[Discussion] A 9kcon Retrospective

Summary:

i know i'm a dumbass, let's try to make it better next time

[Discussion] A 9kcon Retrospective

Naepic 19 Oct 2025 17:56

04: https://scp-wiki.wikidot.com/forum/t-17366073/discussion-a-9kcon-retrospective

The SCP-9000 contest has ended, and with it, I'd like to hear people's thoughts on their experience with said contest, what worked, what didn't, and what we should do going forward.

Side note, I want to thank Disc for their support this contest - even if ultimately we did not find many suspicious votes this time around, and especially none that could alter entire placements like previous contests, their presence was comforting as opposed to having a small group of people shouldering everything.

A lot of things were announced late, mainly because I'm a dumbass: from forgetting to inform people immediately that 9999 was not available due to being placeholdered, the late addition of the ruling to not honor slot placements made within the last 24 hours of the voting period, to even just the chaos regarding slot placements after the voting ended.

But enough from me. I want to hear what you people thought of the contest and how things can improve going forward

Last edited on 19 Oct 2025 17:57 by Naepic Show more

Re: [Discussion] A 9kcon Retrospective

EstrellaYoshte 19 Oct 2025 18:09

I’ll flush out my other thoughts later, but the first thing to come to mind is that if Contest has a docket, add a post-reminder to add to the Contest Archive when everything’s finished.

Re: [Discussion] A 9kcon Retrospective

radian628 19 Oct 2025 20:37

From my point of view as a contest entrant, I thought things went largely smoothly. Creating and uploading an entry was a simple enough process for my team. While I do think the rules for last-minute slot changes could have been communicated more clearly, I think this problem will solve itself moving forward, since, in future contests, entrants will ideally know that these rules exist from the very beginning, instead of them needing to be introduced halfway through as a stopgap measure.

All this said, as the person who recorded everyone's slot preferences and calculated who would get what slot, that process was significantly bumpier. I made a few mistakes in recording users' slot preferences, and, as a result, several slot picks were initially incorrect. Consequently, staff had to temporarily prohibit 9kcon entrants from moving their articles to their new slots, as those slots might have changed as corrections came in.

All in all, I think this problem could be solved if slot preference picking was (1) easier to automate, and (2) provided a mechanism that proactively informs users of what the contest team "thinks" their slot preferences are, so they have the opportunity to make corrections ahead-of-time. To this end, here are my ideas:

Slot preferences should have a consistent format that can be easily picked up by an automated scraping tool. This will save time and reduce the likelihood of data entry errors. To be clear, it doesn't have to be fancy, and it doesn't (and, in fact, probably shouldn't) have to be specific enough to enable total automation. Something as simple as "put all your slot preferences in a blockquote" would suffice, as that would already save me the extra hassle of searching through authors' comments manually to find slot preferences within them. Even if an entrant doesn't follow this format, it isn't a big deal, because I can easily write a script to identify the few contest entries that didn't have properly-formatted slots pickings, and just handle those cases manually.

A contest staff member should send a Wikidot DM to all entrants confirming that they've received their slot preferences. My thinking is that once everyone's slot preferences are identified, I could write a second script that DMs every contest entrant with what the contest team thinks their slot preferences are. The DM would also say something to the effect of "If these slot preferences look correct, do not reply to this DM. If they look incorrect, please send (POINT OF CONTACT) your correct slot preferences." In terms of feasibility, this would require minimal effort on the part of the person actually sending these DMs — they'd just have to plop a script in their console, and it would send all of them automatically.

Entrants should be required to notify Contest Team staff of slot preference changes for future contests. To be clear, they can just put their initial picks in an author post comment, but if they update those picks after the aforementioned confirmation DMs are sent, they should explicitly inform a member of the Contests Team of their changes. I think it'd also be good to have a common-sense "you can't change slot picks in the last X hours of the contest" rule as well, purely for logistical reasons (we don't want to have to take on dozens of slot changes requests in the final few seconds of the contest while ten other things are going on at once!). In case entrants forget these rules, there's no harm in us also including them in the automated DMs.

Overall, despite the above issues, I think 9kcon went quite smoothly.

Last edited on 19 Oct 2025 20:38 by radian628 Show more

Re: [Discussion] A 9kcon Retrospective

Jerden 20 Oct 2025 10:10

It would have made our job as staff easier if people had been able to move their own slots so I think communicating slot allocations but telling people not to move their articles for 24 hours while we confirm things would have been better than what actually happened this time, hopefully a standard format will avoid this kind of confusion in future. The 24 hours would also be a good time to draft up the new series page, as if you don't do that first it's quite easy to lose titles. I'm not sure how someone managed to delete their own article but I think most people can be trusted to do it correctly, and then staff can handle the rest if people are unsure.

