Planning poker

Have you asked the team how they think they should estimate how much work they can take on in a Sprint? Things they might wish to bear in mind include:

• Time based estimation is rarely accurate except at a small scale, which is why relative estimates like story points are preferred.
• A team can relatively estimate using T-Shirt sizes, and then map to points using a scheme of their choosing, such as XS:1, S:2, M:3, L:5, XL:8, XXL:13
• During Sprint Planning a team should always look at a forecast of work in aggregate and decide if the Sprint Goal is realistic and achievable.

Equating points to time can add unnecessary complexity, and can lead to some dangerous assumptions when using those estimates later.

One technique is for the team to take a typical-sized, and well understood item from the backlog, and declare that estimate as a certain number of points (e.g. 2 points).

The team could then use this item as a baseline for relative estimates, until there is a good understanding of the points rating within the team. So an item that is approximately half the size of the baseline would probably be estimated at 1 point.

If there was a need to equate story points back to time you'd surely just use time from the beginning and skip points all together.  What extra value can story points possibly bring if x points = x time?

Thank you for your answers.

So, lets say we go for points and we estimate all the stories.  On the planning we take 25 points to be completed on the sprint.  Since its our first sprint ever, lets say we took too much and we completed 10 points, then for the next sprint that is the base i should take and i should take around 10 points.   And keep trying to improve sprint after sprint.

Am i correct?

That would be one possible way of doing it. Take an empirical measurement and go from there. I'd say you're on the right track.