METHODOLOGY
How the Pulse Score Works
Every event on NextWeekends carries a Pulse Score from 0 to 100. It exists to answer one question — is this worth your time? — and this page documents exactly how it is calculated, including what it deliberately ignores.
This is not AI, and not machine learning. The Pulse Score is a deterministic weighted model: fixed factors, fixed weights, no training data and no model that changes underneath you. The same event, the same search, and the same forecast will always produce the same score — which is precisely why it can be published in full.
The 100-point budget
5 factors, summing to 100 when all of them apply:
| Factor | Points | What it asks |
|---|---|---|
| Event significance | 30 | Is this worth showing up for? |
| Date proximity | 20 | How soon is it? |
| Budget fit | 20 | Does it fit what you said you'd spend? |
| Genre / type match | 15 | How literally does it match what you asked for? |
| Weather | 15 | What's the forecast for that specific date? |
Event significance — 30 points
Starts every event at a baseline of 30% of the available points, then adds: 25% for a festival, 10% per listed performer up to three of them, and 30% for a major-league sporting event. The total is capped, so an event cannot exceed the full 30 points.
Significance never looks at price. A large share of ticketing inventory has genuinely unknown pricing, so deriving significance from price would rank events by how well we happen to have priced them rather than by what they are — and would make an event's rank shift when its pricing data merely went stale.
An honest limitation: community-feed events (library programs, tourism-board listings) usually list no performers, so they tend to score lower here. That reflects what those feeds contain, not a rule that penalizes small events — a festival-tagged community event with multiple performers scores identically to a ticketed one with the same attributes.
Date proximity — 20 points
Scored across a 14-day window. Something happening today earns the full 20points; the score declines linearly to 0 at 14 days out and beyond. Events already in the past score 0. If a date can't be parsed at all, it receives half credit rather than being dropped to zero on a data problem that isn't the event's fault.
Budget fit — 20 points
Budget is treated as a constraint you set, never as a target to spend. The curve is deliberately flat at the top:
- At or under 60% of budget: full 20 points, equally, no matter how much cheaper one option is than another.
- Between 60% and 100%: a gradual decline down to 10 points at exactly your budget.
- Over budget: 0 points.
The plateau exists for a specific reason. A continuous "cheaper is better" reward would quietly turn lowest price into a ranking goal of its own, and a free event would always outrank a great one. Here, a free event is fully budget-compatible but earns exactly what any comfortably affordable option earns — free stops being penalized without becoming automatically best. The 60% threshold rather than 100% reserves headroom for the costs an estimate doesn't capture, like food and gas.
Genre and type match — 15 points
Passing a filter isn't the same as being a good match, so this factor grades how literally an event matches what you asked for: a direct genre match earns the full 15 points, a match through the broader category taxonomy earns 11, and a match that exists only via tags earns 6.
When you haven't asked for a particular genre or type, this factor awards full credit rather than penalizing every event for a preference you never expressed.
Weather — 15 points
Scored from the forecast for the event's actual date, not today's conditions. Chance of rain carries roughly two-thirds of the weight and temperature comfort the remaining third, with the comfort band centered between 65°F and 80°F and tapering outward from there.
When no forecast is available — an event too far out to forecast, or enrichment that hasn't run yet — the factor awards full credit. Missing data is never treated as bad news.
What the score deliberately ignores
Distance is not a scoring factor
This surprises people, so it's worth being explicit: how far away an event is has no effect on its Pulse Score.
The search radius you choose is already a hard filter — anything outside it is excluded before scoring begins. Every event that reaches the scoring stage has, by definition, already satisfied the distance constraint you set. Scoring it again for being 45 minutes away instead of 5 would penalize you twice for a preference you already expressed once.
Distance is still used to filter by radius, to display drive time, and to break ties between events with identical scores.
Unknown prices are not guessed at
When you've set a budget but an event's price is genuinely unknown, the budget factor is removed from that event's score entirely and the remaining 80 points are scaled back up to 100.
It is not treated as free (which would be an unearned advantage), not treated as over budget (which would bury it), and not given an arbitrary partial credit. It competes on everything actually known about it. A combined trip-cost estimate is only trusted when it's complete — a partial estimate is treated as unknown rather than presented as fact.
No paid placement inside the score
The five factors above are the complete set of inputs. Whether a venue is a partner is not among them, and there is no way to buy a higher Pulse Score.
Separately from scoring, partner events can appear in featured placements elsewhere on the site. Those are labeled as partnerships where they appear — a distinct thing from the score, and disclosed as such.
Score tiers
The badge shown on a card is derived from the score:
| Score | Label |
|---|---|
| 95 – 100 | Pulse Perfect Match |
| 90 – 94 | Highly Recommended |
| 80 – 89 | Great Match |
| Below 80 | No tier label shown |
Where our events come from
Event discovery is nationwide: search any city in the country and NextWeekends draws on national ticketing feeds to show what's happening there.
On top of that, we index community sources that never reach a ticketing site — library programs, tourism-board calendars, and venue listings. That deeper layer is live in Evansville, Indiana today, and it's the model we're expanding market by market. We'd rather say exactly where that stands than imply it's everywhere already.