I got some questions about my tweet from last week on 4th down audibles and pulling goalies. Here’s what I meant in more detail…When you start an iteration, the team makes some best estimates on what can be accomplished in an iteration. The estimates are partially based on effort, but also based on risk. When a story’s implementation requires familiar technologies, developers will often rate it low even if it requires significant effort. Developers tend to rate stories high when there is technical risk.

Halfway into an iteration and the team should be thinking about closing. When the iteration is going well, the team should be considering picking up new stories from the backlog. So what happens when the team hits a story (or stories) that has hit blocks or is just harder than anticipated? The product owner can help in these situations. The team has to hit the top theme and stories for the iteration and finish the iteration on time. You still have to score before the clock reads 0:00… But you can review the play call at 4th down by adjusting stories and even dropping a story if necessary.

At the same time, you have to stick with your values. For me, that means giving the QA team the time and input to validate the stories and release candidate. QA is the last line of defense, the “goalie”, before customers get access to your product. If you’re at a point where you need to compromise QA (e.g., “pull the goalie), my advice would be to write off the iteration and regroup.

The concept of closing reminds me of one of my favorite movie lines from Glengarry Glen Ross, “Always be closing”. Seems appropriate for agile iterations…