Category
Scheduling & Rotation Randomisers
Chore rotas, turn order, round-robin fixtures, meeting slots, playlist order.
Five tools that randomise across time rather than in a single moment. That changes what fairness means: a rota is judged on totals across a quarter, a turn order on accumulated first turns across a season, a fixture list on whether it is complete, and a shuffle on whether it clusters.
The common thread is that the output is a structure rather than a result. Pressing the button on a name picker gives you a name; pressing it here gives you a twelve-week grid, a season of fixtures, or an ordering across twenty rounds — and the thing to evaluate is the structure's properties rather than any individual cell in it.
Two of these tools are barely random at all, which is worth saying plainly. The round-robin generator uses the circle method, which is completely deterministic, and shuffles only the starting order. The playlist shuffler applies a constraint that makes some orderings unreachable. In both cases the randomness is doing a small, specific job and the structure is doing the rest.
The other three enforce a rule and report a total. The rota assigner enforces no-consecutive-repeats and prints everyone's load; the turn order generator offers three orderings and prints how many first turns each player got; the meeting slot picker intersects availability strictly and refuses when the intersection is empty. In each case the printed number is the output that matters and the draw is the part that gets the attention.
Chore Rota Assigner
Assign chores across weeks so nobody gets the same job two weeks running.
Turn Order Generator
Straight, reversing, or snake order across rounds — first-player advantage, handled.
Round-Robin Schedule Generator
Everyone plays everyone once, in the fewest rounds — by the circle method.
Random Meeting Slot Picker
Draw fairly from the slots that actually work for everyone — intersection first.
Playlist Shuffler
A shuffle that refuses to stack two tracks by the same artist together.
Rotas: the constraint that matters and the total that gets judged
The one rule the rota assigner enforces is that nobody takes the same job in consecutive weeks. That is a deliberate choice about what to spend a constraint on: a run of two is felt immediately and is the complaint that actually kills rotas, while an imbalance of three across ten weeks has to be counted before anyone notices it.
Totals are reported rather than enforced, and that is an honest limitation rather than an oversight. Each week is generated independently subject to the one rule, so the totals converge toward even without being forced there — and enforcing them would be a substantially harder constraint that would frequently have no solution.
What does most of the balancing is round-robin dealing within each week, which keeps the per-week load even when the numbers do not divide. Five people and three chores means two people are off each week, and which two is randomised, so being off does not always fall to the same names.
The reason to generate a whole quarter at once rather than a week at a time is that a regenerated rota knows nothing about what the previous one said. Regenerating weekly defeats both the consecutive-week rule and the totals, and it is the single most common way this tool gets used wrongly.
Ordering across rounds, and what each shape costs
The turn order generator offers three shapes and prints the consequence of each. Fixed order gives one player every first turn, which is what most groups do by default and is the worst of the three. Snake order reverses on alternate rounds, so the two ends of the draw split the first turns exactly and the middle positions get none. Reshuffling each round distributes them evenly on average and guarantees nothing over a short series.
The first-turn tally is the point of the tool. First-player advantage is measured and published for a great many board games, and it varies from statistically undetectable to large enough that competitive play uses a compensation rule as standard — so the right method is a lookup rather than a principle, and the tally converts the argument into a column of numbers.
The round-robin generator is the least random tool in this category. The circle method fixes one entrant and rotates the rest, producing a complete fixture list in n−1 rounds for an even field — which is provably the minimum, since each entrant plays n−1 matches and can play at most one per round. Only the starting order is shuffled, and that is the only thing stopping the same two names meeting in round one every time you generate a season.
The fixture count is the number worth checking before committing to the format: it grows quadratically, so eight entrants is twenty-eight matches across seven rounds while sixteen entrants is a hundred and twenty matches across fifteen. A round-robin for sixteen is a season rather than an afternoon.
Intersection and constraint: slots and shuffles
The meeting slot picker computes a strict intersection: a slot is proposed only if every single person listed it. That strictness is the design decision, because a tool that ranked slots by how many people could make them would produce an answer every time and that answer would routinely exclude somebody — and a slot that misses one person is not a worse answer, it is not an answer.
Among slots that all work, it draws uniformly, which is genuinely useful and genuinely minor: every slot in the intersection is equally valid by definition, and randomising removes the small bias toward whatever suits whoever set up the poll. Above about five people the intersection is frequently empty, and getting that answer explicitly is more useful than a ranked list that quietly puts somebody at six in the morning.
The playlist shuffler is the clearest example on this site of a deliberately non-uniform reordering. A true shuffle of twenty tracks by five artists produces about three adjacent same-artist pairs, which listeners read as a broken playlist — so it shuffles under a no-adjacent-artist rule and shows the comparison count against a plain shuffle of the same list.
There is a hard limit it reports rather than hides. An artist with more than about half the tracks in a list cannot have them separated by any algorithm — eight tracks of fifteen leaves seven others, which is one short — and the tool marks the result as imperfect rather than presenting its best attempt as clean.
Why date and time generators are not in this category
The random date and time generators read as scheduling tools and live in the numbers category, which is a deliberate decision worth explaining rather than leaving as an oddity.
Mechanically they belong there: both work by converting a range into a flat integer — days since a reference point, or minutes since midnight — drawing a fair integer in that range with rejection sampling, and converting back. That is the same mechanic as the number generator with a display layer, and the numbers hub has an authored section explaining exactly that conversion and why randomising month and day separately would introduce bias.
Moving them here would have silently falsified that section, which is a worse outcome than a slightly odd category boundary. Structure serves the writing rather than the other way round, and both tools are linked from here for anyone who arrived looking for them.
Frequently asked questions
What rule does the chore rota actually enforce?
That nobody does the same job in consecutive weeks. Totals are reported rather than enforced, because enforcing them would frequently have no solution.
Why generate a whole quarter at once?
Because a regenerated rota knows nothing about what the previous one said, which defeats both the consecutive-week rule and the totals.
Which turn-order mode should I use?
Snake if going first is worth something measurable and you want a guarantee; reshuffle if you want unpredictability and can accept that it is only even on average.
How many rounds does a round-robin take?
n−1 for an even field and n for an odd one, and that is provably minimal — each entrant plays n−1 matches and can play at most one per round.
How many fixtures will a round-robin produce?
n(n−1)/2. Eight entrants is 28 matches; sixteen is 120, which is a season rather than an afternoon.
Why does the meeting picker refuse instead of suggesting the best slot?
Because a slot that misses one person is not a worse answer, it is not an answer. A ranked list would produce something every time and routinely exclude somebody.
Why is the playlist shuffler not uniform?
Because a uniform shuffle clusters — about three same-artist adjacencies in twenty tracks — and listeners read that as broken. It gives up some orderings on purpose.
When can it not separate an artist?
When one artist has more than (n+1)/2 of the tracks. Nine of fifteen cannot be separated by any algorithm, and the tool says so.
Why are the date and time generators not in this category?
Because mechanically they are number tools, and the numbers hub has an authored section explaining their day-count conversion. Moving them would have falsified it.