May 19, 2026
6 min read
Estimation and planning poker
How to Calibrate Story Points for a New Team
A practical guide to calibrating story points for a new team so estimates become more consistent without pretending everyone needs perfect agreement from day one.
Start with a shared baseline, not a perfect scale
A new team usually struggles with story points because people bring different past experiences, different technical instincts, and different expectations about what the numbers should mean.
The goal is not to get everyone identical instantly. The goal is to build a shared reference point the team can use consistently enough to have better planning conversations.
Story point calibration
New teams calibrate faster when they build a shared sizing language from real examples.
Shared baseline
Calibration starts by agreeing what small, medium, and large roughly feel like.
Use sample stories
Compare new items against work the team already understands.
Compare relatively
Keep the conversation about relative size, not hidden hour guesses.
Refine over time
The team's sizing language gets more stable through repeated use.
Working calibration
The point is consistency good enough for planning, not a mathematically perfect system.
Use a few real examples from work the team understands
Calibration works better when the team compares actual backlog items or recently completed work instead of abstract fictional examples.
Real examples make it easier to discuss what drove the size: effort, complexity, uncertainty, dependencies, or hidden edge cases.
Pick one or two anchor stories
A helpful move is to agree on one or two anchor stories that the team can use as reference points. For example, if one story clearly feels like a 3, later stories can be compared against it instead of estimated in isolation.
Anchors work best when the team understands them well enough that they feel stable, not controversial.
Talk about why the story feels that size
Calibration improves when the discussion goes beyond the number itself. The useful part is the reasoning behind the number.
If people understand why a story felt like a 5 rather than a 3, the team starts building a shared estimation language instead of just memorizing card values.
- How much effort is involved?
- How much technical complexity is present?
- How much uncertainty is still unresolved?
- What dependencies or risks make the work feel larger?
Expect disagreement early
Wide estimate spreads are normal when a team is still calibrating. They are often a sign that the team is learning how each person sees the work, not that the process is failing.
The mistake is expecting calibration to remove disagreement immediately instead of using disagreement to build shared understanding over time.
Revisit the baseline after real delivery
Calibration gets stronger when the team occasionally looks back at completed work and asks whether the original sizing logic still feels right.
That reflection is not about punishing the estimate. It is about improving the team's shared intuition using actual delivery experience.
Do not over-engineer the scale
New teams sometimes make calibration harder by turning it into a heavy methodology exercise. The team does not need a perfect theoretical model before it can estimate usefully.
A few good anchors, clear discussion, and repeated practice usually do more than an elaborate rulebook.
Where to go next
If your team wants a cleaner way to build shared estimation language instead of trading isolated gut feelings, the estimator is the best next step.
That is where the team can structure the reasoning behind story points and make calibration feel more grounded without turning it into heavyweight process.
TL;DR
- New teams calibrate faster when they use shared anchor stories.
- The useful part is the reasoning behind the estimate, not the number alone.
- Disagreement early on is normal and often helpful.
- Calibration improves when the team builds a shared sizing language from examples instead of chasing a perfect scale immediately.