29
\$\begingroup\$

I've seen a lot of activity about graduation recently. What are the (possible) benefits (and disadvantages) of a graduated site?

\$\endgroup\$
  • \$\begingroup\$ You should edit your about me section @palacsint, we are now graduated :D ;). \$\endgroup\$ – Marc-Andre Oct 15 '14 at 14:23
  • 1
    \$\begingroup\$ @Marc-Andre: I've missed that before your comment, thanks! \$\endgroup\$ – palacsint Oct 23 '14 at 3:33
26
\$\begingroup\$

Graduation does five things for you, mechanically speaking.

  • Reputation privilege tables are changed. That answer really explains it so I don't need to go further. Given the rep distribution at current, this is actually somewhat negative since you have a pretty tiny fraction of 5% that surpasses 1k rep.
  • The site gets a visual design. You get a fancy logo and branding and a lot of things that help make you stand out. This strikes me as a positive.
  • More likely to have 3k-user-based migration paths in and out. This is based, though, on actual volume and success of migrations. Incoming, SO is the only one even remotely considerable and the rejection rate is too high, nevermind the volume being too low. The other choices SO has for their outgoing migrations still trump Code Review and I can see several other site candidates that are far more likely to get a slot than Code Review. For outgoing, you guys don't have quite enough to warrant consideration yet. So this is neither positive nor negative, and I don't see that changing in the near term.
  • Moderator elections happen. This is the main step to the site being more self-governed, in that instead of the moderators being appointed from on high, the community gets to select the moderators. This is advantageous if the community exists that is devoted to moderation, but if you guys are in dry spots for people who actually are up to the task, then this can be seen as a negative.
  • Community Promotion Ads. You might call this underwhelming but I like making these. A graduated site gets a thread that allows them to host ads that are displayed in the right sidebar to visitors. You can see an example from Ask Ubuntu here. It's a fun way to help provide things that you feel visitors to your community will appreciate. I'd call this a positive.

Overall it's mostly a positive - but let's put it this way. Part of the judgment of when we actually push sites to graduation is that all of these should be positive if the site is ready to graduate. If reputation levels changing, or migration paths opening, or moderator elections happening, if these can be seen as negative to the site's progress, then it is unlikely that the site is ready to graduate.

