5

Summary

  1. 90% Answered: Our Area51 %Answered metric should read 90% by May 31st 23:59 UTC.
  2. Improved Voting Economy: The Electorate badge should have been awarded 15 times total by May 31st 23:59 UTC.
  3. Badges: Every 24-hour UTC period from May 1st - May 31st with at least three active bounties counts as a success.

Results: Week 1, Week 2.

Relevant posts: Spring Cleaning 2013, advanced search filters (more), voting strategies, how to write a good answer, vote early vote often.

Spring cleaning

As I announced for the first chat event, I think our site should make a coordinated effort of cleaning up the site during the month of May. While we had some very productive discussions during the chat, we didn't actually manage to cover much ground for me to base this announcement on, which, as I said in the summary, is mostly my fault.

In any case, I want to move on this project, so I'm going to try to provide some goals and a general framework, make some suggestions, and encourage people to post their own answers. There will be a 'halfway discussion' at some point from the 14-16th of May to go over how we are doing, and to discuss any ideas that have come up during spring cleaning, as well as whatever other topics people have.

Goals

Spring cleaning has two complementary purposes:

  • Signaling a discrete period of high site engagement to formerly active users who have gone passive for whatever reason.
  • Improving CogSci's Beta metrics, user participation/retention and general site quality.

(The first goal is basically achieved with this post, success notwithstanding, but I wanted to list it explicitly.)

To make these abstract goals tangible, I am setting the following concrete objectives for Spring Cleaning 2015.

Goal 1: 90% Answered

Target: Our Area51 %Answered metric should read 90% by May 31st 23:59 UTC.

Since late March, we've gone from 77% to 82% on April 19th, and then from 82 to 84% in less than two weeks more. A lot of the early gains came from serial closing nearly everything in our 'question purgatory,' true, but recent gains have primarily come from new answers. It's not a low bar to set, to be sure, but there is every possibility that we can reach the Promised Land with only a little elbow grease here!

To raise the chances we will meet this goal, I am personally setting off time to write two answers a day on average through May. It would be nice if a few other regulars would join me in making a (smaller, I would expect) public commitment, but I will do it regardless of what happens.

This is the most important goal, but also the most labor-intensive to contribute to. We should heed Jeromy's suggestion from my original spring cleaning proposal two years ago: answers don't need to be perfect to contribute substantial value to the site, they just need to be good. Another thing to keep in mind is that pursuing negative answers (i.e., answers which affirm that no answer exists) can be a strong strategy here, because most of the work goes into the preceding literature search, which can easily be done in bits and pieces. I suspect many unanswered questions have a negative answer.

Goal 2: Improved Voting Economy

Target: The Electorate badge should have been awarded 15 times total by May 31st 23:59 UTC. We each get 30 votes every day, so this can be accomplished by any user within the month.

enter image description here

I've talked about our voting economy before here. Votes are as important for an SE site as high quality questions and answers are, because the vast, vast, vast majority of people who come to this site are not in a position to evaluate the merits of an answer in any way except votes.

Relative comparisons with other Stack Exchange sites notwithstanding, the SE framework imposes an objective framework on every SE site. Answers go grey when they reach an absolute score of -3, regardless of what -3 means. Questions and Answers are "good" for SE's definition of "good" at an absolute score of 10. We can't change these things, so our voting economy needs to reflect them. (I doubt we can actually reach this ideal distribution, but it's as tangible a collective voting goal as any.)

Presently, I think it's reasonable to say our voting economy does not reflect these norms. In other words, we need more upvotes and downvotes both. Downvoting answers cost rep, but it's a miniscule -1 reps, you can downvote 10 answers for every upvote your answers get. If you don't have time to write answers, try to vote early and vote often.

Goal 3: Three Bounties

Target: Every 24-hour UTC period from May 1st - May 31st with at least three active bounties counts as a success.

