Penguins (Student, Wise, Shaman etc.)

Hello,
During my readings I see that some users hold a Penguin title next to their username : “Student Penguin”, “Wise Penguin”, “Shaman Penguin” and other.

I thought that was perhaps a private joke relying on a profile text option but there is none corresponding.

Searching the topics was informative but these are very old and things may have changed, e.g. Penguins & Penguins..... However, these archeological findings say the titles are only related to the post count (topics created + replies sent) (source). The list was something like that :

Puzzled Penguin 0-49 (new users start with this level)
Student Penguin 50-99
Explorer Penguin 100-249
Busy Penguin 250-499
Parent Penguin 500-999
Wise Penguin 1000-1999
Shaman Penguin 2000-2999
Omniscient Penguin 3000

At this point, scientific method leads me to the user list (to order by decreasing post count) :slightly_smiling_face:

(bad shortcut, this was send too quick : the end of the message is in writing, sorry !)

2 Likes

This is something that was carried over from our previous platform, which assigned these titles automatically. Discourse doesn’t seem to have the ability to do that (or we haven’t looked at setting it up yet).

But that’s the history and where it comes from. :slight_smile:

My conclusions at list analysis were the system changed (maybe some new penguins like “New or Quiet Penguin” and “Flux Capacitor Penguin” beyond 4k posts) but probably was down because some transitions don’t fit the thresholds and some people don’t have any title.

So, thanks @hendersj for confirmation ! You are too fast for me :+1:

Too bad this is broken ! Maybe something counting not only the posts but also likes for example ? That would be great to discourage flood but reward regular and liked helpers. I like things that gamify the experience :slight_smile:

And here I thought I had to grind away with posts and likes to get a title … jk :stuck_out_tongue:

Technically, it’s not “broken”, just a different system. Discourse uses something called “trust levels”, but looking ahead at the 3.0 release, they’re making some changes to that so it’s more group-focused. As individuals participate more (or less) over time the trust level changes to reflect “engagement” with the community. :slight_smile:

Yes, I know about the trust levels but the granularity is very poor.
Basically, if you don’t have much time every month to regularly participate at high level (come every 2 days max, read 50% of topics, make 10 replies), you are stuck (or retrograded) to simple member, no matter you were a “Shaman” a month ago or someone contributions can be trusted for long.
For me, this system is as imperfect as single ranks are. Combining the two could be better : trust levels & total posts and likes rewarded. :slight_smile: What’s your thoughts ?

it’s more group-focused

And groups can be used as title.

1 Like

I think the poor granularity might be why they’re changing trust levels to groups in 3.0.

As a community grows and evolves, people come and go - but the community still evolves. So being around for a long time and then leaving for a while means you’re not necessarily in touch with the community when you come back. It makes sense for the current members of the community to know that those who are identified as being active (which is really what trust levels help measure) are current and active.

I don’t like thinking in terms of “engagement”, but there is an argument to be made for its importance in community health.

No system is going to be perfect, but I personally think (not speaking for the rest of the forums team here :slight_smile: ) that current active membership is something that can be extremely valuable for someone evaluating the quality of answer they get. That doesn’t mean the most active or longest-serving members are going to be giving perfect answers every time (or even ever, for that matter).

Maybe a measurement based on those whose answers are most often selected as the solution would be beneficial as well - we introduced the “solved” extension a little while ago - not long enough for it to have been used enough to give a good, consistent measure of “answer trustworthiness”, but over time, that will become, I think, a key measurement that can be used to highlight members who are helpful.

1 Like

To complete, I’m not a Discourse specialist at all and I don’t know about your constraints with this forum so this is just a very humble proposal/contribution.

  1. Automatically granted badges with, if necessary, custom SQL rules as criterias to achieve. Some commons requests are provided as examples.
    A badge can be used as a title (but I think this is on a user choice basis).

  2. Automatically populated groups. This is possible on the basis of badges. Source : this and that, see “User Group Membership through Badge”. Then, groupe used a title.

My small stone, if it can help :slight_smile:

We do try to keep the number of plugins and customizations to a minimum, just to help ensure that we’re not going to be dealing with issues during upgrades. We’ve added a few plugins recently, and as we see how upgradability is affected, we may look at doing more customization.

I have 3 sandbox environments of my own that I play around with and test things in - so I’ve looked at some of this, but not all of it (time is a finite resource, after all :wink: )

That said, the input is greatly appreciated. It’s useful to know what the community might find informative or useful, even if it’s not something we have the “bandwidth” to implement.

Then I can throw in my opinion that for me it is al humbug. :wink:
E.g. making 3000 useless posts in Soapbox will give a person some rating. I really do not hope that others think this will guarantee technical skills to get help. :laughing:

And giving technical help is what these forums are about.

@hendersj Sorry, we posted almost in same time.
Thanks for this long answer on your feeling about “measuring” and “valorize” (good) participation !

I think the poor granularity might be why they’re changing trust levels to groups in 3.0.

