Tell the Whole Truth – @ken_power #xp2011

This was yet another fabulous talk from Ken “The” Power about estimating. For Ken though, if possible, the best way is to not estimate at all. Just trust that the team will do the work as fast as they can.

Unfortunately most teams have people asking “when” things will be done, so they need estimations. We’ve moved away from giving man hour estimations to using points, but that’s a whole other blog post.

But rather than give your story a single point estimation like “that’s 5 points”, give it a range. This adds a lot more richness and depth to your planning sessions, igniting much more interesting discussions. At the beginning of the project stories will be harder to estimate as you need more knowledge, therefore the ranges will be wider, as the projects move on they will become much smaller.

Thinking in range forces you to be considerate of Risk. Allows an upper and lower bound to your velocity, if you track the actual it should be in the middle. If you need to give an estimate use the value 2/3rd between the lower and upper bounds.

How do you do this in a planning poke session? You hold up two cards instead of one!

Happy estimating!

2 thoughts on “Tell the Whole Truth – @ken_power #xp2011

  1. I’ve tried this approach in the past, and I understand the reason for it, but find that Novice teams have real trouble with it, and so I don’t recommend it to them. In particular, Novice teams already worry enough about the accuracy of their estimates, and while you’d think that presenting a range would calm them, it appears to inflame their fears and leads to estimating paralysis. Perhaps Ken has a magic formula for dealing with that, and if he does, I’ll buy a copy!

    I teach a different technique: for teams that like Planning Poker, I ask for an initial estimate from the group. If all the estimates fall into consecutive groups (example: 5, 5, 8, 8, 5, 8, 5), then I instruct the team to go with the higher estimate (here, 8). They move quickly, and they see that they needn’t fuss over low-level estimates. If, however, the estimates fall in a larger range (3, 5, 8, 5, 13, 3, 3, 8, 8), then I instruct the team to spend 5-10 minutes writing out tasks. As they do this, they’ll exchange information about the previously-unstated assumptions they’d put into their estimates. They tend to converge quite quickly, estimate again, then invoke the first of these two rules.

    Most teams who learn this technique estimate stories more quickly without any abnormal problems with estimate accuracy over time.

    I do worry, though, about total team blindness. From time to time, they’ll have similar estimates, all have it completely wrong, and the story will blow up in their face. That’s when I invoke the “ridiculous simplicity”[1] rule and stop worrying.


What did you think?

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s