the button.
Nothing is remembered between presses — every answer is independent.
How often the same answer comes up in a row
| Same answer in a row | Chance | As a percentage | In practice |
|---|---|---|---|
| 2 | 1 in 4 | 25.00% | Happens every fourth time you press twice. |
| 3 | 1 in 8 | 12.50% | Common enough that you will see it most sessions. |
| 4 | 1 in 16 | 6.25% | About once in every sixteen pairs of presses. |
| 5 | 1 in 32 | 3.12% | Still unremarkable — roughly one run in thirty-two. |
| 6 | 1 in 64 | 1.56% | Starts to feel like the thing is stuck. It is not. |
| 7 | 1 in 128 | 0.78% | Roughly one in a hundred and twenty-eight. |
| 8 | 1 in 256 | 0.39% | Rare in one sitting, ordinary across many. |
| 9 | 1 in 512 | 0.20% | About one in five hundred. |
| 10 | 1 in 1024 | 0.10% | Now genuinely unusual. |
| 11 | 1 in 2048 | 0.05% | About one in a thousand. |
| 12 | 1 in 4096 | 0.02% | One in four thousand: you will probably never see it here. |
These are the odds of a specific run starting on a given press. Across a long session, runs turn up far more often than the individual odds suggest, which is exactly why a fair coin feels rigged.
How this list was built
Each press is independent, and that is the whole design. There is no balancing, no memory and no attempt to make the totals level out. If you get four noes in a row, the chance of the fifth being a no is still exactly one in two. Anything else would not be random.
The running count is there so you can check us. Most yes-or-no pages ask you to take the fairness on trust. This one shows the tally as you go. Over a few hundred presses it will sit close to half and half, and it will still wander — that wandering is what randomness looks like.
Streaks are normal and people badly misjudge them. Asked to fake a coin-flip sequence, people produce far fewer long runs than a real coin does. Six the same in a row feels broken and happens about once in sixty-four starts. The table above is there because the commonest reason people distrust a generator is a streak that is entirely ordinary.
It cannot answer your question, only give you a coin. That is not a disclaimer, it is the honest description: the value of a coin flip is that it forces you to notice which answer you were hoping for. If you are disappointed by the result, you have your answer, and it did not come from the button.
Common questions
Is it really fifty-fifty?
Yes. Each press draws independently with an even chance, and the tally under the button lets you watch it settle near half and half over a few hundred presses.
I got the same answer six times. Is it broken?
No. Six in a row starts about once in every sixty-four presses, so in a long session you should expect it. A generator that avoided streaks would be the rigged one.
Does it remember my previous answers?
No, and that is deliberate. The count on screen resets when you reload the page, and nothing is stored or sent anywhere.
Should I use this to make a real decision?
Use it the way people have always used a coin: flip it, and notice which result you were hoping for. That reaction is usually the more useful answer than the coin.
Other generators
- Random date generator — A date between any two you choose, in every common format
- Random thing generator — An everyday object, with its size and how hard it is to draw or act out
- Random emoji generator — One emoji, with its real name and what it is used for
- Random animal generator — A real species, with where it lives and what it eats