Why the Fibonacci sequence?
Story points climb 1, 2, 3, 5, 8, 13 because you can't tell a 13 from a 14. The gaps are the point.
The gaps grow because uncertainty grows. A scale that lets you say "7" is a scale that lets you pretend.
Fibonacci — 1, 2, 3, 5, 8, 13, 21 — spaces the numbers further apart as they get bigger. That spacing isn't decorative. It encodes the one thing every estimate is really about: how confident you are. Telling a 2 from a 3 is a real distinction your team can argue and resolve. Telling a 21 from a 22 is theatre. The sequence widens the gaps so the scale stops offering a precision it can't deliver.
Why not 1 to 10?
Because a linear scale promises distinctions that don't exist. Give a team ten buckets and someone will spend ten minutes arguing 6 versus 7 — a difference that evaporates the moment anyone opens the codebase. Fibonacci gives you roughly six usable buckets and forces the decision: this is a 5 or it's an 8. There's no 6 or 7 to hide in, so the conversation moves to what actually matters — what makes this bigger than the reference story.
Why are story points 1, 3, 5, 8?
Some decks drop a number or two; the classic planning poker deck is 1, 2, 3, 5, 8, 13, 21. The exact membership matters far less than the shape. What every variant shares is the widening gap — each step is roughly the sum of the two before it, so the jumps get bigger exactly where your ability to estimate gets worse. Pick a deck and keep it; don't let the team renegotiate the scale every sprint.
Modified Fibonacci, and Powers of 2
Most planning poker tools ship a modified Fibonacci deck — ½, 1, 2, 3, 5, 8, 13, 20, 40, 100 — that rounds the top end. Nobody distinguishes a 34 from a 21 at that size, so the deck rounds to 20, 40, 100 and stops pretending. Powers of 2 (1, 2, 4, 8, 16) work for the same reason: exponential spacing. Pick Fibonacci unless your team is already fluent in doubling — the difference is cosmetic, and switching decks mid-stream just resets everyone's calibration.
When the scale is the wrong fix
If every story comes back an 8 or a 13, the scale isn't the problem — the stories are too big. Split them; don't reach for a 34. And if you find yourself wanting a 6, you've started estimating in hours with extra steps. The missing 6 is a feature.
Use Fibonacci because uncertainty isn't linear. The numbers spread out because your confidence does the opposite.
Common questions
Why are story points 1, 2, 3, 5, 8?
Because the gaps mirror how uncertainty grows with size. The difference between a 2 and a 3 is real and arguable; the difference between a 21 and a 22 is noise. Fibonacci widens the gaps as the numbers grow so the scale stops offering precision it can't deliver.
Why use Fibonacci instead of a 1-to-10 scale?
A linear 1–10 scale invites arguments over 6 versus 7 — a distinction that won't survive contact with the work. Fibonacci gives you about six usable buckets and forces a decision: it's a 5 or an 8, with nothing in between to hide in.
What is the modified Fibonacci sequence?
A rounded version — ½, 1, 2, 3, 5, 8, 13, 20, 40, 100 — used by many planning poker decks. It rounds the large numbers because no team meaningfully distinguishes a 34 from a 21 up there. The rounding just admits it.
Why does planning poker use the Fibonacci sequence?
Planning poker decks use Fibonacci so the cards force a choice the moment estimates get fuzzy. Because the gaps widen as the numbers grow, there's no value to hide an each-way bet on — you commit to a 5 or an 8, and the discussion moves to why.
Adjacent: planning poker card decks for the mechanics of each deck; why no 1-point stories for the other end of the scale; story points vs hours for the conversion trap that the missing numbers protect against.