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

Re: Release Planning based on Velocity

Expand Messages
  • MarvinToll.com
    Ron, hmmm... Truthfully, I ve pretty much ignored the velocity nuances that have sparked debate for many years - not that veloctiy doesn t have value. I ve
    Message 1 of 18 , Jul 1, 2012
    • 0 Attachment
      Ron,

      hmmm... Truthfully, I've pretty much ignored the 'velocity' nuances that have sparked debate for many years - not that veloctiy doesn't have value.

      I've also pretty much ignored Scrum. (My wife accuses me of selective hearing - but that is another topic.)

      However, I was in a Scrum class (as a participant joining a new Scrum team) this past week and was a bit unnerved by the strong emphasis on velocity... and more specifically by the context that was established... that management needs predictability so 'velocity' is (ultimately) the answer to THAT problem.

      Call it a Scrum smell... which led your statement to resonate sharply:

      "an unhealthy focus on velocity is often a big impediment to doing Scrum well."

      So I'm trying to come to terms with why I was intuitively bothered and your strong statement.

      Said another way, I'm wondering if I was bothered for the same reason(s) you made your statement?

      So... with your indulgence... let me combine my emerging thoughts with your elbatoration and see if we are (however unlikely) on the same page:

      1. That determining velocity is an enabler - not an end in itself.
      2. That velocity enables workload planning - not predicting.
      3. That extending a workload planning tool into a predicting tool can damage the effectiveness of the planning tool.
      4. That the type of predictability leadership (aka management) needs is related to amount of value... not volume of work... so there is no reason to damage the effectiveness of velocity as a tool for planning.

      _Marvin
      http://PatternEnabled.com


      --- In extremeprogramming@yahoogroups.com, RonJeffries <ronjeffries@...> wrote:
      >
      > Hello Ajithesh,
      >
      > Your question explores an interesting area, tracking velocity and using it for prediction. I often wish that the idea of velocity had never been imported into Scrum from XP, and I often wish we had not invented it or made such a big deal of it in XP.
      >
      > That said, the same issues will arise even if all one does is use the old idea of the Product Burndown, so I suppose it's not fair to blame velocity for all the misconceptions that arise.
      >
      > However it came about, and I blame myself, an unhealthy focus on velocity is often a big impediment to doing Scrum well. It is important that velocity be consistent. Increasing it is not usually the most important thing a Scrum Team can do.
      >
      > Anyway ...
      >
      > On Jun 30, 2012, at 3:02 AM, Ajithesh Hegde wrote:
      >
      > > A team's velocity might vary if:
      > > i) the iteration length varies,
      > > ii) the team size/team composition varies
      > iii) the story size changes;
      > iv) the stories involve a domain the team is more, or less, familiar with;
      > v) the code becomes more clean, or becomes less clean;
      > vi) the team does more, or less, testing;
      > vii) the team does more, or less, refactoring;
      > viii) the team undergoes more, or fewer interruptions;
      > ix) the team injects more, or fewer, defects;
      > x) ... i can think of more ... can you?
      > >
      > > If a team is to ramp up progressively in number (and later progressively
      > > ramp down) over iterations, how to do the release planning based on
      > > velocity? The same question when the iteration length is not kept constant
      > > over iterations.
      >
      >
      > Do you mean that the team is supposed to ramp up velocity, and ramp it back down? If so, why would they be supposed to do that?
      > Or do you mean that they in fact do that without being asked? What is causing the ramping?
      > Or do you mean something else?
      >
      > Basically, predictability is only possible in the presence of consistent velocity. Consistent velocity is only possible if all of the items in the list above are controlled rather tightly. In the case of a software project, the best known ways to do that are thorough testing, mostly automated, continuous design improvement, and a number of other well-known if rarely done technical practices.
      >
      > Furthermore, the desire for predictability often indicates that the Scrum process is being used in an ineffective fashion. It usually carries the idea of "when is the team going to be done with all this". Scum requires the Product Owner, not the team, to be responsible for maximizing the value of the project by the due date. This is accomplished by the PO managing the value of what the team does rather than trying, somehow, to manage the volume of what they do.
      >
      > Ron Jeffries
      > www.XProgramming.com
      > I'm not bad, I'm just drawn that way. -- Jessica Rabbit
    • RonJeffries
      Oddly enough, yes R ... Ron Jeffries www.XProgramming.com There s no word for accountability in Finnish. Accountability is something that is left when
      Message 2 of 18 , Jul 1, 2012
      • 0 Attachment
        Oddly enough, yes
        R
        On Jul 1, 2012, at 9:54 AM, MarvinToll.com wrote:

