On Fri, 2013-02-01 at 09:06 +0100, Vincent Untz wrote:
...

> All credits for this should still go to James, as we used the same
> script, and his explanations for each table are used below.

Excellent work again, kudos guys.

...

> The following table lists the sources of votes for the top 8 candidates
> classified by the employer of the voter. Four categories are listed;
> the first, "%Employer", is the percentage of voters for that candidate
> with the same employer as the candidate. The second, "%Major", is the
> percentage of voters from a "major" employer, which is to say an
> employer that was specified by at least 5 voters overall (there are
> about 20 "major" employers in the data set by this calculation). The
> third, "%Minor", is the percentage from a "minor" employer, i.e., one
> specified by less than 5 individuals. The last, "%None", is the
> percentage who listed no affiliation.
>
>               Source of Votes for Top 8 Candidates
> +-----------------+----------------------+-----------+--------+--------+-------+
> |       Name      |       Employer       | %Employer | %Major | %Minor | %None |
> +-----------------+----------------------+-----------+--------+--------+-------+
> | Monty Taylor    | HP                   |    57 %   |  22 %  |  14 %  |   7 % |
> | Rob Hirschfeld  | Dell                 |    59 %   |  23 %  |  13 %  |   5 % |
> | Troy Toman      | Rackspace            |    68 %   |  18 %  |  10 %  |   4 % |
> | Hui Cheng       | SINA                 |           |        |  29 %  |  31 % |
> | Tim Bell        | CERN                 |           |        |  23 %  |  10 % |
> | Lauren Sell     | OpenStack Foundation |           |        |  19 %  |   8 % |
> | Mark McLoughlin | Red Hat              |           |        |  24 %  |   9 % |
> | Tristan Goode   | Aptira               |           |        |  30 %  |  12 % |
> +-----------------+----------------------+-----------+--------+--------+-------+
>       [Some values have been omitted to preserve anonymity.]

What I'd focus on here is %Employer

Last time around it was:

| Rob Hirschfeld | Dell      |    86 %
| Monty Taylor   | HP        |    75 %
| Joseph George  | Dell      |    88 %
| Troy Toman     | Rackspace |    76 %

The smaller numbers this time around means the candidates' votes were
from a more diverse set of voters.

That's definitely a solid improvement, but it still looks to me like the
big blocks of affiliated voters have an overly large influence over the
result. If the affiliated blocks who voted for the top 3 candidates were
a more average size, I'm guessing they still would rank highly (which is
great) but it would be a much more closely run thing.