Per my Q&A with Slate, I was given the following recommendation for how to proceed:
My advice, I think, is gonna be: write it up how you'd want to see it. But, be clear about the intents at each step, and underlying reasons why you think it would help. At worst, if it's not technically achievable, or if we don't have a way to allocate time to it, having those intents & root causes documented will help guide us to figure out what elections may look like in the future. And if some kind of partial solution is achievable, that might work as well.
With this in mind, I'm proposing the following for each phase of the election process:
- Election Questions - Proposed Duration: 4 weeks
This may come off as the most absurd duration, however, as a whole our community takes awhile to come to consensus on things. Furthermore, we have seen at least a few times that meta Q&A has one answer that floats to the top almost immediately and then sinks back down after several weeks.
Per the site's overall analytics for the past few months, we see very large spikes in visits primarily on Fridays and Saturdays; and if I had to guess, that's probably game night for a lot of folks.
As such, we likely have a lot of folks whom might be learning about the election for the first time on Friday night and then have to get back to the table or something. If they're not visiting daily, the next time they will get a reminder might be next week.
Giving these folks a few weeks to provide their input on questions might be very helpful for garnering a list of questions that are of interest to them.
There are some folks who wish to make the point that the weekend spikes sync with new visits and those are correct. But that doesn't mean there aren't a significant number of regulars who visit the site during that time frame. Taking the time frame from September 30th through today (November 13th), we see the following increases in site visits over the preceding week's average after removing the new visits:
- October 1st: +2,627
- October 2nd: +5,981
- October 8th: +1,688
- October 9th: +4,461
- October 15th: +1,853
- October 16th: +5,136
- October 22nd: +1,913
- October 23rd: +5,013
- October 29th: +327*
- October 30th: +298*
- November 5th: +2,706
- November 6th: +6,872
- November 12th: +1,637
- November 13th: +4,830
*These values were as low as they were due to a maintenance outage.
- Nomination Phase - Proposed Duration: 3 weeks
This most recent election had me contemplating on running myself, but a big hangup for me was having enough time for answering the questions. I suspect I'm not the only one in that position, either. Between a very busy work schedule, home life, and other obligations it is often difficult to find a few hours to read, think, and write up responses to questions.
This may mean that I'm too busy to take on another obligation, or it may mean that I need more time to figure out how to fit this potential obligation into my life. Either way, time is necessary for this.
- Primary Phase - Proposed Duration: 2 weeks
To the best of my knowledge, we've only had 1 primary. However, with the increased time for the nomination phase the plausibility of a crowded field becomes more practical. I don't consider this to be a bad thing, though.
I recommend the 2-week schedule simply because there's a lot of candidate responses to comb through at this point.
- Election Phase - Proposed Duration: 1 week (2 weeks if there's no Primary)
At this point, the community as a whole is effectively aware of the election and has had ample time to review the candidates' responses. Thus, this is proposed to remain the same 1 week duration as before. This allows those folks who pop in just on Friday night to quickly make their picks and then get back to roll bones against the goblin horde.
In the most recent election (election 6) 7,510 were eligible. Of those, 1,320 visited during the election (17.5% of electorate). Of those, 719 visited the election page (9.57% of electorate). And of those, only 323 voted (4.30% of electorate). Based on folks showing some interest in the election and those whom cast a vote, there's a significant disparity of nearly 55%.
We may be able to close that disparity by giving those folks more time to provide their input. This may be doubly necessary if they're part of that group whom seem to visit most often on Fridays and Saturdays.
This phenomenon isn't unique to this election, rather it seems to be about par for most of them:
- Election 5: 23% visit site; 13% visit election; 6.3% voted (52.8% disparity)
- Election 4: 33% visit site; 18% visit election; 10.6% voted (41.4% disparity)
- Election 3: 30% visit site; 26% visit election; 12.2% voted (52% disparity)
- Election 2: 40% visit site; 25% visit election; 17.9% voted (27.4% disparity)
- Election 1: 40% visit site; 22% visit election; 16.4% voted (26.8% disparity)
Disparity is defined as 1 minus number of people who visit the election page divided by the number of people who cast a vote.
If possible, I'd like to see this disparity number as close to 0% as practicable. I note that elections where we had the least amount of disparity were those early on and the candidate responses were pretty brief (maybe just a 2-3 paragraphs).
Given the much more extensive responses we expect from candidates, affording voters more time to consider those responses can be helpful for garnering greater participation.
Is this too long a period?
A concern has been expressed that a scheme like this would leave the site short of moderators for too long. On this site, at least, elections have rarely been run quickly to fill a perceived urgent need. Specifically:
Election 2: unknown whether a time-pressure to elect SSD and WaxE was felt--perhaps mxyzplk could chime in?
Election 3: Wax Eagle had been largely inactive for six months before the election; mxyzplk and SSD reported at the time that the site wasn't too much for them without WaxE (but that a broader representation of viewpoints was desirable).
Election 4: SSD and mxyzplk had decided well in advance to step down and did so when SE was able to run an election. (source: nitsua60)
Election 5: doppelgreener and nitsua60 had decided well in advance to step down and did so when SE was able to run an election. (source: nitsua60 and doppelgreener in comment)
Election 6: V2 stolen by SE, Inc.; election was scheduled at SE's convenience and then-current mods relay no urgency to run any faster than that. (source: Someone_Evil & Oblivious Sage, in mod-room conversation of 15 Nov 2021)