        > So... with your indulgence... let me combine my emerging thoughts with your elbatoration and see if we are (however unlikely) on the same page:
        >
        > 1. That determining velocity is an enabler - not an end in itself.
        > 2. That velocity enables workload planning - not predicting.
        > 3. That extending a workload planning tool into a predicting tool can damage the effectiveness of the planning tool.
        > 4. That the type of predictability leadership (aka management) needs is related to amount of value... not volume of work... so there is no reason to damage the effectiveness of velocity as a tool for planning.


        Ron Jeffries
        www.XProgramming.com
        There's no word for accountability in Finnish.
        Accountability is something that is left when responsibility has been subtracted.
        --Pasi Sahlberg



        [Non-text portions of this message have been removed]
      • Jonathan Harley
        Hey Ron, I ve always viewed velocity as a fact rather than a goal, and use it to try to have a sense of what the team has been capable of, capacity-wise rather
        Message 3 of 18 , Jul 1, 2012
        • 0 Attachment
          Hey Ron,

          I've always viewed velocity as a fact rather than a goal, and use it to try
          to have a sense of what the team has been capable of, capacity-wise rather
          than a promise of delivery.

          Is your regret about use of velocity in Scrum and XP that it takes so long
          for management to understand the nuances thereby causing undue heartburn
          all around? Or is there some other deep seated reason that I've missed
          along the way? (My fault for not reading all the posts regularly?)

          Thanks - Jon

          On Sun, Jul 1, 2012 at 11:04 AM, RonJeffries <ronjeffries@...> wrote:

          > **
          >
          >
          > Oddly enough, yes
          > R
          >
          > On Jul 1, 2012, at 9:54 AM, MarvinToll.com wrote:
          >
          > > So... with your indulgence... let me combine my emerging thoughts with
          > your elbatoration and see if we are (however unlikely) on the same page:
          > >
          > > 1. That determining velocity is an enabler - not an end in itself.
          > > 2. That velocity enables workload planning - not predicting.
          > > 3. That extending a workload planning tool into a predicting tool can
          > damage the effectiveness of the planning tool.
          > > 4. That the type of predictability leadership (aka management) needs is
          > related to amount of value... not volume of work... so there is no reason
          > to damage the effectiveness of velocity as a tool for planning.
          >
          > Ron Jeffries
          > www.XProgramming.com
          > There's no word for accountability in Finnish.
          > Accountability is something that is left when responsibility has been
          > subtracted.
          > --Pasi Sahlberg
          >
          >
          > [Non-text portions of this message have been removed]
          >
          >
          >


          [Non-text portions of this message have been removed]
        • RonJeffries
          ... Yes. It was originally used as yesterday s weather to help the team decide how much work to take on in the next iteration. Then we began using it in a
          Message 4 of 18 , Jul 1, 2012
          • 0 Attachment
            On Jul 1, 2012, at 11:25 AM, Jonathan Harley wrote:

            > I've always viewed velocity as a fact rather than a goal, and use it to try
            > to have a sense of what the team has been capable of, capacity-wise rather
            > than a promise of delivery.

            Yes. It was originally used as "yesterday's weather" to help the team decide how much work to take on in the next iteration.

            Then we began using it in a healthy way -- yes, that is actually possible -- to give an indication of progress, to get a sense of whether we would be done on time. I don't remember now why we felt the way we did.
            >
            > Is your regret about use of velocity in Scrum and XP that it takes so long
            > for management to understand the nuances thereby causing undue heartburn
            > all around? Or is there some other deep seated reason that I've missed
            > along the way? (My fault for not reading all the posts regularly?)


