How much should we sharpen the saw?

You probably know the classic story of the man sawing the tree. You walk up to him.

You: "How's it going?"
Sawer: "Wow. It's really hard. This is one big tree, and I have been working at it for hours now."
You: "Sounds tough. Why is it going so slowly?"
Sawer: "Well, this is hard wood. [He is still sawing.] (Puff.) And my saw has become dull. (Puff)"
You: "Why don't you take the time to sharpen it?"
Sawer: "Can't. (Puff.) (Puff.) I am in a rush."

And the point is that the sawer is not thinking well. He could easily have saved time by sharpening the saw.

But, they always feel that sharpening is taking too much away from doing 'real work'.

And the truth is that sometimes "they" are right about this.

So, how much team time should we allocate to 'sharpening the saw'? In Scrum, say.

Some of you will notice that I have a picture of Coach K above. Coach K, for those who don't know, is the ScrumMaster of the Duke Blue Devils basketball team. A real team. His team won the NCAA Championship last Spring (in what they call March Madness). And his team is #1 in US college basketball today, as I write this. And is undefeated. He is a very, very, very good coach.

So, if we use him, his team, his other coaches, his managers, his trainer, etc, etc, as an example, what is a reasonable allocation to "impediment removal" for one team? If you really want to win.

My answer in practical Scrum terms: one full-time person (ScrumMaster) on impediment removal per full-time team of 7 people is quite reasonable. Probably an under-allocation, but quite reasonable.

And I take the view that none of us has won the NCAA Championship yet (or our equivalent), so until we do, we have plenty of impediments to remove. (And by the way, our team members virtually always do not join the team with the same raw ability in our sport with which Coach K's recruits join his team.)

And I notice that Duke does not fire Coach K after they win the NCAA. And New England does not fire Coach Belichick after they win the Super Bowl. Just repeating at that high level is almost impossible. But again, I personally have never seen a Scrum team that could argue that it was the best team (or even thought that it might be), so this worry that we have 'topped out' seems like an issue not worth talking about for a long time.

Now, if the SM is not very good or if the organization or some other factor means that the SM's efforts will not increase team velocity, then maybe 1 to 7 is too high an allocation. But it is often the first job of the SM to learn how to remove impediments effectively, and to teach the organization (the managers) how worthwhile that can be. Those are the first two impediments, sometimes.

Doesn't this allocation seem reasonable now?