Fibonacci sequence vs linear: why the scale matters
The Fibonacci sequence (0.5, 1, 2, 3, 5, 8, 13, 21) is the most used scale in Planning Poker. But why not use a linear scale (1, 2, 3, 4, 5, 6, 7, 8)? The answer lies in uncertainty.
The Weber-Fechner law
In psychology, the Weber-Fechner law states that human perception of difference between stimuli is logarithmic, not linear. The difference between 1kg and 2kg is easy to perceive. The difference between 100kg and 101kg, not so much.
With software estimates it’s the same. Differentiating between 1 and 2 complexity points is feasible. Differentiating between 12 and 13? Almost impossible. Uncertainty grows along with the task.
Why Fibonacci works
Growing gaps reflect growing uncertainty
- Between 1 and 2: gap of 100% — clear difference
- Between 5 and 8: gap of 60% — perceptible difference
- Between 13 and 21: gap of 61% — still perceptible, but with margin
The gaps force the team to make binary decisions: “Is this 8 or 13?” — no misleading middle ground.
Without linear scale, false precision
With a linear scale (1-10), the team wastes time debating whether something is 6 or 7 — a difference nobody can truly justify. It’s false precision that wastes time.
Fibonacci is naturally limited
The sequence stops at 21 (or 34, at most). This is a signal: if something is 21+, it’s probably too large for a single story and needs to be decomposed.
Popular alternatives
T-Shirt Sizing (XS, S, M, L, XL)
Less precise, but more intuitive. Good for backlog triage, bad for velocity calculation.
Powers of 2 (1, 2, 4, 8, 16)
Similar to Fibonacci in terms of growing gaps. Some teams prefer it for being simpler to remember.
Modified Fibonacci (0.5, 1, 2, 3, 5, 8, 13, 20, 40, 100)
The classic Planning Poker version. The larger jumps (20→40→100) further reinforce that giant tasks need to be broken down.
When each scale works best
| Scale | When to use |
|---|---|
| Fibonacci | Sprint planning, teams with some experience |
| Linear | Never recommended for real estimates |
| T-Shirt | Initial backlog triage, new teams |
| Powers of 2 | Teams wanting simplicity with large gaps |
Conclusion
The Fibonacci sequence isn’t arbitrary — it reflects how humans perceive complexity and uncertainty. Using the right scale forces good decisions: decomposing large tasks, accepting uncertainty, and avoiding false precision.