Method
How we rate a day
Every day of the year, in each of 132 cities, gets a score from 0 to 100 and one word. This page says what the words mean, what goes into a score, how sure we are, and where every figure comes from.
The words
We say “not the time” only on firm grounds: when a day is below “fair” even allowing for our margin of error, or when something holds it there whatever else is good: dangerous heat, a rainy season, very unhealthy air, the worst of a storm season, severe cold, a closed season or government advice against travel. In the northern-lights season of a city where the lights are overhead on ordinary nights, a day without such a condition is “fair” at worst: the cold and dark are what people come for. A month is “Avoid” when at least half of its days are “not the time”; otherwise “Fair”, unless it is good or among the best.
What goes into a score
Five things, each rated 0 to 100 for the day, then weighed together. The weights are our choice, stated here; they are not a measurement. Weather counts most because it changes a day most.
Best for
Those weights are the All-round reading, the one every page opens on. “Best for” offers three more, which weigh the same five things differently:
One rule keeps them honest: on a day when less than a third of the daytime is comfortable, a focused reading never rates the day above All-round. A quiet or cheap day in poor weather is not a great day.
When one thing spoils the day
Cheap hotels do not make up for a hurricane. So some conditions hold a day’s score down whatever else is good: government advice against travel; dangerous heat or severe cold likely; storm season, by how likely a storm or hurricane is; unhealthy air on most days (at best “fair”; very unhealthy or hazardous air, “not the time”); a rainy season with frequent downpours; extreme crowds (except in the Lively reading); and a closed season, when less than half of the place’s hotel capacity is open. Close to a rainy season, or where the air is unhealthy on some days but not most, a day is lowered a little, by how likely it is, and says why.
How sure we are
Every score has a margin of error. We measure it: each input is moved by the error we have measured for it (for example, how far our weather figures sit from the city’s own station), and the day is scored again, 200 times. A typical day’s margin is ±3 points; it is wider where the data is weaker, up to about ±5 for one day in ten. That is for the All-round reading; Lower prices rests on hotel prices, which vary more, and its typical margin is about ±6.
It covers the error in our data, not our own choices such as the weights. Where the data is weaker, a day’s score also stays closer to the city’s own average instead of swinging on figures we are less sure of.
That margin decides what we say. One city or week is called better than another only when the gap is bigger than the margins. Otherwise we say it is too close to call.
What a score is not
It is not a forecast. A score says what is usual for those dates, from years of measured data, with that year’s holidays and events on the calendar. It cannot know about a heatwave, a strike or a storm that has not happened yet. Check the forecast before you travel.
Money
TripTiming carries no adverts and no booking links, and no one pays for a rating. If we add booking links later, every rating will still be worked out before and apart from them, and this page will say so.
The data
414 sources, all public and free to read but one: the hotel price checks on big event nights, which come from a hotel-price service whose terms are not public. Most are open data or official publications (holiday and school calendars, whose dates are facts), used under their terms; 120 are reports or services whose terms do not cover reuse or could not be settled, and from those we take only a pattern (how the seasons rise and fall, or how much an event raises prices), never their figures (each is marked in the full list). The widest-used are below; the full list follows. Where a city publishes no usable monthly series, we use its airport traffic (sometimes averaged with Airbnb review dates), or official monthly figures reported in reputable news (each one cited), and the margin of error is wider there. How much most events other than street festivals add to the crowds still comes from our earlier estimates of their size; we are replacing them with measured ones.
Corrections
If a figure looks wrong for a place you know, tell us through the contact form and we will check it against the source.