Loading ...
Sorry, an error occurred while loading the content.

Re: [XP] A different scheme for estimating

Expand Messages
  • Kim Gräsman
    Hi James, ... In my previous team, we tried ordering by difficulty -- mostly to avoid the focus on time. However, I think that bred more confusion than clear
    Message 1 of 58 , Apr 2, 2009
    • 0 Attachment
      Hi James,

      On Thu, Apr 2, 2009 at 20:02, James Carr <james.r.carr@...> wrote:
      >
      > This idea isn't quite new, I'm sure some other shops out there have tried it
      > and I'd like to get some feedback if possible. I've been thinking a bit
      > about what bugs me when it comes to sizing. I don't know if it's sometimes
      > the expectation of sizings related to time, the question of accuracy,
      > etc.... but I for some reason dislike the notion of estimating points that
      > equate to ideal development days (or half days, or sessions). One story's
      > sizing of 1 may be done in an hour, another 1 point story might take a day.
      >
      > What if instead of sizing like that, what if we just sized only 0,1,2,3? 0
      > for something so simple it's more of a chore than a story, 1 for something
      > easy, 2 for something of medium difficulty, and 3 for something that might
      > be a bit hard to do. I think the benefit would be less complex estimation
      > and a much closer "feel" for the how hard each feature will be, or how much
      > bumpiness there will be in the project.

      In my previous team, we tried ordering by difficulty -- mostly to
      avoid the focus on time.

      However, I think that bred more confusion than clear thinking. We had
      a lot of arguments over what the numbers actually meant; how you would
      express estimates that were simple but time-consuming, for example.

      I know some people gave up on trying to use the abstract points and
      back-tracked: "OK, this will probably take us three days, we do 6
      points/iteration, so I'll call it 2". This seems to me to be a
      cardinal sin, as it derives velocity from itself in an incestuous
      manner.

      Anyway, after trying ideal days estimation (and having it interpreted
      as actual days, mind you), I still think the discussion on ideal vs.
      actual is more valid and interesting than trying to define
      "complexity".

      But Mike Cohn makes some convincing arguments for trying to estimate
      size rather than time, here for example:
      http://www.youtube.com/watch?v=fb9Rzyi8b90
      http://www.youtube.com/watch?v=jeT0pOVg0EI

      Cheers,
      - Kim
    • Justin Daubenmire
      Hi Andy, ... From: andy.pickler I think I am seeing the real problem with my environment and why it drives the (perceived) need for these charts. Individuals
      Message 58 of 58 , Apr 6, 2009
      • 0 Attachment
        Hi Andy,

        ----- Original Message -----
        From: andy.pickler
        I think I am seeing the real problem with my environment and why it drives the (perceived) need for these charts. Individuals are spread in individual offices around the program, making "team boards" less helpful or realistic (along with the fact that there is no clear place to put them). Additionally the traditional command-and-control management feels the need to know how the teams are doing. It wouldn't be feasible for them to walk around the various team boards.

        I need to do some more thinking on this and see how in my environment we could help the team out the most (shared, visible charts being the goal) while still meeting the needs of management. I realize that numbers and graphs in software are not as helpful to the team as the index card team board (I've done both approaches on different programs).

        See my post "extreme blind programming update" on using network folders and text files as a possible dispersed team solution.

        People in various parts of the building/office could access the notes folder on the network. Our team set it up like this:

        \\server\notes

        in the notes folder is the project name/names:

        notes\Project1
        notes\Project2
        ...

        In each project folder is a Pending, In Progress, and Completed folder. Our PM monitors the completed folder from her desk looking over the folder to see what completed cards she can work on and also what is in progress and pending.

        Cards are .txt files jotted down in notepad. So you'd see:

        \\ServerName\Notes\ProjectName\Pending
        \\ServerName\Notes\ProjectName\In Progress
        \\ServerName\Notes\ProjectName\Completed

        My team names the text file (story card) after the name of the story such as "pay with a credit card.txt" or "can transfer funds over web service.txt" etc.

        When a team player wants to take on a story, they cut it from the pending folder and paste it into the in progress folder. Likewise, when an in progress story is completed, they cut it and paste it into the completed folder which our pm looks at.


        Regards,
        Justin


        [Non-text portions of this message have been removed]
      Your message has been successfully submitted and would be delivered to recipients shortly.