> ## Documentation Index
> Fetch the complete documentation index at: https://docs.rolearn.dev/llms.txt
> Use this file to discover all available pages before exploring further.

# Reading Forecasts

> How to interpret RoLearn's Roblox CCU forecasts — understand prediction charts, confidence intervals, and forecast accuracy

Forecasts predict future CCU for your tracked games using advanced time-series machine learning.

## Understanding the Forecast Chart

The forecast chart shows three lines:

* **Predicted CCU** (solid blue): The model's best estimate
* **Upper bound** (dashed, lighter): Optimistic scenario
* **Lower bound** (dashed, lighter): Conservative scenario

The wider the confidence interval, the more uncertain the prediction.

## What Affects Forecast Accuracy

<AccordionGroup>
  <Accordion title="More history = better predictions">
    Games with 30+ days of CCU history produce more reliable forecasts than recently tracked games.
  </Accordion>

  <Accordion title="Seasonal patterns matter">
    The model automatically detects weekly patterns (weekday vs. weekend) and seasonal trends (holiday spikes).
  </Accordion>

  <Accordion title="Updates can break patterns">
    A major game update can shift CCU dramatically. Forecasts based on pre-update data may not reflect the new reality until enough post-update data accumulates.
  </Accordion>
</AccordionGroup>

## Forecast Horizons

| Plan     |         Forecast Range         |
| -------- | :----------------------------: |
| Explorer |                —               |
| Builder  |             7 days             |
| Studio   | 30 days + confidence intervals |

## Tips

* **Compare forecast vs. actual**: Check back after a few days to see how the prediction held up
* **Watch the trend, not the number**: The direction matters more than the exact CCU value
* **Use with benchmarking**: Combine forecasts with [genre benchmarks](/features/game-benchmarking) to understand if growth is above or below genre norms