Home / Planning poker

Online planning poker, free and without an account

Estimate user stories together, remotely or in the same room, without installing anything and without anyone signing up.

What is planning poker?

Planning poker is a consensus-based estimation technique used by agile teams. Each person secretly picks a card that represents their estimate, then everyone reveals at the same time.

The simultaneous reveal is the whole point: nobody is influenced by the first number said aloud, whether it comes from the most senior person or the loudest voice. It avoids anchoring and brings out real differences of opinion.

How a round works

1. Present the item. The Product Owner explains the story and the team asks questions.
2. Pick a card in private. Everyone chooses a value without showing it.
3. Reveal together. All cards are turned over at the same moment.
4. Explain the gaps. If estimates differ, the lowest and the highest voters explain their reasoning.
5. Vote again until the estimates converge. Most items settle after one or two rounds.

Fibonacci or T-shirt sizes?

Fibonacci (1, 2, 3, 5, 8, 13…) is the classic deck. The gaps widen as the numbers grow, which reflects a simple truth: the bigger an item, the less precisely you can estimate it.

T-shirt sizes (from small to large) suit early sizing, such as roadmap items or epics, where a number of points would suggest false precision.

Scrum Island offers both decks, so you can pick the one that fits the conversation.

Planning poker at the Tavern on Scrum Island

On Scrum Island, planning poker happens at the Tavern, a place on your team's island.

• The Scrum Master gathers the team at the Tavern and starts a round, with an optional name for the item.
• You choose the deck (Fibonacci or T-shirt) and, if you like, a round timer.
• Votes stay hidden until the cards are revealed, and you can change yours until then.
• The result shows a verdict (consensus, close or significant spread), the average and the distribution of votes.
• When the votes are far apart, the app points at the lowest and the highest voters and asks them to explain. Then you revote.

It runs in real time, in your browser, with no account. Available in English, French and Dutch.

Tips for better estimates

• Estimate effort relative to other items, not hours. Compare with a story the team already knows.
• Treat a spread as information. A wide gap usually reveals a different understanding of the item, which is better discovered now than mid-sprint.
• Keep rounds short. A timer helps to avoid long debates about a single story.
• Split the big ones. An item that stays at the top of the scale after discussion is often too large to start.