A minor thing we should recommend people don't do is put the full article URL in image links - if its localised to the page (as it should be,) then just the Filename is fine, this does get more complex with listpages offsets though. I think we should clearly communicate that if people are not going to move their own article we will just change the URL and find+replace "9000" with the new slot number - I was trying to just do "SCP-9000" at first but there were too many times when it was just the number (e.g. ACS headers)

Listpages are annoying to migrate but I don't think there's any way to streamline that process, as long as they are tagged correctly as _listpages we can just make sure that they're done carefully first so that everything else can be done quickly.

Just pragmatically I found it was easiest to just go through all the contest entries at the end to add and amend all the Wikiwalk footers, it's a real hassle to try to do that during the move and I kept forgetting to check for it before moving on to the next one.

Last edited on 20 Oct 2025 10:12 by Jerden Show more

Re: [Discussion] A 9kcon Retrospective

radian628 20 Oct 2025 17:37

It would have made our job as staff easier if people had been able to move their own slots so I think communicating slot allocations but telling people not to move their articles for 24 hours while we confirm things would have been better than what actually happened this time, hopefully a standard format will avoid this kind of confusion in future.

Agreed. That said, if it weren't for the fact that there were a few data entry errors in users' recorded slot preferences, we could even tell them to move their slots sooner — I wouldn't be surprised if some people would forget to follow the 24-hour waiting period. This is why I think it'd be good to have a system in place to work out slot preference errors ahead-of-time.

I'm not sure how someone managed to delete their own article but I think most people can be trusted to do it correctly

While I might be overcomplicating something that happens too rarely to warrant specific code, in theory we could add the following CSS to the mainsite, which will cause extra text to appear when they attempt to delete a page, telling them to just use the "Rename" option instead of deleting&reposting.

#action-area:has(#rename-option-delete):not(:has(em)) h1::after {
  content: "Warning: If you just want to rename the page, you do NOT need to delete and repost it. Instead, use the 'Rename' option.";
  display: block;
  font-size: 1rem;
  margin-top: 1lh;
  font-family: InterVariable,RedactRect,Inter,BlinkMacSystemFont,'Segoe UI',Roboto,Oxygen,Ubuntu,Cantarell,'Fira Sans','Droid Sans','Helvetica Neue',sans-serif;
}

Re: [Discussion] A 9kcon Retrospective

aismallard 11 Nov 2025 09:45

I absolutely do agree that we should have a standard slot preference format, but the fundamental issue is that unless it can be automatically enforced / verified from the outset of the contest, there will be no standard format. There are reasonable patterns like "a 9X00 article" and then once there's free text you inevitably get complicated or ambiguous human-written descriptions.

My proposal is that the next K-con should have an HTML utility that uses a standard format which can output a list of preferred slots. This allows contest runners to use the same tool to automatically assign all the slots based on contest order and avoids complications like needing to backtrack / recalculate slots if something needed to be fixed.

I developed a grammar which I think is suitable for providing good flexibility but without making it Turing-complete (as to avoid undecidability and arbitrarily-complicated rules), I think something like that could form the basis of such a utility.

Re: [Discussion] A 9kcon Retrospective

S D Locke 20 Oct 2025 00:01

Without any major shenanigans this is one of the smoother kcons I’ve witnessed.

Unrelated to the direct running of contests but I wonder what more could be done to encourage users to read and engage with more works with a more open mind, as, though rarer than in the past, we’ve still users (on the newer/casual side) who comment to the effect of supporting/not supporting an entry for the slot itself — even in cases where it’s wholly nonsensical (not wanting to positively rate a work in the low tens while a dozen entries were 1-300 past the low-ranking submissions).

This obviously doesn’t include the unknown number of users who feel/prefer to vote this way but simply don’t broadcast it.

Though a bit bizarre, it seems that a non-zero number of readers consider the rating they can give out as the only one they have¹² — anything to disabuse casual readers of this notion might (hopefully!) allow more entries to receive a higher degree of recognition.

Warnings on the contest page of DQs are effective for entrants, and to possibly ward off the most egregious of downvoters, but I’m hoping some form of positive reinforcement/encouragement to support any and all entries one enjoys, rather than just their favored, might help casual readers to be more generous and not zero in on only the top couple of entries.

Given IO already promotes every entry as best as they can, I’m unsure what else could be done in this regard, but I’d appreciate any ideas to shift engagement to a wider array of entries than seems to usually be the case.

