Open Stack

Anne Gentle wrote:
> You point to an idea number of 11 for the group size, and I would like
> some citation for where that number comes from. I know of the two-pizza
> rule in tech, and there's group research around "ten-groups" (saying
> eight to fourteen people in a group is about right for overcoming human
> nature and collaborating effectively, citing book /Corporation Man/ by
> Antony Jay (Pelican 1975)). What causes you to land on 11? Could we say
> 10 or a range of 8-14 instead?

It's the lower bound of Dunbar's "sympathy group" size, rounded to the
next odd number to avoid stand-off during voting. It ensures we all know
each other well and therefore favors consensual decisions. It also
matches my experience of efficiency in online groups meeting on IRC.
Above 11 it's starting to get more difficult to have a discussion that
would reach consensus rather than bikeshedding. 13 is still ok, 15 is
starting to cause problems.

> Defining the ideal group size will help with scoring. For example, I
> don't think the difference between 11 members and 13 members merits a +2
> vs. a +1 score.

Yeah, I think 13 is definitely OK too. Maybe both should get +1 scoring :)

> I also think the ideal number of members will help determine whether
> categories are useful, and further defining categories to discover how
> many there may be will help score that one better.

Personally I think anything between 8 and 14 would be acceptable. The
trick with categories is that we may just have to add more in the future
to cover for the evolution of our community, so aiming a bit lower to
keep room for growth might be wise.

-- 
Thierry Carrez (ttx)