9
\$\begingroup\$

Q: Is the recentness of a user's site activity relevant to their candidacy for moderator?

(Nominations) (Candidate's Answers)


I collected some data for a simple metric for this election, and thought I would share the results.

Please NOTE: This is but one piece of information about the candidates.

Below is a chart showing all site activity (tab: ?tab=activity&sort=all) relative to how recent the activity was. This is a admittedly crude metric to show how recent, a candidate's activity (and potentially interest) in the site is.

As a simple a measure of how long ago a user's activity was, the horizontal axis is the the 100th. 250th, ... etc, previous activity. The vertical axis is how many months it has been since that specific activity number occurred. Users with lower numbers have been more active recently.

This shows the 6 current candidates, and also myself as a reference. As a reference to interpreting this chart, you can see that I have performed 250 site actions in approximately the last 1.3 months.

This is a simple count of all publicly visible site activity. There may be some activities that are more relevant to being a moderator than others. This metric ignores any of that. I would encourage sharing of other metrics that you might feel are more relevant for a moderator.

enter image description here

\$\endgroup\$
  • 1
    \$\begingroup\$ If I understand the graph correctly, of the candidates, Mast has been the most active recently, while Malachi has been least? The way it's graphed is kind of confusing... \$\endgroup\$ – Phrancis Apr 1 '18 at 2:08
  • \$\begingroup\$ You interpretation is correct. I went back and forth trying to find a reasonable presentation. This was the best I came up with... \$\endgroup\$ – Stephen Rauch Apr 1 '18 at 2:10
  • \$\begingroup\$ Would it be easier to graph the cumulative number of actions against a timescale?. Invert X and Y so to say? \$\endgroup\$ – Vogel612 Apr 1 '18 at 2:14
  • \$\begingroup\$ @Vogel612, The trouble I had with time as the horizontal axis is that this chart is just bins, it is not XY. I could have sampled at a given time instead of event count, but I did this by hand... \$\endgroup\$ – Stephen Rauch Apr 1 '18 at 2:27
  • \$\begingroup\$ wow that's ... A lot of work. Very impressive :) thanks for starting this "thread" to get people to share what metrics they consider important. \$\endgroup\$ – Vogel612 Apr 1 '18 at 2:32
  • 1
    \$\begingroup\$ This is a fascinating way to look at recent activity. From my perspective, it looks very representative of my visible activity since for the past couple of years I have been quieter. I'm not trying to defend my quiet activity because what you're measuring is important, and relevant, but what is not measured is the monitoring-without-visible-actions here - for example, I check the review queues daily and always process whatever reopen, low quality, suggested edit, and late answer queues, but often when I check, those are empty. \$\endgroup\$ – rolfl Apr 1 '18 at 4:22
  • 2
    \$\begingroup\$ Out of interest, @StephenRauch - where's your nomination for the elections? \$\endgroup\$ – rolfl Apr 1 '18 at 4:41
  • \$\begingroup\$ @rolfl, While I feel I have enough interest and user moderation activity in the site to run, with only 3k rep I don't feel like I have participated in the site enough yet to run for moderator. Maybe next time? \$\endgroup\$ – Stephen Rauch Apr 1 '18 at 4:59
  • 2
    \$\begingroup\$ @rolfl, oddly enough you were the case that made me hold off putting this post up for several days. I have spent quite a bit of time in the reviews queues this last year, and you were at the top of basically all of them that entire time. So you are kinda my CR hero. Well after Jamal at least. But he is already a Mod. I recently ran for Mod on SO, and I did this same analysis there (I did not share, seemed self serving) because some of the Nominees had been activity in the stone ages.... \$\endgroup\$ – Stephen Rauch Apr 1 '18 at 5:13
  • 3
    \$\begingroup\$ On SO the Mod work load is kinda crazy, so super active involvement seems crucial. But after consideration, my impression here is that the workload is not so nuts, so understanding how things work, and a willingness to be here consistently, seemed more important than willingness to shoulder a crazy load. And after your nominee statement acknowledged that you have been less active than the very high levels of the past, I decided this data wouldn't hurt your candidacy. So, the discussion begins. \$\endgroup\$ – Stephen Rauch Apr 1 '18 at 5:13
  • 1
    \$\begingroup\$ And if you check the activity for the current moderators, how does that look? (I'm not sure I want to know...) \$\endgroup\$ – Simon Forsberg Apr 1 '18 at 8:50
  • \$\begingroup\$ @SimonForsberg - I actually did check that out... and I was surprised ;-) ( I only checked out the review acivity). \$\endgroup\$ – rolfl Apr 1 '18 at 12:49
  • \$\begingroup\$ would be interesting to see your query. or description on how you gathered the data. just curious. and where is the Activity Graph for the Current Moderators? \$\endgroup\$ – Malachi Apr 5 '18 at 19:01
  • \$\begingroup\$ @Malachi, This was by hand gathered from the user's activity tab, by simply stepping back in time a certain number of pages, and then getting the date for the item on that page. The current moderators have a different role in that a lot of moderation activity is not public. Additionally, if they are active in any of the queues which require multiple votes their votes are binding and thus circumvent community involvement. Thus I prefer, that except for very obvious, close it now cases, the mods let the community handle the normal stuff, so that more voices are heard on these things. \$\endgroup\$ – Stephen Rauch Apr 5 '18 at 19:14
  • 1
    \$\begingroup\$ @Malachi, Since I stopped doing first posts reviews after doing 2000 of them, I can understand the sentiment of letting others get involved. I hope no one uses this metric as the only way to evaluate candidates. But my reference to moderator actions was that the moderator queues do not show up in the public stats. So looking at them in the same light as non-moderators doesn't make as much sense. \$\endgroup\$ – Stephen Rauch Apr 6 '18 at 5:08
9
\$\begingroup\$

This graph is a bit more detailed. What you see are the number of actions taken within buckets of 8 days, starting (at the left) on Jan 1st 2016

That is, if I can get my SQL right at 4:30 AM...

