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

# Bounded shapes

> Triangle, rounded peak (PERT) and even spread: distributions for hard limits.

Use a bounded shape when you can name a **lowest possible** and a **highest possible** value. Nothing outside those limits can occur in the model.

## The three shapes

| Shape | Looks like | Good for |
| - | - | - |
| **Triangle** | A straight rise to the most common value, then a straight fall | Lowest, most likely and highest, with no other information. Gives the ends a fair amount of weight |
| **Rounded peak (PERT)** | A smooth hump around the most common value, with thinner ends than the triangle | A most common value you are fairly sure of |
| **Even spread** | Flat: every value between the limits is equally likely | When you have no idea which value is most common. The most common value is not used |

The triangle is chosen automatically for hard limits. PERT and even spread are never chosen automatically; select them yourself.

## Choosing between them

1. Do you truly know the most common value? If not, use **even spread**.
2. Are you fairly sure of it, and do extreme values seem unlikely? Use **PERT**.
3. Otherwise keep the **triangle**.

> [!TIP]
>
> The shape chooser shows each curve, its average and its range side by side. Compare them before choosing.

## When not to use them

If your low and high are a *likely range* rather than true limits, use a [skewed or percentage shape](https://docs.meerietechnology.com/estimate/skewed-shapes.md) instead. Real limits are rarer than people think: a cost can almost always exceed your highest guess.

**Next:** [Skewed and percentage shapes](https://docs.meerietechnology.com/estimate/skewed-shapes.md)


This documentation is built and hosted on [Mintlify](https://mintlify.com), a developer documentation platform.