Friday, September 30, 2011

Bring me a rock

"Bring me a rock" is an exercise in which a manager or client asks an analyst for a particular piece of work. The analyst runs off and does the work. The conversation looks like this:
  • Mgr: "Bring me a rock."
  • Analyst: Finds a rock and shows manager the rock.
  • Mgr: "Not this rock. I need a different rock."
  • Analyst: Finds another rocks and shows it to the manager.
  • Mgr: "No, this isn't the right rock either. Bring me another rock."
...and so on. This can end in any number of ways, but it usually doesn't end well unless the analyst is able to get the manager to explain his goals for acquiring and using the rock. I've written about one version of this pattern before, and called it generate me some alternatives.

I'm going through a bring-me-a-rock exercise right now that includes a strong element of the value-add demotivation technique. It's interesting, and it really helps to have models of these scenarios to help understand what's going on under the surface.

Saturday, September 10, 2011

Counting points, or "stats are for losers"

Sports fans are familiar with the phrase "Stats are for losers." If your favorite team got hammered last Sunday you can count on a few people jumping into a discussion with an argument about how the team "should have won because we outgained the other guys 500 yrds to 250 blah blah blah..." The simple rejoiner from the winning side is, "Stats are for losers."

I was thinking about this in the context of an agile project that was comparing its iteration point count with another team's. "Yeah, we got 25 points last iteration, the other guys only got 21." Agile practitioners know that if you're doing estimating correctly, the points are only measures of relative effort, comparable only to the team's effort within a release. They can't be used to compare to, for example, another team's point estimate. The team that was congratulating itself had delivered a design that was flawed (and they knew it) but didn't fix it once the flaw was pointed out, because it would have affected their point count.

Stats are for losers. Winning means delivering high quality product that is of value to the business, not having a high score on some arbitrary measure like points.