Data obtained from this query with the following parameters:

UserIDs: 31503,52915,42632,37660,42401,18427
WindowSize: 8
LeftDate: 2016-01-01 00:00:00

Feel free to play around a bit with this as well :)

The user IDs map as follows: rolfl, Mast, Phrancis, Me, Peilonrayz and Malachi

\$\endgroup\$
  • 1
    \$\begingroup\$ That is a LOT more detailed... \$\endgroup\$ – Stephen Rauch Apr 1 '18 at 2:24
  • 5
    \$\begingroup\$ I think @StephenRauch 's manual process is more relevant for moderator-relevant actions since it includes things like actions in review queues, etc. The above query from Jason C only includes posts (questions/answers), and (suggested)edits (and ignores comments, review queues, etc). \$\endgroup\$ – rolfl Apr 1 '18 at 4:12
  • 1
    \$\begingroup\$ this shows that there are some spots where two users almost flip flop on their actions. which is very interesting to note, because when one person is more active, the actions that need to be taken are not as available because someone has already taken care of them. so the number of actions available at the time should also be taken into account when looking at these numbers, @Rolfl also mentions this in his comment on the question \$\endgroup\$ – Malachi Apr 5 '18 at 18:55
  • \$\begingroup\$ @rolfl, does this query include Close Votes and such? I believe that the Review Queues consider those as VTC's and VTD's don't they? if you made edits, those also count as edits. so I think there is going to be some overlap. \$\endgroup\$ – Malachi Apr 6 '18 at 13:24
  • 1
    \$\begingroup\$ @Malachi - the query only covers edits that high-rep users make, suggested edits that low-rep users make, and actual questions/answers posted. Other meta-activities like close-voting, commenting, review queues, etc are completely ignored by this query/graph. In other words, this query covers only the things that a typical user would no, not the things we expect to be part of a good moderator. \$\endgroup\$ – rolfl Apr 6 '18 at 14:20
1
\$\begingroup\$

Normal user activity does not equate to Moderator activity.

I was curious, and wanted to give a larger picture of what the activity looks like so I changed some of the parameters.

Activity Since Revival

I think it is very important to note that there are major events scattered throughout this timeline, things like

  • New Pro Tempore Moderators assigned by SE
  • Graduation which led to the election of moderators and also more moderator positions
  • More users becoming active on our site and handling actions, distributing the workload

I think that these are important to note because they affect the amount of activity as it is defined here (Edits and posts).


Here is the same set with the Addition of Jamal

Activity Since Revival with Jamal

and because I was curious I did one with the Current Moderators and Myself, same parameters (other than the user ids)

Current Moderators and Malachi

Please keep in mind that the colors for the users have changed between the graphs. Also, I am using the same SEDE query that Vogel used.

if you take into account Votes (this includes votes to: close, reopen, delete, undelete, spam flags, approve edit suggestions, offensive flags) and Comments then the previous graphs look like this, (same order)

1.)

with Votes and Comments

2.)

With Votes and Comments and Jamal

3.)

ID's (accidentally included @Rolfl even though he is not a current Moderator) 31503,18427, 22222,9357, 12390, 23788, 31562

Current Mods and Malachi and Rolf

I think this is one of the most important ones though because it shows that Moderators activities don't truly change that much when they become a moderator.

take a look at Mathieu Guindon's marks on this graph, there are not any drastic up and down movements, same with Simon and 200. I do see some dips that could be attributed to vacations but other than that their behavior as a normal user doesn't change.

the big thing that we should take away from this is that the moderator responsibilities are not the same as normal user actions, so we should keep that in mind while looking at these metrics

\$\endgroup\$
  • 5
    \$\begingroup\$ Interesting insights, again, but I caution the use of the SEDE query you have used. It only covers a small fraction of actual activity on the site, and it is skewed. For example, Much of Jamal's activity is in editing, wheras 200_success is more a combination of editing and posting answers. None of the charts represent posting comments, or review queues, etc. I know from my time as a mod that different mods tended to focus on different aspects of the site, so if the aspect the mod focused on is not covered in the query it's no longer representative \$\endgroup\$ – rolfl Apr 6 '18 at 14:17
  • 4
    \$\begingroup\$ I believe these graphs are totally irrelevant and doesn't say at all how "good" of a moderator anyone is. This just measures some activities, but how important is it whether a moderator has twenty edits in a week, or two hundred? A moderator's job is not to vote or edit posts. \$\endgroup\$ – Simon Forsberg Apr 6 '18 at 16:21
  • \$\begingroup\$ I agree, if you change the number of days or Window Size, you can change the look of that graph drastically, especially over greater time periods. and I agree with you about what makes a "good" moderator. statistics/metrics can show whatever you want them to show. but normal user activity does not equate to Moderator activity \$\endgroup\$ – Malachi Apr 6 '18 at 17:26
  • \$\begingroup\$ I added an upvote to get this out of the negative, but IMO this entire Q&A line is pretty irrelevant \$\endgroup\$ – Raystafarian Apr 8 '18 at 23:01
  • \$\begingroup\$ Thank you @Raystafarian. I agree with you. the metrics don't really measure what was hoped for. \$\endgroup\$ – Malachi Apr 8 '18 at 23:07

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .