Why Planning Poker Uses Fibonacci Numbers (And When to Use Other Scales)

What Is the Fibonacci Sequence?

The Fibonacci sequence is a series of numbers where each value is the sum of the two preceding it: 0, 1, 1, 2, 3, 5, 8, 13, 21, 34, 55, 89 and so on, extending to infinity. It was described by the Italian mathematician Leonardo of Pisa (nicknamed Fibonacci) in his 1202 book Liber Abaci, though the pattern had been documented centuries earlier in Indian mathematics.

The defining characteristic of the sequence is that the ratio between consecutive terms converges on approximately 1.618 — the golden ratio — as the numbers grow. This ratio appears throughout nature in spiral patterns, plant growth, and animal proportions. In planning poker, however, what matters is not the golden ratio but the growing gap between values. As you move up the scale, the distance between adjacent points increases, making it progressively harder to distinguish between consecutive values.

Why Fibonacci Works for Estimation

The core insight is that human estimation accuracy degrades as scope increases. Estimating whether a task takes one hour or two hours is reasonably tractable — you can reason about the specific steps involved. Estimating whether a large feature takes three weeks or four weeks is far less reliable, because at that scale there are too many unknowns, dependencies, and emergent complexities to reason about precisely.

The Modified Planning Poker Fibonacci Scale

In practice, planning poker uses a modified Fibonacci sequence rather than the pure mathematical one. The standard planning poker deck looks like this: 1, 2, 3, 5, 8, 13, 21, 34, ?, ∞, ☕

Notice a few things. The sequence starts at 1 rather than 0, because a story of zero effort should not be in the backlog. The 34 card is included as a signal that a story is very large but still theoretically estimable. In many teams, a 34-point story is an automatic trigger to break the story down before the next sprint.

The ? card means "I don't have enough information to estimate this story." When multiple team members play the ? card, it is a strong signal that the story needs more refinement before estimation is useful. Playing ? is not a failure — it is the correct response to an under-specified story.

The infinity card means "this story is too large and too uncertain to estimate as a single unit." It is not a number but a request to split. If the whole team plays infinity, the story should go back to the product owner for decomposition.

The coffee cup card is a request for a break. Estimation sessions that run too long without breaks produce declining estimate quality. The coffee cup acknowledges that humans are not estimation machines and that a five-minute break will improve the quality of subsequent estimates.

When to Use T-Shirt Sizes Instead

T-shirt sizes — XS, S, M, L, XL, and sometimes XXL — are an excellent alternative to Fibonacci for specific contexts. They strip away any lingering numerical intuition and make it nearly impossible for anyone to convert sizes into hours, which is their primary advantage.

T-shirt sizing works best during early discovery and roadmap planning, when the team is trying to size epics or themes rather than individual user stories. At that level of abstraction, the precision of a Fibonacci scale is not just unnecessary — it is actively misleading. Saying an epic is "L" is more honest than saying it is "21 story points" when neither estimate will survive contact with actual implementation.

T-shirt sizes are also valuable when non-technical stakeholders are present. Business stakeholders often do not have an intuitive sense of how story points map to effort, but everyone understands what "large" and "small" mean in relative terms. Using t-shirt sizes in mixed-audience sessions reduces the translation overhead and keeps conversation focused on priorities rather than point totals.

Powers of 2

The powers-of-2 scale (1, 2, 4, 8, 16, 32) provides even wider gaps between values than Fibonacci and is favoured by teams doing coarse-grained capacity planning across quarters or program increments. Because each step doubles the previous value, the scale forces extremely honest acknowledgement of uncertainty at the high end.

When to Use Hour-Based Estimation

Hour-based estimation abandons relative sizing entirely and estimates in absolute time units. This is appropriate in some contexts — particularly when billing clients or fixed-price contracts. The significant downside is that hours correlate estimates to individual capacity rather than story complexity.

Which Scale Should Your Team Use?

For most agile teams running sprint-level planning, the modified Fibonacci scale is the right default. It is widely understood and its growing gaps keep estimation honest at every size level.

Use t-shirt sizes when you are working at the epic or theme level, or when non-technical stakeholders are estimating.

Use powers of 2 if your team tends to inflate estimates and you want structural pressure to decompose stories more aggressively.

Use hours only when external constraints require absolute time tracking.

Whatever scale you choose, consistency matters more than perfection. A team that uses the same scale for six sprints and builds a velocity baseline will produce more reliable forecasts than a team that switches scales frequently.

Try Different Scales with Your Team

Corgi Planning Poker supports Fibonacci, T-shirt sizes, powers of 2, and custom decks. You can experiment with different scales in different sessions to find what works best for your team's context.