Generate a random Yes or No in Excel by either typing a formula like =CHOOSE(RANDBETWEEN(1,2),"Yes","No") into a cell, or by using a browser-based Web Crypto tool and pasting the result. In-cell formulas recalculate on every workbook change because RAND and RANDBETWEEN are volatile, which means a Yes you saw ten seconds ago can flip to No while you are still working on the file. A separate generator outside the spreadsheet produces one fixed answer per click and leaves the rest of the workbook untouched, so the value you paste stays put exactly where you placed it. For quick decisions during a session — should I clean this column, should I send this draft, should I ship this version — the external approach is often faster because you do not have to remember the formula syntax, escape the volatility, or freeze the value before something else in the sheet changes it. Both routes can produce a fair 50/50 result, but only the browser tool displays the source, the mapping rule, and the running history in the same view.
Why Excel Formulas Aren't Always the Fastest Yes or No
Excel's built-in random functions are powerful, but they were designed for sampling data, not for settling casual decisions. RAND returns a decimal between 0 and 1, and RANDBETWEEN returns an integer between two bounds. Wrapping either of them in a CHOOSE or IF expression gives you a Yes or No, but the result is recalculated every time the workbook calculates — and that includes the moment you type into another cell, press F9, sort a range, or save the file. The same Yes you saw a moment ago can silently turn into No without warning.
That volatility is fine for simulations and Monte Carlo work, where you actually want the values to keep changing. It is awkward for a one-shot decision. To freeze a generated Yes or No in Excel, you have to copy the cell and paste it back as a value, which adds a step and a place to forget. A dedicated Yes or No Generator sidesteps the whole recalculation issue because the answer lives outside the workbook and only enters Excel when you deliberately copy it in.
How the Yes or No Generator Works
The Yes or No Generator is a small browser page that produces a 50/50 answer using the W3C Web Cryptography API rather than the older Math.random function. Each click calls Crypto.getRandomValues on a single-element Uint32Array, which fills the buffer with a uniform 32-bit integer from 0 through 4,294,967,295. The generator then maps that integer to an outcome: even values become Yes, odd values become No. Because two divides 2^32 exactly, every valid Uint32 value is accepted and each side of the binary split receives exactly 2,147,483,648 source values — the cleanest possible 50/50 split the source range allows.
Rejection sampling is built into the same tested mapper, even though ordinary binary use never triggers it. For outcome counts that do not divide 2^32 evenly, the tool would reject values at or above the largest multiple of the outcome count and try again, capped at 128 draws so a broken source cannot hang the page. The mapping for the two-outcome case is a direct modulo, but the underlying math is exercised at the three-outcome boundary during tests so the rejection path stays verified.
If Web Crypto is unavailable, blocked by the page, or throws for any reason, the tool reports a clear failure and produces no answer at all. It deliberately does not fall back to Math.random, because silently switching the random source would contradict the displayed methodology and hide the problem from you. Per the MDN documentation, getRandomValues is the standard interface browsers expose for cryptographically strong random integers and is the basis for the result you see.
Generate a Random Yes or No in Three Steps
- Optionally type a low-stakes question into the question field. The field accepts plain text and is capped at 1,000 UTF-16 code units, so a long prompt is allowed but a tiny one is fine too. Leaving it empty is also valid and produces a bare Yes or No.
- Click the Generate button. The page requests one Uint32 from your browser's Web Crypto implementation and maps it through the even/odd rule. The current answer appears prominently, alongside a short note about the source.
- Use the newest-first history if you want to compare recent answers, or clear it when you are done. The history holds at most 20 entries; once the twenty-first generation runs, the tool removes the oldest entry and tells you exactly how many earlier answers have been discarded so nothing disappears silently.
Editing the question field clears the displayed current answer so a Yes or No is never mislabeled with wording that no longer matches it. History entries stay where they are, recorded with the exact text you typed at the time. The question, the answers, and the history never leave your tab, and no account, network call, or remote service is required.
Putting the Yes or No Result Into Your Excel Sheet
Once you have an answer, the most reliable way to land it in Excel is to copy the text from the current-answer display and paste it as a value. A few practical patterns work well for this workflow:
- Single decision cell. Type or paste "Yes" or "No" directly into a labelled cell such as B1. Because the value is a string, it will not recalculate or change while you keep working on the rest of the workbook.
- Decision column for a list. Paste the same answer down a column to mark every row at once, for example labelling all rows where you decided "Yes, follow up this week." This is faster than entering a formula and gives you a stable audit trail.
- Reference-based workflow. If you have a question whose answer you want to keep linked to other cells, type the answer in a named cell and reference it with =B1 elsewhere. The reference inherits the pasted value, not a fresh draw, so the rest of the sheet stays consistent.
- Sampling for training or demos. Generate twenty Yes or No values one at a time, then paste them into a column to build a small training set or example sheet. The visible history is useful for picking a varied batch because you can see streaks before they go into the spreadsheet.
If you would rather keep the work inside Excel, the 3 formula approaches guide walks through CHOOSE, IF, and RANDBETWEEN patterns. That approach is the right choice when you genuinely want every cell to redraw on recalculation, which is useful for simulations and very different from settling a single decision.
Generator vs. In-Cell Excel Formula
| Aspect | Yes or No Generator (browser) | In-cell Excel formula |
|---|---|---|
| Random source | Web Crypto Uint32, displayed in the UI | Excel's internal PRNG, not user-visible |
| Recalculation | One fixed answer per click, no recalculation | Volatile — value changes on every workbook recalculation |
| Where the value lives | Outside the workbook, copied in by you | Inside the cell, tied to the formula |
| History | Up to 20 newest entries with discarded count | None — only the current cell value exists |
| Network and storage | No upload, no account, no remote storage | Stored in the workbook file itself |
| Best suited for | Casual one-off decisions and small audit trails | Live simulations and bulk sampling that should keep changing |
The two tools serve different jobs. Use the generator when you want a single frozen decision, and use a formula when you want a sheet full of redrawing values.
Fairness, Streaks, and What to Watch For
With two outcomes dividing 2^32 exactly, the generator's mapping is as balanced as the source range allows: each side gets half the integers. Independence is also part of the contract — every click draws a fresh Uint32 with no influence from the question, the time, the previous answer, or the history list. A streak of three Yes answers in a row is not a sign of bias; it is the normal behaviour of independent fair draws. The recent history is descriptive only and never rebalances later answers to "make up for" a streak.
The tool is also deliberately small in scope. It does not learn preferences, weight alternatives, prove fairness over a small sample, or predict future events. The question field is a label, not a factor. If you need a different probability — say a 70/30 split, or a yes/maybe/no range — the same tested mapper supports a general outcome count, but the standard Yes or No button here is locked to 50/50.
When a Random Yes or No Is the Wrong Tool
A casual 50/50 prompt is not a substitute for judgment, expertise, or proper analysis. Do not use the result for medical, legal, financial, safety-critical, emergency, security, consent, employment, eligibility, or compliance decisions. Web Crypto quality does not make the question appropriate, validate the available choices, or protect you from consequences. A mathematically balanced answer can still be unsuitable, unlucky, or harmful when used outside a low-stakes context. For anything that matters, talk to a qualified person, read the actual data, or use a structured decision framework rather than a random prompt.
For quick spreadsheet decisions, demos, and small audit trails, generating a Yes or No outside Excel and pasting it as a value is a reliable pattern. The answer stays put, the rest of the workbook keeps working, and the source of the randomness is shown in the same view where the result appears.
Related reading: Bar Chart Maker Cheat Sheet: Syntax, Limits, and Output.