Fibonacci vs T-shirt sizing: which estimation scale should you use?
Fibonacci story points or T-shirt sizes? How each agile estimation scale works, when to use which, and how to convert T-shirt sizes to points for your tracker.
· planning poker / story points / estimation / agile
Every team that starts planning poker hits the same question in the first five minutes: which cards do we use? The two most common answers are the Fibonacci sequence and T-shirt sizes. Both work. They answer slightly different questions, and picking the wrong one for the job makes estimation feel pointless.
The short answer
- Use Fibonacci when you’re estimating stories for the next few sprints and you track velocity in story points.
- Use T-shirt sizes when you’re sizing a roadmap, epics or ideas that aren’t broken down yet.
- Many teams use both: T-shirt sizes early, Fibonacci once a story is ready for a sprint.
Why not just 1, 2, 3, 4, 5…?
Both scales exist for the same reason. Uncertainty grows with size. You can tell a 2 from a 3 because both are small and well understood. Nobody can honestly tell a 14 from a 15.
A linear scale invites fake precision and long debates over one point. A scale with growing gaps forces a coarser choice as the work gets bigger: is this an 8 or a 13? That’s a question a team can actually answer.
Fibonacci story points
The deck is usually 0, 1, 2, 3, 5, 8, 13, 21, plus ? for “I don’t understand it” and ☕ for “I need a break”. Each number is roughly the sum of the two before it, so each step is about 60% bigger than the last.
Points are relative, not hours. A 5 is “about as much effort, complexity and risk as the other 5s we’ve done”. Pick a well-understood reference story, call it a 3, and size everything against it.
Strengths
- Numbers add up, so you can measure velocity and forecast sprints.
- Fine-grained at the small end, where you have the most information.
- Widely known, so new team members recognise it immediately.
Weaknesses
- Numbers look like hours. Stakeholders start asking “so 8 points is one day?”
- Teams argue over 5 vs 8 when either would do.
- Anything above 13 is mostly a signal to split the story, not a real estimate.
Modified Fibonacci
A common variant is 0, ½, 1, 2, 3, 5, 8, 13, 20, 40, 100. The ½ is for trivial changes. The rounded top values (20, 40, 100) make it obvious that large numbers are rough guesses. It’s useful if one backlog mixes one-line fixes with whole features.
T-shirt sizes
The deck is XS, S, M, L, XL, XXL. There’s no arithmetic, just buckets.
Strengths
- Fast. People decide “medium” far quicker than “5 or 8”.
- Nobody confuses an L with hours, so conversations with stakeholders stay about relative size.
- Works for things that aren’t broken down yet: epics, roadmap items, ideas for next quarter.
Weaknesses
- You can’t add up sizes, so there’s no velocity without converting.
- The buckets are coarse. Two “M” stories can differ by a factor of two.
- Teams interpret sizes differently unless they agree on reference items.
Side by side
| Fibonacci | T-shirt sizes | |
|---|---|---|
| Cards | 0, 1, 2, 3, 5, 8, 13, 21 | XS, S, M, L, XL, XXL |
| Best for | Sprint-ready stories | Epics, roadmaps, early ideas |
| Velocity and forecasting | Yes, points add up | Only after converting |
| Speed of a round | Moderate | Fast |
| Risk of confusing with hours | Higher | Low |
| Precision | Finer at the small end | Coarse buckets |
Converting T-shirt sizes to story points
If you size with T-shirts but your tracker wants numbers, agree on a mapping. A common one follows Fibonacci:
| Size | Points |
|---|---|
| XS | 1 |
| S | 2 |
| M | 3 |
| L | 5 |
| XL | 8 |
| XXL | 13 |
Treat the result as a rough forecast, not a commitment. An XXL usually means “split this before anyone builds it”.
Planning Poker uses this mapping by default when you estimate with the T-shirt deck, so the totals and the Jira CSV export still get numeric story points.
How to choose for your team
Ask two questions:
- Do you need to add estimates up? If you plan sprints by velocity, use Fibonacci for sprint work.
- How well do you understand the work? If the items aren’t broken down yet, T-shirt sizes fit the level of detail you actually have.
And whichever you pick, keep the rules that make planning poker work: vote in secret, reveal together, and let the people with the highest and lowest cards explain first. The remote facilitation guide covers how to run those rounds on a video call.
Try both in one session
You don’t have to commit up front. In Planning Poker, the facilitator can switch the deck at any time: Fibonacci, Modified Fibonacci, T-shirt sizes, or your own list of cards. Run one round of each on the same story and see which one gets your team to a decision faster.
Start a free room. No sign-up, and nothing to install.