Pine Cove Shores · Activity Grid Review
Short version: the scheduling tool isn't broken. It's following every rule we gave it — to the letter. The team's hand-built grid looks better because, in a handful of spots, a person quietly bent those same rules to make everything fit. This page shows exactly where, and asks how you'd like to proceed.
The setup, in plain numbers
Each cabin gets 6 morning activity classes across the week. The tool fills those classes from a fixed menu of 12 activities, each with its own size limit. Before any rule about gender or repeats even enters the picture, the raw arithmetic is already tight:
57 seats for 48 cabins means the room is ~84% full at best, every single slot. That's before we require boys and girls to be separated, before we protect the must-do lake activities, and before we try to avoid sending a cabin to the same activity twice. Each of those rules removes options. The question isn't whether the tool is trying hard enough — it's whether the rules we wrote leave a solution at all.
The rules the program was told to follow
When we set up the scheduler, these became hard rules — the tool will never violate one, even if that means leaving a cabin with a repeat. That's the crux of everything below.
In any given class, an activity serves one gender at a time — unless it's specifically marked as mixed (only Huma & Big Chillers are).
Each activity has a hard cap (Boats 4, Pool 4, Huma 4, Slick Track 4, Golf 3, Flight Scooters 2, and so on). Never exceeded.
Craft Time is girls-only. Flight Scooters is limited to certain cabin groups. The tool honors these absolutely.
Don't send a cabin to the same activity twice; make sure everyone gets the headline lake activities. The tool optimizes for these after the hard rules are satisfied.
Grid 1 — the team's hand-built grid
This is the grid the team built by hand. It's genuinely good: not a single cabin repeats an activity. But look closely at the amber cells. In those spots, the grid quietly breaks one of the "hard" rules above — putting boys and girls together, or squeezing extra cabins past an activity's size limit. These bends are what create the breathing room that makes zero repeats possible.
Grid 2 — the computer's grid
Same week, generated by the tool. It never breaks a rule: no mixed-gender classes, no over-capacity activities. But because it can't bend, the leftover demand has nowhere to go — and 7 cabins end up doing an activity twice. The red cells are those repeats. Notice they cluster on the girls' side, and on a small handful of activities.
Why the repeats land on the girls
This week has 30 girls and only 18 boys. Because boys and girls can't share most activities, the girls effectively compete for a smaller pool of seats each slot. The activities a girl can reliably land on without a gender clash are just three: Craft Time (girls-only), Big Chillers and Huma (the two mixed activities). That's about 14 seats — for 30 girls — every slot.
Within the rules it was given, the tool leaves exactly 9 seats empty per slot — the unavoidable minimum. It is packing the room as tightly as arithmetic allows. The repeats aren't a sign of a lazy or buggy tool; they're the mathematical consequence of strict rules + a lopsided week. Even the human team couldn't hit zero repeats without bending a rule nine times.
The honest verdict
The hand-built grid and the computer's grid are solving two different problems. The human solved a slightly looser one — bending gender separation and capacity where it felt safe. The computer solved the strict one we actually wrote down. Both did about as well as their problem allows.
So if we want the tool's output to look like the hand-built grid, we don't need to rebuild the tool. We need to decide which rules are truly firm, and which were always meant to flex a little — and tell the tool that.
Where the resolution begins
Each of these tells the tool to relax in the same way a person already does by hand. None require rebuilding the scheduler. They're listed roughly from highest-impact to smallest.
If older cabins sharing Camp Classics, Shports, or Slick Track is actually fine (the hand-built grid assumes it is), mark those as mixed. This is the single biggest source of breathing room.
The hand-built grid put 5–6 cabins in Huma against a limit of 4 — four separate times. If Huma (or others) can actually hold more, raise the cap. If not, that's a rule the humans were over-bending.
A 30/18 week is the underlying pressure. Where overflow cabins land (the Mahalo split) changes the math. Worth knowing how lopsided weeks typically run.
More options on the menu means more fresh choices before a cabin is forced to repeat. Last year's grid had several activities not in the current set.
The one decision we need from you
The hand-built grid tells us that, in practice, the team treats gender separation and a couple of size limits as flexible in specific spots. The tool can do exactly the same thing — but only once we tell it which spots are okay to flex.
Answer those three, and we can tune the tool to produce grids that look like the one your team builds by hand — automatically, every week.