Attendance Forecasting

Attendance forecasts from an
8,000-agent crowd simulation

Most attendance forecasts are built on ticket sales. Thoosie is built on 48M+ measured wait times across 56 US parks — the actual behavior of crowds, not the purchase intent that preceded them. Day-ahead precision for operations. Month-ahead visibility for marketing and finance.

48M+
Real wait-time readings
8K
Agents in crowd simulation
56
US parks tracked
Day-ahead
Operational forecast horizon

What drives forecast accuracy

Ticket sales tell you who paid. Wait times tell you who showed up, when, and where they went. That distinction is the source of Thoosie's forecast edge.

Primary Signal

Historical Wait-Time Data

48M+ readings across 56 parks, updated every 15 minutes for years. This is measured crowd behavior at the attraction level — the richest signal available for predicting what crowds do, not just how many show up.

Calendar Signals

School Calendars & Events

Local and regional school calendars are ingested automatically. Park-specific events, holiday schedules, and regional competitors are layered in — capturing the demand spikes that trip up models built on ticket sales alone.

Environmental

Weather Adjustment

Forecast models are conditioned on weather — not just rain/no-rain, but temperature, cloud cover, and hour-by-hour deviation from seasonal norms. Hot days at water parks and cool days at coaster parks behave very differently.

Simulation Core

8,000-Agent Crowd Simulation

The Kings Dominion simulation model runs 8,000 individual agents through a full park day — each agent with their own decision rules, tolerance thresholds, and attraction preferences. The result is a crowd distribution forecast, not just a gate count.

Forecast horizons

Different decisions need different time windows. Thoosie produces forecasts at three horizons calibrated to how parks actually plan.

1 Day Ahead

Operational Staffing

The highest-precision forecast. Produced each evening for the next calendar day. Used by operations to finalize staffing levels, ride opening schedules, and position assignments before gates open.

1 Week Ahead

Workforce Scheduling

Week-ahead crowd bands give scheduling teams enough lead time to adjust shift templates before they're published. Reduces last-minute callouts, overtime, and the scramble caused by surprise peak days.

1 Month Ahead

Marketing & Capacity

Longer-horizon forecasts identify slow stretches worth targeting with promotions and busy periods where capacity constraints may limit revenue. Gives marketing and finance teams a shared ground truth for planning.

Who uses attendance forecasts

Attendance forecasting isn't one team's problem — it's the foundation that operations, marketing, and finance all build on.

Operations

Day-ahead and intraday forecasts drive staffing levels, ride cycling schedules, and real-time crowd management decisions. The duty manager sees tomorrow before it arrives.

Marketing

Month-ahead forecasts identify low-demand windows where promotions can move the needle — and high-demand periods where discounting just cannibalizes full-price visits.

Finance & Revenue

Reliable attendance forecasts anchor revenue projections. Finance teams can model variable-cost exposure and flag weeks where attendance risk is elevated before it hits the P&L.

Frequently Asked Questions

How accurate are theme park attendance forecasts?+

Day-ahead forecasts from pattern-based models typically achieve 85–92% accuracy at well-established parks with 3+ seasons of historical data. Thoosie's 8,000-agent simulation is validated against 48M+ historical wait-time readings.

What methods do theme parks use to forecast attendance?+

Time-series regression on historical gate counts, advance ticket sale velocity models, weather correlation models, agent-based crowd simulations, and ML ensemble models. Thoosie combines agent-based simulation with live wait-time signal data.

What is an agent-based crowd simulation?+

An agent-based simulation models individual guest behaviors — arrival patterns, ride preferences, break timing, exit triggers — across thousands of synthetic guests simultaneously. Thoosie runs 8,000 agents, producing ride-level demand curves rather than just gate count forecasts.

See a forecast for your park

We'll pull a day-ahead attendance forecast for your property — or a comparable park in our dataset — and walk through what drives it.

Request Access