            In my opinion, "predicting" is not good thinking. Steering, by selecting high value things to do, is important. Having a potentially shippable "product increment" all the time, is important.

            Generally -- almost always -- management's focus on velocity is a focus on "doing more". This is -- almost always -- an indication that they are not focused on having an always-ready increment and always putting the most important new features into it. They are managing the cost end of the equation, not the value end.

            That way lies pain, mediocrity, and bad software.

            Ron Jeffries
            www.XProgramming.com
            I know we always like to say it'll be easier to do it now than it
            will be to do it later. Not likely. I plan to be smarter later than
            I am now, so I think it'll be just as easy later, maybe even easier.
            Why pay now when we can pay later?



            [Non-text portions of this message have been removed]
          • Ajithesh Hegde
            Hi, My clarifications on the two varying parameters: 1. Gradual increase in the team size in the beginning and gradual decrease in the team size towards the
            Message 5 of 18 , Jul 2, 2012
            • 0 Attachment
              Hi,

              My clarifications on the two varying parameters:

              1. Gradual increase in the team size in the beginning and gradual decrease
              in the team size towards the end of the project. Why should this happen?:

              Answer: I have seen such a practice in many of the projects as a
              tradition. In the beginning of a project, the core team is formed with a
              fewer members (generally the senior members). The initial ground work
              such as high level requirements and architecture starts happening.

              With the high level details reasonably worked out, the team size is
              increased with more members (generally relatively junior members).

              A size down happens when the project is nearing completion on similar
              grounds. As the project is well on the track and most of the project is
              done by now, more senior members are pulled out gradually and are put on
              newer projects.

              2. Iteration length. Why is this varied?
              I have seen the teams taking such a decision as part of their sprint
              retrospectives to experiment with different iteration lengths to find out
              which is the more optimal iteration length for their work. Sometimes,
              based on the theme that they are taking for the next iteration, I have seen
              the teams dynamically changing the iteration length as well.

              Rgds
              Ajithesh


              On Sun, Jul 1, 2012 at 9:06 PM, RonJeffries <ronjeffries@...> wrote:

              > **
              >
              >
              >
              > On Jul 1, 2012, at 11:25 AM, Jonathan Harley wrote:
              >
              > > I've always viewed velocity as a fact rather than a goal, and use it to
              > try
              > > to have a sense of what the team has been capable of, capacity-wise
              > rather
              > > than a promise of delivery.
              >
              > Yes. It was originally used as "yesterday's weather" to help the team
              > decide how much work to take on in the next iteration.
              >
              > Then we began using it in a healthy way -- yes, that is actually possible
              > -- to give an indication of progress, to get a sense of whether we would be
              > done on time. I don't remember now why we felt the way we did.
              >
              > >
              > > Is your regret about use of velocity in Scrum and XP that it takes so
              > long
              > > for management to understand the nuances thereby causing undue heartburn
              > > all around? Or is there some other deep seated reason that I've missed
              > > along the way? (My fault for not reading all the posts regularly?)
              >
              > In my opinion, "predicting" is not good thinking. Steering, by selecting
              > high value things to do, is important. Having a potentially shippable
              > "product increment" all the time, is important.
              >
              > Generally -- almost always -- management's focus on velocity is a focus on
              > "doing more". This is -- almost always -- an indication that they are not
              > focused on having an always-ready increment and always putting the most
              > important new features into it. They are managing the cost end of the
              > equation, not the value end.
              >
              > That way lies pain, mediocrity, and bad software.
              >
              > Ron Jeffries
              > www.XProgramming.com
              > I know we always like to say it'll be easier to do it now than it
              > will be to do it later. Not likely. I plan to be smarter later than
              > I am now, so I think it'll be just as easy later, maybe even easier.
              > Why pay now when we can pay later?
              >
              >
              > [Non-text portions of this message have been removed]
              >
              >
              >