\$\endgroup\$
  • 6
    \$\begingroup\$ You say that the percentage of 1K users matters. May I ask, why is that, i.e. why isn't it only the absolute number of privileged/active users that matters? For example there are about as many 1K users here as there are on Skeptics.SE and Bicycles.SE which graduated long ago. The percentage of 1K users is lower here (only) because there are 3 or 4 times as many users in total (i.e. a much longer tail). \$\endgroup\$ – ChrisW Feb 27 '14 at 17:38
  • 2
    \$\begingroup\$ We check both, really - it's just that the absolute numbers we look for are relatively small. The relative values are regarding the size of the community - more people means there are more voices to be heard and so it's preferable to have more people able to act on those voices. 10-15 people can technically do all the arguing and closing and reopening on a site by themselves, but there's a wide difference between those 10-15 people speaking for a community of 1500 users versus one of 30000 users. \$\endgroup\$ – Grace Note Feb 27 '14 at 17:48
  • 5
    \$\begingroup\$ 1000 reputation might look like an odd threshold to look at, considering it doesn't actually give any actionable privileges. But it's sorta like a litmus test. If enough people are seen hitting that level, they tend to hit the higher levels. Seeing a good percentage in both the 1000 level and the 500 level usually indicates a healthy site that sees a nice number of people rising through the ranks in proportion to the incoming population. \$\endgroup\$ – Grace Note Feb 27 '14 at 17:50
  • 6
    \$\begingroup\$ Does having 10,000+ inactive & long-gone "avid" users (150-200) from the Dark Ages of CR (haven't counted them, really - just assuming) mean we have to compensate with 10,000+ brand new fresh active avid users to make up? How much is the Dark Ages (2011-2012, early 2013) affecting [non-]graduation? We've addressed the unanswered zombies, but we really have no way of addressing zombified users... Yeah, a quick look says we have 30K+ users, but... what's the picture like if we factor out zombie users that haven't visited in, say, a year's time? \$\endgroup\$ – Mathieu Guindon Feb 27 '14 at 18:38
  • 5
    \$\begingroup\$ @MatsMug We check absentee values on all levels across a number of reputation levels. Though you don't have remotely to deal with 10,000 zombies. In terms of avid users, you have 2337 in the bracket from 150 and higher rep. Of them, only 1001 have been seen in the past month, with 98 of those being in the 1k+ bracket. That's basically 10% for the month. If we look at users who visited some time in the past 6 months, it's more like 7% (115 1k+ users out of 1685 qualifying users). We aren't going to hold super-long-gone users against the site's potential. \$\endgroup\$ – Grace Note Feb 27 '14 at 18:57
  • \$\begingroup\$ Great, thanks for clarifying. That was my biggest concern with the numbers coming out of the "Grace Note Query" and us "hardly made a dent" on those numbers recently (although I do see a significant %increase in the 1K cluster since last November - see the yellow chunk); if we have a whole bunch of inactive users counted as "avid", making a remarkable dent is going to take a while! \$\endgroup\$ – Mathieu Guindon Feb 27 '14 at 19:14
  • \$\begingroup\$ Obviously our closing performance will be hit after the rep threshold increase, but, about how much close-voting would you expect for a graduated site? Should obvious off-topic questions be closed rather quickly by the community, or is some delay okay? For us specifically, we should close questions with linked code as quickly as possible. This is because the question, along with the answers, will have to be deleted if the link dies. I personally like to close these quickly. But I also feel that, on a graduated site, moderators shouldn't need to do a lot of closing. \$\endgroup\$ – Jamal Feb 27 '14 at 19:24
  • 2
    \$\begingroup\$ @Jamal There isn't really an "expectation" or anything in terms of how much close voting happens. Technically we want the number to be as low as possible because higher numbers mean too much bad material is showing up on the site. Speed needs to be what's comfortable for the community - you only need speed when there's no need to stop and chat with people about why there's closing happening and such (for example, your given example). Do note, however, that you'll also take a blow to the number of direct editors. That's probably more valuable. \$\endgroup\$ – Grace Note Feb 27 '14 at 19:37
  • \$\begingroup\$ Good to know. Fortunately, it seems that the number of off-topic questions due to misinformed SO users is rather low, and most are from users new to SE who haven't read the About page, and post non-working code or no code at all. If it's just code behind a link, and it otherwise works, we can work with that. Personally, as a very active moderator, I still feel that I don't need to one-hit KO each off-topic question, perhaps unless it's of the type that are just plain low-quality (or has code behind a link). \$\endgroup\$ – Jamal Feb 27 '14 at 20:01
  • 1
    \$\begingroup\$ 200_success' answer about privilege levels has been updated recently. I believe it changes the overall picture. \$\endgroup\$ – Mathieu Guindon Aug 31 '14 at 15:45
22
\$\begingroup\$

Disadvantage: Existing users are probably going to experience a demotion in privilege levels, since graduated sites have higher standard thresholds:

  • 20000: Trusted user
  • 15000: Protect questions
  • 10000: Access to moderator tools
  • 5000: Approve tag wiki edits
  • 3000: Cast close and reopen votes

I think we currently have enough users at the ≥ 5000 level, but I would be concerned about the current population of the top three tiers.

With 7 users above 20000 points, 17 users above 10000 points, and 62 users above 3000 points as of August 2014, I would say that this is no longer a concern. In fact, allowing users to create tags at 150 points instead of 300 is becoming problematic.

\$\endgroup\$
  • 8
    \$\begingroup\$ +1 for a good valid point, but other successful (perhaps smaller) sites are known to have graduated with fewer >5K users than we currently have. \$\endgroup\$ – Mathieu Guindon Feb 26 '14 at 20:14
21
\$\begingroup\$

All the pretty colours .... that are not BETA blue ... Isn't that enough?

Oh, a serious answer:

Pro:

  • easier migration to/from other sites
  • better promotion from other sites
  • no more axe over the head....
  • we get to have a bun-fight each year voting for moderators.
  • pretty colours
  • did I mention pretty colours?

