The article buried it a bit: cycle time. Reducing cycle time means more value faster. This is a good metric for driving discussions. As for points, the best sy

The article buried it a bit: cycle time. Reducing cycle time means more value faster. This is a good metric for driving discussions. As for points, the best system I've worked with is 1, 2, or 3 points. 1: we know exactly what to do. 2: we have an idea what to do and we feel the story should not be broken up any further. And 3: we don't know how to solve this; break it up or create a spike to study the problem to break it down or accept the unknown. Everything practically becomes a 2, and is thus in line with the article - the number of stories is pretty reliable between sprints and for projections. The advantage of 1, 2, or 3 over "no points" is that it drives conversions on if something should be broken down. The team quickly starts creating tickets that generally are 2s and pointing a story typically takes a team of 6-8 about 10 seconds. A 3 may be talked about for a minute before it is accepted or broken up as agreed to by the group.