              [Non-text portions of this message have been removed]
            • George Dinwiddie
              Ajithesh, ... That s a tradition from a phased, plan-driven project lifecycle. It s a tradition that treats software development as an endeavor of a group of
              Message 6 of 18 , Jul 2, 2012
              • 0 Attachment
                Ajithesh,

                On 7/2/12 4:55 AM, Ajithesh Hegde wrote:
                > Hi,
                >
                > My clarifications on the two varying parameters:
                >
                > 1. Gradual increase in the team size in the beginning and gradual decrease
                > in the team size towards the end of the project. Why should this happen?:
                >
                > Answer: I have seen such a practice in many of the projects as a
                > tradition. In the beginning of a project, the core team is formed with a
                > fewer members (generally the senior members). The initial ground work
                > such as high level requirements and architecture starts happening.

                That's a tradition from a phased, plan-driven project lifecycle. It's a
                tradition that treats software development as an endeavor of a group of
                individuals rather than as a team activity.

                > With the high level details reasonably worked out, the team size is
                > increased with more members (generally relatively junior members).
                >
                > A size down happens when the project is nearing completion on similar
                > grounds. As the project is well on the track and most of the project is
                > done by now, more senior members are pulled out gradually and are put on
                > newer projects.

                That's always a good idea, so that the more senior members don't get
                tarred by the project failure during final integration and test. ;-)
                Seriously, I've seen that play out on a number of occasions in just such
                a fashion. It's one of the frequent plot lines of a serial project
                lifecycle, and one of the problems that an adaptive, team-driven
                lifecycle is intended to avoid.

                > 2. Iteration length. Why is this varied?
                > I have seen the teams taking such a decision as part of their sprint
                > retrospectives to experiment with different iteration lengths to find out
                > which is the more optimal iteration length for their work. Sometimes,
                > based on the theme that they are taking for the next iteration, I have seen
                > the teams dynamically changing the iteration length as well.

                Why do these teams dynamically change their iteration length?

                - George

                --
                ----------------------------------------------------------------------
                * George Dinwiddie * http://blog.gdinwiddie.com
                Software Development http://www.idiacomputing.com
                Consultant and Coach http://www.agilemaryland.org
                ----------------------------------------------------------------------
              • Tim Ottinger
                Goodhart s law applied to software productivity. :-) Roughly, If you lean on a gauge, it quits providing useful information.   Tim Ottinger
                Message 7 of 18 , Jul 2, 2012
                • 0 Attachment
                  Goodhart's law applied to software productivity. :-)

                  Roughly, "If you lean on a gauge, it quits providing useful information."

                   
                  Tim Ottinger
                  http://agileinaflash.blogspot.com/
                  http://agileotter.blogspot.com/
                • Curtis Cooley
                  Sorry for the top post, using my phone :-( Others have hinted but I ve not seen anyone say, building projects is easy, building teams is hard. On the off
                  Message 8 of 18 , Jul 2, 2012
                  • 0 Attachment
                    Sorry for the top post, using my phone :-(

                    Others have hinted but I've not seen anyone say, building projects is easy,
                    building teams is hard. On the off chance you manage to build a high
                    performance team, you already have plans to tear of down.

                    I suggest you don't do that.
                    On Jul 2, 2012 1:55 AM, "Ajithesh Hegde" <ajithesh.gh@...> wrote:


                    [Non-text portions of this message have been removed]
                  • MarvinToll.com
                    Is this akin to the notion that: If you push a metaphor too hard the wheels kind of fall off. [Richard Greene]
                    Message 9 of 18 , Jul 3, 2012
                    • 0 Attachment
                      Is this akin to the notion that:

                      "If you push a metaphor too hard the wheels kind of fall off." [Richard Greene]

                      --- In extremeprogramming@yahoogroups.com, Tim Ottinger <linux_tim@...> wrote:
                      >
                      > Goodhart's law applied to software productivity. :-)
                      >
                      > Roughly, "If you lean on a gauge, it quits providing useful information."
                      >
                      >  
                      > Tim Ottinger
                      > http://agileinaflash.blogspot.com/
                      > http://agileotter.blogspot.com/
                      >
                    Your message has been successfully submitted and would be delivered to recipients shortly.