Con:

  • easier migration to/from other sites
  • we get to have a bun-fight each year voting for moderators.
\$\endgroup\$
  • 1
    \$\begingroup\$ Missing Con: pretty colours? \$\endgroup\$ – Mathieu Guindon Feb 26 '14 at 20:05
  • 10
    \$\begingroup\$ I don't see the mod thing as a factor. If you all vote for me as Supreme Mod Dictator, you won't have to worry about recurring elections. Code Review Stack Exchange is best Stack Exchange. \$\endgroup\$ – Jamal Feb 26 '14 at 20:18
  • \$\begingroup\$ Pretty colours? Do you mean a new favicon? I'd love that. \$\endgroup\$ – palacsint Feb 26 '14 at 21:10
  • 5
    \$\begingroup\$ @palacsint that also - plus a whole artwork (background, buttons, badges, tags, etc.) for the site. \$\endgroup\$ – Mathieu Guindon Feb 26 '14 at 21:31
  • 7
    \$\begingroup\$ There really isn't an axe over your heads even now. You'd have to, like, torch everyone on the site and then set exploding bear traps for new visitors in order to drop things low enough to get you guys in any sort of danger zone. \$\endgroup\$ – Grace Note Feb 27 '14 at 14:50
  • 3
    \$\begingroup\$ Well then, no TNT for the monkey... and I was sooo looking forward to that ;-). Good to hear. \$\endgroup\$ – rolfl Feb 27 '14 at 14:54
13
\$\begingroup\$

Disadvantage: Questions can be migrated more freely. If Stack Overflow's off-topic popup lists Code Review as a possible destination for migration, that's going to result in many poor questions being migrated on the grounds that "this code is a mess, send it to Code Review to sort out". We don't want to review code that's a mess, we want code that's as good as the coder can make it.

(Counterpoint to this answer by 200_success.)

\$\endgroup\$
  • 1
    \$\begingroup\$ I think this will be a 1-time process of educating the moderators on SO. Reject, reject, and reject the migrations with educational messages. Remember, we only have to educate the people who press the 'migrate' button, not the people who flag for migration. That's the job of the SO moderators. \$\endgroup\$ – rolfl Feb 27 '14 at 14:18
  • 1
    \$\begingroup\$ For the sites listed on the "Migration" popup (currently meta, superuser, tex, dba and sharepoint) users need 3,000 reputation to vote for migration. There are more than 20,000 such users, so it might take a while to educate them all. (So long as Code Review stays off the "Migration" popup, I agree that we're safe.) \$\endgroup\$ – Gareth Rees Feb 27 '14 at 14:54
  • 2
    \$\begingroup\$ See, SO users can flag for migration, but the SO mods are the ones that pull the trigger... there are only 17 of those... and they are responsible for educating the 20K \$\endgroup\$ – rolfl Feb 27 '14 at 14:56
  • \$\begingroup\$ @rolfl Doesn't matter who's responsible for educating them if we still end up getting the questions that doesn't belong here. \$\endgroup\$ – Simon Forsberg Feb 27 '14 at 17:53
  • \$\begingroup\$ Since the migration list is set manually (and typically only after discussion on meta), you simply need to oppose adding CR to the SO migration list. So I wouldn't count this as a disadvantage. \$\endgroup\$ – CodesInChaos Feb 28 '14 at 10:55
  • \$\begingroup\$ Regarding the "this code is a mess" part, it would really be problematic if some SO users happily migrate such a question that happens to be actual crap. As soon as you start breaking that golden rule of migration, something needs to be changed. \$\endgroup\$ – Jamal Mar 1 '14 at 22:23
9
\$\begingroup\$

Another "advantage" that no one mentioned yet:

Beta sites aren't listed at the bottom of every page on Stack Exchange

Beta sites aren't listed at the bottom of EVERY PAGE ON STACK EXCHANGE!

\$\endgroup\$
8
\$\begingroup\$

Advantage: Questions can be migrated more freely. If Stack Overflow's Off-Topic popup lists Code Review as a possible destination for migration, that's a huge amount of free publicity.

\$\endgroup\$
  • 2
    \$\begingroup\$ I don't see this as a good thing. There are plenty of users already redirect users to CR just because they don't get what CR is about. See this comment. Also, just because we graduate doesn't mean that they will set up a migration path. \$\endgroup\$ – Simon Forsberg Feb 27 '14 at 17:50

You must log in to answer this question.

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