🔚 [Long weekend people] All 10 last days are "falling" for rooms?

The end of a long weekend reads like a travel-home day, so people write it off. Sort by room rate instead and the last day turns out to sit on the convenient side of the ledger, every time. It is the exact mirror of the middle days.

On the last day of a run, the crowd and the room rate point the same way for once. Of the 724 days we score, 10 are last days, one from each of 10 runs. All 10 land on the falling side for room rates. On crowds, 9 of them — 90% — sit below the yearly average.

The table lists the 9 last-day dates still ahead, quietest first. Days on the same level are not ranked against each other. Each level is what we expect of that date, not a record of how many came.

• Jan 11MonComing of Age DayEasyroom↓
• May 5WedChildren's DayEasyroom↓
• Jan 10MonComing of Age DayEasyroom↓
• Oct 12MonSports DayFairly easyroom↓
• Oct 11MonSports DayFairly easyroom↓
• Nov 23MonLabour Thanksgiving DayNormalroom↓
• Jan 3SunNormalroom↓
• Feb 13SunNormalroom↓
• Mar 22MonSubstitute HolidayFairly busyroom↓

🏁 Only 10 days a year are last days!

The factor fires on 10 of the 724 days we score, one from each of 10 runs. That is 1% of the window.

They average Fairly easy against Normal for the year, 1 level apart. None of the year's 32 busiest days lands in here. Still inside a run, yet already below the yearly average.

💰 All 10 sit on the falling side for rooms!?

Split those 10 days by room rate: 10 on the falling side, 0 on the climbing side. Run by run, all 10 of them come out the same way. Not one last day lands on the expensive side.

Mid-run days invert this completely. Inside a single run, the room rate flips end to end.

🔍 So is the last day always quiet and cheap?

Not that far. 9 days (90%) came in below the yearly average, leaving 1 above it. The lightest is Easy, the heaviest Fairly busy, 3 steps apart.

And 10 days from 10 runs is a small pile to begin with. This is not "last days always behave this way" — it is what the 10 days inside this window did.

🗺️ It flips because tomorrow is a workday!

A last day has, by definition, a workday after it. Room rates read tomorrow rather than today, so last days land on the falling side automatically. Inside one run, the hotel treats these dates as a different animal.

The mechanism is in park crowds and room rates; which day of a run to pick is runs and school breaks; the comparison against the middle sits in mid-run days.

• othree / CC BY 2.0🗓️[Before you book] Tickets stop at two months, forecasts do not!Tickets stop two months out, Disney hotels four. Our forecast runs past both, so you can settle the date before anything goes on sale. Here is the order.
• othree / CC BY 2.0📍[June 15 plans] Only 2 June 15ths are in the forecast — what can they show?Only 2 June 15ths sit inside our forecast window: Jun 15, 2026 at Normal, Jun 15, 2027 at Fairly easy, on different weekdays. Too few days for an average.
• othree / CC BY 2.0🌙[Staying till close] 19 early closes, and most of them on Friday?Of the 696 entries we hold, 19 close before the usual 21:00, and 16 of those fall on Friday. The data has no field for the reason, so we do not give one.
• othree / CC BY 2.0🌆[Arriving late] Walk in at 15:00 and a short date costs you 50% of what is left?Opening is fixed at 09:00, so a standard day runs 720 minutes and the shortest 540. Those 180 minutes all come off the evening: 50% of a late arrival.
• othree / CC BY 2.0🧭[First trip] 463 days sit at or below the yearly average — how do you get to one?Of the 724 days we forecast, 463 sit at or below the yearly average. Decide in this order: crowd level, weekday, holiday runs, room rates.
• othree / CC BY 2.0📌[Only free on holidays] 35 holidays against 485 weekdays — how many steps apart?Across 35 public holidays the average is Normal; the 485 ordinary weekdays average Normal. The same step — and the bands still overlap.