How to Run Secret Santa With Exclusions and No Self-Matches
A Secret Santa assignment is a constrained permutation: every participant gives once, receives once, and may not be assigned to a prohibited recipient such as themselves.
A worked example with numbers
Six participants A, B, C, D, E, and F each need one recipient. A valid result is A→C, B→A, C→F, D→B, E→D, F→E: every recipient appears once and no person receives themselves. If A cannot draw B and C cannot draw D, test those two exclusions as well. There are !6 = 265 self-free assignments before extra exclusions; that count comes from the derangement sequence documented by Wolfram MathWorld’s derangement reference. Extra prohibitions reduce the valid set, so a tool must check that at least one complete assignment remains.
How to set the rule before the result
Collect the participant list and exclusions before generating. State whether couples, housemates, or previous-year pairs are exclusions, and clarify spelling so each person has one identity. Use the Secret Santa generator for the full assignment, not a one-winner tool. After the draw, verify three counts: six givers, six distinct recipients, and zero prohibited edges. Deliver only each person’s own recipient rather than publishing the entire map.
Common mistakes that change the odds or the process
Do not let each person pick a slip independently: the last person can be left with themselves. Do not add an exclusion after names have been sent and then silently reroll only one pair. Do not assume no self-match is enough when a household or safeguarding rule prohibits other pairs. Do not publish recipient names in a shared log merely to prove the draw was random.
Where this method stops being appropriate
This process suits a voluntary, low-stakes gift exchange. It cannot solve coercion, unequal budgets, privacy, travel, or a participant who does not want to take part. Make the spending limit and opt-out rule a human agreement before randomisation; a valid permutation only says that the matching constraints were met.
How the random source fits into the rule
The mathematical object is a derangement, a permutation with no fixed point, as defined by Wolfram MathWorld’s derangement reference. Browser randomness can be used to choose among valid assignments through MDN’s Crypto.getRandomValues reference, but the protection comes from validating every constraint before private delivery. Keep only the minimum record needed to resolve an error.
Test constraints before sending any recipient
| Check for 6 people | Valid result |
|---|---|
| Givers | 6 distinct names |
| Recipients | 6 distinct names |
| Self-links | 0 |
| Declared exclusions | 0 violated links |
All four checks concern the completed map. A sequence can look plausible at its beginning and still fail at the final participant.
Handle an impossible exclusion set honestly
If one participant is permitted to give only to themselves or to a prohibited person, there is no valid assignment. Ask the group to relax a named restriction, change participation, or use a different activity; do not hide an exception in a private reroll. That is a constraint problem, not bad luck.
Recalculate the stated case before using the result
Six participants A, B, C, D, E, and F each need one recipient. A valid result is A→C, B→A, C→F, D→B, E→D, F→E: every recipient appears once and no person receives themselves. If A cannot draw B and C cannot draw D, test those two exclusions as well. There are !6 = 265 self-free assignments before extra exclusions; that count comes from the derangement sequence documented by Wolfram MathWorld’s derangement reference. Extra prohibitions reduce the valid set, so a tool must check that at least one complete assignment remains.
Use the matching tool rule, not a convenient substitute
Collect the participant list and exclusions before generating. State whether couples, housemates, or previous-year pairs are exclusions, and clarify spelling so each person has one identity. Use the Secret Santa generator for the full assignment, not a one-winner tool. After the draw, verify three counts: six givers, six distinct recipients, and zero prohibited edges. Deliver only each person’s own recipient rather than publishing the entire map.
Keep this allocation within its stated boundary
Do not let each person pick a slip independently: the last person can be left with themselves. Do not add an exclusion after names have been sent and then silently reroll only one pair. Do not assume no self-match is enough when a household or safeguarding rule prohibits other pairs. Do not publish recipient names in a shared log merely to prove the draw was random. This process suits a voluntary, low-stakes gift exchange. It cannot solve coercion, unequal budgets, privacy, travel, or a participant who does not want to take part. Make the spending limit and opt-out rule a human agreement before randomisation; a valid permutation only says that the matching constraints were met.
Keep the stated source and inputs with the result
The mathematical object is a derangement, a permutation with no fixed point, as defined by Wolfram MathWorld’s derangement reference. Browser randomness can be used to choose among valid assignments through MDN’s Crypto.getRandomValues reference, but the protection comes from validating every constraint before private delivery. Keep only the minimum record needed to resolve an error. The numerical example above must remain attached to its population, units, and replacement rule; otherwise a later reader cannot tell which probability or match count it describes.