When to use the point card "0" and "?"?

[Qunfeng Wang]

It's my first Agile meeting: Requitments are mapped into stories, we are asked to give point for each story. And then devide every story to pieces of tasks, assign some work hours to each task. That's really interesting.

I notice the point cards "0" and "?" are seldom used. Do you often use these card in your practice?

Thanks.

[Ilja Preuss]

What point cards? Is that the thing Mike Cohn teaches?

[Qunfeng Wang]

Moutain Goat
Isn't this card widly used in Agile?

[Peer Reynders]

Are you talking about the Story Points on the User Story Cards?

I checked in both Mike Cohn's books User Stories Applied: For Agile Software Development and Agile Estimating and Planning and could not find a particular reference to the 0 and ? Story Point values.

I would guess that:

0 - simply indicates that the User Story depends on another one - i.e. the functionality will come into existence when another user story is completed. Otherwise a 0 would make no sense - you would be asserting that a user story can be realized without any effort.

? - indicates that the User Story cannot be currently estimated. In this case you have to allocate some time in the next iteration to implement a spike (Tracer Bullets) for that User Story, so that you can later assign a more educated estimate to it.

In any case I would assume that a ? is more likely than 0 and that both of these are much less common than a positive, non-zero Story Point Value.

[Qunfeng Wang]

Sorry, I think I should describe it more clearly:

Requirements---->Many stories
A story<--------->A point
A story----------> Many tasks
A task<----------> Work hours

The Moutain Goat Software has a series of Point Card:0, 1, 2, 3, 5, 8, ?
We give each story a point with a card. We try my best to implement these stories to get a good point in one iteration. We repeat this in the next.

It's very interesting. I don't find the card on MoutainGoatSoftware's webpage either. But we use it in our meeting.

[Peer Reynders]

Mike Cohn alludes to the Point Cards in his Presentation (Slide 20) Agile Estimating and Planning Slides (PDF) but doesn't mention them in his books.

However these seemed more as a device to determine how many stories you can implement in a single iteration.

Also I'm a bit surprised by "tasks <-> work hours". Story Points are an abstraction of effort that is supposed to be decoupled from (personal/environmental) velocity. The point system allows you to score the implementation effort for all parts of the system relative to one another.

One organization may choose 1 story point to represent one ideal day of uninterrupted work (no meetings) for the ideal developer, while another may chose a quarter of a day in the same environment.

Hours only result once you assign the story to someone who has a known velocity in the current environment.

You make it sound more like a competition - whoever can implement the most points during the iteration 'wins'. 'Lines of code written' has never been a good measure of productivity, 'Story points' implemented is only slightly better as reality will usually show that not all story points are created equal (no matter how hard we try).
[ April 10, 2007: Message edited by: Peer Reynders ]

[Ilja Preuss]

As an aside, what seems to be most widely used is a more simple point system with just the values 1, 2 and 3.

[Lasse Koskela]

Originally posted by Ilja Preuss:
As an aside, what seems to be most widely used is a more simple point system with just the values 1, 2 and 3.

I don't remember seeing a team doing that in the past couple of years while I have seen multiple teams using the logarithmic scale story points (1, 2, 3, 5, 8, ?).

[Ilja Preuss]

Originally posted by Lasse Koskela:

I don't remember seeing a team doing that in the past couple of years while I have seen multiple teams using the logarithmic scale story points (1, 2, 3, 5, 8, ?).

That's interesting - I wasn't aware that so many teams are using it.

How well is it working in your experience? Somehow I wouldn't trust myself to really know the difference between a 5, 8 and ?...

[Stan James]

We used 1,2,4,8 because the bigger they get the poorer your accuracy. 8's were rare and indicated some risk of being very wrong. I think that's the same feeling Ilja was expressing?

[Ilja Preuss]

Originally posted by Stan James:
We used 1,2,4,8 because the bigger they get the poorer your accuracy. 8's were rare and indicated some risk of being very wrong. I think that's the same feeling Ilja was expressing?

Exactly. I feel it might be better to break down the 8 into several stories of smaller size. Which leaves you with 1 2 4 - which is nearly 1 2 3 already...

[Qunfeng Wang]

Originally posted by Lasse Koskela:

I don't remember seeing a team doing that in the past couple of years while I have seen multiple teams using the logarithmic scale story points (1, 2, 3, 5, 8, ?).

Yeah.

That's the card I'm using. I have a "0" in addition.