Secret Santa Rules That Actually Work
Settle the budget before anything else
The draw is trivial. The budget is where Secret Santa lives or dies, and it fails in a predictable way: somebody proposes “about ten to twenty”, half the group hears ten and the other half hears twenty, and on the day one person unwraps a keyring while the next unwraps a bottle of something proper. Nobody did anything wrong and everyone feels awkward.
State one number. Not a range, not a guideline, not “whatever you like”. One number, in one message, before names are drawn β because once people know who they are buying for, the budget conversation becomes personal.
Two additions that cost nothing:
- A deadline before the event, not on the day. Post arrives late, and a missing gift at a reveal is worse than a cheap one.
- A tone, stated plainly. Joke gifts and real gifts are both fine, and a group where half do each is not.
Run the draw so nobody can draw themselves
The classic method β everyone pulls a folded name out of a hat β fails more often than people realise. The chance that a completely random pairing leaves nobody holding their own name is about 36.8%. That figure is the derangement probability, it converges on 1 divided by e, and it barely moves once the group is larger than about five: a group of six and a group of sixty fail at very nearly the same rate.
| Group size | Chance nobody draws themselves |
|---|---|
| 3 | 33.3% |
| 4 | 37.5% |
| 5 | 36.7% |
| 6 | 36.8% |
| 10 or more | 36.8% |
So roughly two draws in three need redoing, and every redraw is another opportunity for somebody to glimpse a name they should not have seen.
The Secret Santa generator sidesteps the whole thing. It shuffles the list and reads it as a single closed circle: first buys for second, second for third, and the last person buys for the first. Self-matches are impossible by construction, not prevented by retrying, so there is no redraw and no edge case at small group sizes. A side effect worth knowing is that mutual pairs cannot happen either β nobody buys for the person buying for them, which most groups prefer.
Exclusions: be honest about the redraw
The most requested feature in any Secret Santa tool is exclusions β couples who should not draw each other, siblings, the two people who sit together and would find it obvious.
The practical approach is unglamorous and works: generate, look at the result, and generate again if a forbidden pairing appears. With eight people it takes a couple of seconds, and with a single chain the whole assignment is on screen at once so the check is instant.
What matters is the order of operations. Announce the exclusion rule before the draw. A redraw that follows a stated rule is administration; a redraw that follows somebody seeing the result is an organiser choosing the pairs, and everyone can tell the difference afterwards.
If exclusions start multiplying β several couples in a group of eight β the honest move is to accept that the constraint is doing most of the work and assign by hand, openly, rather than pretending chance produced it.
Drawing when everyone is remote
Two arrangements work well, and they differ in who holds the secret.
One organiser draws and tells people privately. Paste the names, generate once, then message each person their recipient individually. Fast and reliable, with the accepted cost that the organiser knows everything. Most workplaces run this way, and it is fine as long as the organiser is disciplined about not steering conversations.
Everyone on a call, one shared screen. Generate once with the group watching, and have each person read only their own line while the others look away. Nobody holds the full picture, but the etiquette is fragile β one screenshot into the group chat and the whole thing collapses. It works best in small, well-behaved groups.
Either way, the link is the useful part. Your entries travel in the page address, so sending the link gives everyone the same participant list to check before the draw. Note that each press of the button produces a fresh chain, so exactly one person should generate the result β two people generating separately will produce two different, equally valid assignments and considerable confusion.
For the arithmetic behind why a chain beats a hat, see how random is random; for the tool itself, the Secret Santa generator takes any list from three people upwards.
Frequently asked questions
How much should the budget be?
Whatever the group agrees, stated as one firm number rather than a range. A range is read as its upper end by half the group and its lower end by the other half, which produces the classic mismatch where one person unwraps a novelty mug and the next unwraps something genuinely expensive. Say the number, in one message, before the draw.
How do I stop couples drawing each other?
Generate the assignment, look at it, and regenerate if a forbidden pairing appears. Because the whole chain is produced at once, checking takes seconds. Announce the exclusion rule before you draw, though β a redraw that happens after everyone has seen the result looks exactly like an organiser reshuffling until they got what they wanted.
Can I run the draw when everyone is remote?
Yes. Paste the names, copy the page link so everyone has the same list, and agree that exactly one person presses the button β each press produces a fresh chain, so two people generating separately would produce two different assignments. Then either share results privately or run it on a screen share where each person reads only their own line.
What if someone joins or drops out after the draw?
Redraw the whole thing and say so. Patching a chain by hand β giving the newcomer the leaver’s recipient β usually breaks the structure and creates a self-match or a stranded person. A complete redraw takes seconds, and announcing it plainly is far better than everyone slowly discovering that the pairs changed.
Should the gifts stay anonymous at the reveal?
Decide in advance and tell everyone. Some groups enjoy the guessing round; others find it excruciating, particularly at work where the gift and the guess both become performance. Neither approach is wrong, but discovering halfway through that people expected different things is the single most reliable way to flatten the mood.
How the randomness works
Sources: