Types of Estimating: Time vs. Complexity vs. Multidimensional

When agile teams need to estimate the effort required to complete a job (such as a user story), there are several possible approaches.

The most common ones can be grouped into three main categories: time-based estimates, complexity-based estimates (relative size), and multidimensional estimates (combining several factors).

Each approach has its own characteristics, advantages and disadvantages.

The choice of estimation method can significantly influence team dynamics, planning accuracy and the ability to use historical data to predict future performance.

Time Based Estimating (Man Hours/Days) and its Disadvantages

This is the more traditional approach, where work is estimated in units of absolute time, such as man-hours or man-days. For example, a story is estimated to require "five man-days".

Although it seems straightforward, this method has several significant disadvantages in agile contexts:

• It does not scale linearly: Five man-days does not mean five people will finish it in one day; adding more people often increases complexity and may even lengthen the time.
• Unhelpful Historical Data: Knowing that the team completed 180 hours last week and 220 this week does not clearly indicate whether they improved, worked more hours, or simply had different tasks. It makes future prediction difficult.
• Tied to the Person: The estimate depends on who performs the task (a senior will take less time than a junior), which requires knowing the assignment beforehand or generates inaccuracies.
• Creates Pressure: Can create an unhealthy dynamic where team members feel pressured to meet time estimates, especially the less experienced ones, or compete to give the lowest estimate.

While useful in classic project management, it is often avoided for day-to-day agile planning.

Complexity-Based Estimating (T-Shirt Sizing)

This approach estimates work using relative size categories, such as T-shirt sizes (S, M, L, XL, etc.).

It focuses on the complexity or perceived size of the job itself, decoupling it from the specific time or person who will perform it.

It is a step toward relative estimation, overcoming some disadvantages of time-based estimation.

However, its main limitation is similar: the difficulty of using historical data in an actionable way.

Knowing that the team completed "2M, 1L and 5S" in one iteration does not make it easy to predict how many "L's" or "S's" they can complete in the next.

Multidimensional Estimation (Story Points)

Story Points are the most common example of multidimensional estimation.

This approach seeks to capture a relative measure of effort by considering multiple factors simultaneously: approximate time, technical or domain complexity, and associated uncertainty or risk.

Like T-shirt sizes, they estimate the work itself, unlinked to the specific person.

Their great advantage is that, being numerical