Difference between affinity estimation vs relative sizing

Affinity estimation can leverage many types of relative scales, including T shirt sizes (e.g. XS, S, M, L, XL, XXL), story points based on the Fibonacci sequence scale (e.g. 1, 2, 3, 5, 8, 13, 21...), or other scales (e.g. dog sizes, gummy bears).

Affinity estimation is a quick way to visualize your Product Backlog into groupings of relative sizes. The Dev Team typically places Product Backlog items in the right relative group. This approach is usually much quicker and easier than trying to estimate your Product Backlog with Planning Poker. The downside is it is less accurate compared to Poker.

Another common approach to relative estimation is to use Planning Poker (based on the Delphi Method). Planning Poker focuses on estimating one Product Backlog item at a time, and is consensus drive. The same relative scales used with Affinity estimation can be used. While Planning Poker takes longer to estimate a large batch of Product Backlog items, it is considered more accurate than Affinity estimation.

Whatever approach you choose to estimate, consider the conversations more valuable than the estimates themselves.

Hope this helps!

Chris

There is no real difference, there are numerous ways to relatively estimate complexity and workload. The one technique might prove to be more accurate and helpfull than others. 

In my teams right now, we use Tshirt sizes (S, M, L, XL) in early stages, and close to planning we use storypoints. This is because close to planning, more details are known due to refinement, and a more accurate value can be assigned.