Planning poker vs
T-shirt sizing

Two of the most popular agile estimation techniques go head to head. Learn the strengths of each so you can pick the right method for your team.

6 min read

Two popular estimation methods

Agile teams rely on relative estimation to size their work without getting bogged down in hours or days. Two of the most widely adopted techniques are planning poker and T-shirt sizing. Both compare stories against each other rather than assigning absolute time values, but they differ in precision, speed, and the depth of discussion they encourage.

Planning poker uses a numbered scale, most commonly the Fibonacci sequence, and a structured reveal-and-discuss process. T-shirt sizing uses labels like XS, S, M, L, and XL for a faster, more intuitive sort. Choosing between them depends on where your team is in the planning cycle and how granular you need your estimates to be.

Below we break down how each method works, their strengths and weaknesses, and when to reach for one over the other.

How planning poker works

Each team member holds a set of cards with values from a numbered scale (e.g., 0, 1, 2, 3, 5, 8, 13). A story is presented, everyone selects a card simultaneously, and the cards are revealed at once. If estimates differ, the highest and lowest voters explain their reasoning, the team discusses, and another round is played until consensus is reached.

Pros

  • Precise estimates with numerical values
  • Reduces anchoring bias with simultaneous reveal
  • Encourages in-depth discussion
  • Works well for sprint-level items

Cons

  • Takes more time per item
  • Requires all team members present
  • Can feel slow for large backlogs

How T-shirt sizing works

Stories are placed into buckets labeled XS, S, M, L, and XL based on relative size. The team reviews each item and quickly agrees on a bucket. There is no formal reveal step; most teams simply discuss and place. The focus is on speed and rough categorization rather than precise numerical values.

Pros

  • Very fast per item
  • Easy to learn for any team member
  • Great for roadmap-level estimates
  • Low overhead and minimal ceremony

Cons

  • Less precise than numbered scales
  • Harder to track velocity over time
  • Limited discussion compared to poker
  • Ambiguous boundaries between sizes

Side-by-side comparison

CriteriaPlanning pokerT-shirt sizing
PrecisionHigh (numerical)Low (categorical)
SpeedModerateFast
Discussion depthDeepLight
Best forSprint planningRoadmap planning
Scale typeFibonacci / numbersXS - XL labels
Learning curveModerateLow
Remote-friendlyYes, with toolsYes

When to use each method

Use planning poker when…

  • Sprint planning and backlog refinement
  • When precision matters for capacity planning
  • Small-to-medium backlogs (under 30 items)
  • Teams that value deep estimation discussions

Use T-shirt sizing when…

  • Roadmap and quarterly planning
  • High-level estimation for stakeholders
  • Large backlogs that need quick sizing
  • New teams learning relative estimation

Frequently asked questions

Yes. Some teams use T-shirt sizing for initial roadmap estimation and then switch to planning poker with Fibonacci values for sprint-level refinement. This hybrid approach gives quick high-level estimates while maintaining precision where it matters.

T-shirt sizing is generally faster per item because the scale has fewer options and the discussion is lighter. Planning poker takes more time per story but produces more precise and well-discussed estimates.

Both work well remotely with the right tools. Planning poker is particularly well-suited because online tools like Scrum Poker handle the simultaneous reveal and vote tracking automatically.

Start with T-shirt sizing. The simple scale is easier to learn and less intimidating for teams new to relative estimation. Once the team is comfortable, consider switching to planning poker for more granular estimates.

Try both methods free

Scrum Poker supports Fibonacci, T-shirt sizes, and custom scales. Start a session in seconds with your team.