Oh, nice, I didn’t understand that change was coming. I thought that was current behaviour with the 4 levels.

but the community still evolves. So being around for a long time and then leaving for a while means you’re not necessarily in touch with the community when you come back. It makes sense for the current members of the community to know that those who are identified as being active (which is really what trust levels help measure) are current and active.

Yes, I totally agree with that. But in the mean time, it’s valuable for all the communuty to identify the expected “quality” of a answer given by a “elder”, even if his trepident life doesn’t allow him to come often. The current levels 3 and 4 only render the information of the actual participation (which can also be faked by free time to read and post in the “open chat”).

Maybe a measurement based on those whose answers are most often selected as the solution would be beneficial as well

Yes, this is a good point and the reason why I also mentioned the likes. There are much difficult to “fake” than a post count taken alone.

we introduced the “solved” extension a little while ago - not long enough for it to have been used enough to give a good, consistent measure of “answer trustworthiness”, but over time, that will become, I think, a key measurement that can be used to highlight members who are helpful.

I wish you it will work but in my (professional) experience with this kind of functionality relying on an active intervention of the user receiving help, this never takes to a proportion that allows to measure efficiency at global level. But, even if not, that’s good and time saver for other people searching for the same problem.

E.g. making 3000 useless posts in Soapbox will give a person some rating

@hcvv Ahahah, yes of course ! :grin: For this, again, likes “lies” less and if necessary, excluding the Soapbox of the counts is possible with the good request.

Of course, no method is perfect, and there are always ways to game the system, regardless of what the system is.

1 Like

Thanks for your messages, I really appreciate !

We do try to keep the number of plugins and customizations to a minimum, just to help ensure that we’re not going to be dealing with issues during upgrades.

I understand this and laught at myself because this is the kind of answer I sometime give (at work) too.

so I’ve looked at some of this, but not all of it (time is a finite resource, after all
or the time/hands to do

Same here, I thought so, that’s why I searched for the links. If one day next you find time & interest for that and need a basic user to test, you can ping me.

Meantime, I stop filling this topic and let potential other interested user give his/her opinion :dotted_line_face:

1 Like

No worries. :slight_smile: I’m always glad to have a productive discussion about things like this. :slight_smile:

Yeah, in the administrative interface, I see frequent reminders that changes are coming to the trust system. It’s not been implemented as ‘groups’ up to this point, and that makes it something of a ‘special’ system when it comes to doing things. 3.0 is still a ways off (and I’m going from memory about when it’s being introduced…actually, let me check…

It was announced as coming in October 2023, but I have the version number wrong - it’s 3.3 that’s getting it, not 3.0 (we’re on 3.2.x right now, and yeah, various site settings that relate to trust levels have notes added that say that they’re being removed in 3.3.

The old way it was done was based on posts only. As noted, it’s easy to ‘game’ it if you want, whereas having to ‘game’ it over time in order to maintain the ‘status’ takes a lot more effort, and ultimately can weed out those trying to game the system.

Similarly, with as fast as openSUSE changes, I’m not sure that someone who posted 3,000 messages 10 years ago (these forums have a history of messages going back about 16 years IIRC) doesn’t really mean that their advice may be useful (it also doesn’t mean that it isn’t useful either, to be fair), but for members weighing answers received and whether or not they’re useful, it’s one of many indicators they may be able to see in order to help make that determination.

Yes in both cases. Users can’t like their own posts, nor can they identify their own “solution” as a solution in a topic they didn’t start (if one starts a topic and then comes up with an answer, they can reply with their solution and make it a solution). Only the user opened the topic (and staff) can identify a solution in a topic. So those become much more difficult (but as all things of this nature, not impossible) to game in order to raise one’s reputation artificially.

It’s actually something that we’ve been persistently asked for for years - the old vBulletin forum didn’t have that functionality natively, and we were skeptical of the value (because one person’s solution isn’t necessarily going to be the solution for another, vaguely similar issue), but I was persuaded that there were other benefits to having this functionality (in particular - those looking to help could filter out already-solved issues, allowing them to help more people who were still looking for help).

I think you misunderstood me. I do not object to the impreciseness, or the not-perfectness of the system, I object to it (whatever the goal may be) being there at all.

1 Like

No, I picked up on your subtlety there. :wink:

So do I.

“Rating” individual forum members can never be done in a reliable way. So “rating” will only discourage those who feel “rated incorrectly”.

To give some reasons why I think that “rating” forum members reliably is impossible:

  • The matters discussed in this forum are subject to constant changes. No one will have an “up-to-date” knowledge at every point of time.
  • No one is an expert in all areas.
  • Some members can spent more time on the forum than others. However the input created by the latter might still be valuable.
2 Likes

Exactly. Some members will have plenty of time to be on the forum and others will only have a little time to spend. Forum ‘rating’ systems are biased and therefore not trustworthy.

2 Likes

Some people find it useful; if you don’t, that’s fine - you don’t have to use it. :slight_smile:

Well, that is only one half of the story.

The other half is all those who get judged wrongly by those who trust the rating.

3 Likes