How to use story points in planning poker
Story points give a team a shared way to compare work items. They are relative estimates: a card describes how a piece of work feels beside other work the same team has discussed, considering complexity, uncertainty, and effort.
Choose a deck your team can use consistently
Refinement Pointer offers Fibonacci cards: 1, 2, 3, 5, 8, and 13. It also offers T-shirt sizes: XS, S, M, L, and XL. Both decks include Uncertain. Pick one scale for the Session so everyone sees the same choices. A later Session in the same Room can use a different scale.
The spacing in either deck is useful because it leaves room for uncertainty. A team does not need a card for every fine distinction. When two items are hard to separate, the discussion may be more valuable than a more precise label.
Compare work, not people or calendars
A point is not an hour, a delivery date, or a score for a person. Avoid setting a conversion such as "one point equals a day." That invites false precision and turns a team estimate into a productivity comparison. Instead, use a familiar completed item as a reference and ask how the new item compares.
Fictional example
A team remembers that adding a simple preference to an existing screen felt like a 2. A proposed change touches a new integration, has unclear error behavior, and needs several screens to agree. Some people choose 5 and others choose 8. The difference highlights uncertainty and cross-system complexity, not who is expected to work faster.
Consider three questions before choosing a card:
- How much problem-solving or technical complexity is involved?
- What is uncertain or still needs to be learned?
- How much coordinated effort is likely across design, development, testing, or review?
Use private votes to surface assumptions
After the item is clear enough to discuss, have each participant select a card before anyone announces a number. Refinement Pointer keeps votes private until the team reveals them. That makes it easier to notice independent views rather than following the first confident estimate.
When cards differ, ask one or two people at the edges what they saw. The group may discover a missing dependency, a simpler approach, or a reason the work should be split. Revote when the new information changes how the item compares to the reference work. If uncertainty remains substantial, Uncertain can be an honest result and a prompt to learn more before committing to a relative estimate.
Record the decision where your team already plans work
Once the team has reached a useful estimate, record it in the planning system you already use. Refinement Pointer is a focused conversation tool: it has no accounts, ticket syncing, or export workflow. Its Room controls support vote, reveal, and reset. Ending a Session clears its current votes while preserving the Room link and recorded history.
Create a room and choose a deck
For a simple meeting flow, read how to run backlog refinement with planning poker.