Bounties are not that expensive in terms of reputation, and are a great way to signal site activity. Bounties are also effective: I have set a half dozen bounties on my own questions or ones I simply found interesting in the past month, and all of them received a satisfying answer. If you're a previously active user sitting on thousands of rep, who doesn't have time to answer questions and doesn't have time to vote, finding questions that deserve a bounty is a great way to help the site.

We can do this, so let's do it!

  • 1
    I don't fully agree on goal 2. I don't believe you can make any generalizations across sites on what vote counts mean. The only real way to assess what a vote count means is by taking site specific details into account, as well as the amount of viewers a particular question got. – Steven Jeuris Apr 30 '15 at 13:05
  • @StevenJeuris Normal users don't have access to that information, so I went with the objective features of the SE framework. In any case, goal #2 is just to get people to vote more, so you can always vote more for a different reason, if you like. – Christian Hummeluhr Apr 30 '15 at 13:07
  • Or propose a new formulation, I'm not married to this stuff. :v – Christian Hummeluhr Apr 30 '15 at 13:09
6

This is great! Just remember that quality still comes first.

Hats off to Christian for this very well thought out initiative. It's great to see this kind of push from within the community. Just a friendly reminder from the Stack Exchange Community Team (of which I'm a member); while you go through this (awesome) exercise, just make sure you don't sacrifice quality in pursuit of these targets.

The way we think about things on the Community Team, having well written, expert Q&A is always goal #1. Every SE site is creating an archive of top notch domain knowledge for future readers. So please, band together, have fun and take ownership of your site! Just keep the larger goal in mind too.

  • Praise from Caesar is praise indeed. Maybe you could aptly throw a call to vote more in there, since we can't just reach into the Platonic realm to identify our own quality answers from not-quality a priori. I mean, I always do my best, but sometimes my best isn't good enough―hence voting. – Christian Hummeluhr May 2 '15 at 6:43
  • Also, are you our handler, or just interested? :v – Christian Hummeluhr May 2 '15 at 6:47
  • 1
    We don't exactly have handlers for Stack Exchange sites, I'm just one of several who floats about the network and sleeps with one eye open. ;) You're raising several potential points in your first comment and instead of assuming I'm understanding you correctly, I's rather make sure. Care to speak in chat? – Ana May 4 '15 at 20:48
  • I'm almost always in chat (definitely during spring cleaning), so sure. – Christian Hummeluhr May 5 '15 at 6:02
5

Just to throw out some numbers on goal #1:

There are currently 2,674 questions on the site. Getting to 90% would require answering 2,407 of them. There are 412 unanswered questions right now, so we would need to answer about 145 of those questions in order to reach the goal.

Of course, we get new questions at a rate of about 3/day right now. At the end of May, this will mean an addition 90 questions. So we would also need to answer about 80 additional questions to keep up with the new questions.

All in all, achieving goal #1 would mean about 225 questions without an answer get answered before the end of May.

(Closing questions can also factor into this, but I'm ignoring that to simplify).

  • Thanks @Josh, concrete numbers are extremely helpful. – Christian Hummeluhr Apr 30 '15 at 14:26
  • To add to this, we answered 111 questions during the month of April so far, so we need to approximately double our answer rate for one month, and then we'll only need to maintain the old rate to make slow gains into 90%+. – Christian Hummeluhr Apr 30 '15 at 14:56
4

WEEK 2: The Numbers

%Answered: 86% (+2% since 01/05/2015).

Unanswered questions: 369 (-43 since 01/05/2015).

New Electorate/Civic Duty badges: 1 Civic Duty.

Days with 3 or more bounties: 1/7.

Here are a couple of plots. Because I spent a lot of time yesterday making convenience plotting functions in R, so why not use them for this too? YAY, PLOTS.

enter image description here

enter image description here

3

WEEK 1: The Numbers

%Answered: 85% (+1%).

Unanswered questions: 392 (-20). Note that this is despite losing some ~10 when a suspended user deleted their account.

New Electorate/Civic Duty badges: 0. Some progress: a question actually managed to reach 10 score in less than a week!

Days with 3 or more bounties: 7 (100%).

Other: Questions per day is up almost a whole point.

You must log in to answer this question.

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