One possibility might be community posts that encourage users (on and offsite) to pick out and discuss their favorite five entries? Something that openly encourages the support of multiple entrants rather than favoring one?

[Finally, this is all obviously anecdotal it would take a ton of work and scpper being functional to even have a chance at determining how many users vote on a single/only top few entries and if this is even an actual ‘problem’.]

Footnotes

1. or alternatively, that each vote constitutes a direct vote for the winner itself. (Which yea sure it does in a way but only insignificantly, given it’s determined through aggregate)

2. ETA: of course there’s technically nothing wrong with only casting a single upvote across all entries so long as one isn’t engaging with others maliciously. I just think anything that could help bring more attention to a wider array of works would be a boon to all those who put a lot of effort into these competitions but for one reason or another do not see much in the way of attention/reads/votes/etc.

Last edited on 20 Oct 2025 00:07 by S D Locke Show more

Re: [Discussion] A 9kcon Retrospective

Sinking_Otter 20 Oct 2025 16:35

Speaking on the entry promos, I think the largest issue in ensuring more equal engagement across a wider array of entries is the sheer number of entries. During this contest I scheduled the promos to go out every four hours on Facebook and Instagram and even with this rapid pace, the promos barely finished before the voting period ended. Had there been maybe 10 or 15 more entries, there would've been promos being posted after the contest finished, which is something I think should be avoided to ensure all possible votes generated by the promos are counted in the contest.

Another issue is promo burnout. Taking a look at the data on Instagram, of the top 10 best performing promos posts, only one was published outside the first two weeks of promos running. Expanding that to the top 20, and the number only increases to three. Anecdotally, I've also noticed a drop in the number of views¹ and interactions² promos have been receiving, especially in the final week of promos going live.

I do like the idea of encouraging the community who engages with the socials to share their favourite entries. However, I worry that it may significantly cut into time for the promos, and potentially lead to a scenario were some promos are posted outside the voting period. Another issue is these posts might end up getting comments featuring for the most part works that already have a significant amount of attention. Many people generally only read the top entrants, and thus if asked we may see the same few articles being mentioned, which could come across as unfair and extra promo for the articles that don't need it as much. Therefore, if this kind of post is to work, we'd need to find a way to have a wide-range of articles mentioned in the comments.

Given all this, I think the best option available to help ensure more equal engagement is to offer a guide that showcases promos that performed well in previous contests. Even though there are a lot of factors that impact a promos performance, looking at the worst and best performing promos, it's clear those with custom posters/art/images generally faired far better than those that were simply a picture from the article or a place like Wikimedia commons. Having a selection of these promos in a guide can help provide inspiration so users can have an easier time crafting their promos. Further, the guide can include important information like image sizing for platforms, what is required for promo posts and a set of rules for promos to ensure no platform guidelines are violated.

I'd also recommend that when promos are going to happen, users be made aware via an announcement of some kind in advance. There were a lot of users who I contacted that either never responded or weren't prepared with anything. Letting people know in advance gives them the time to prepare promo materials and helps spread awareness that they're happening, which may also help with engagement.

Also, to be clear, my above suggestions only apply when promo materials are being sourced directly from entrants, rather than being constructed directly from contest entries by the IO team, which I believe has happened previously.

Footnotes

1. The number of times a post is seen. Views aren't unique, and if a person sees a post multiple times, it will count as multiple views.

2. The number of times a post is interacted with via likes, shares, comments, shares or reactions.

Re: [Discussion] A 9kcon Retrospective

S D Locke 20 Oct 2025 17:47

This is quite helpful and fascinating. The promos themselves seeing less engagement as time wears on does track with entires themselves— with those posted early on (for the most part) receiving more engagement than those posted in the middle/end of the posting period.

I wonder if it might be worth it to try out a different manner of sharing: to limit promos to the first week (or so) of the voting period, but to promote different articles on different platforms. Rather than say, 160 insta posts, 160 twitter posts, etc, 40 entries get shared on insta, a different 40 on twitter, yet a third batch shared on facebook. Of course each platform has a different level of engagement so it’s an imperfect solution — and of course coordination has an additional level of complexity, but perhaps it’d help keep any one audience from experiencing burnout/disinterest as time wears on.

A standardized guide would be quite useful to have handy for contests moving forward, hopefully easing a bit of a burden for IO having to explain each time how authors should construct promos/who to send to, etc. New contests could have a simple bullet point linking the guide as a reminder, and naming who would be the recipient of promos when authors are ready to send.

*Any* sort of guide in this vein would be nice, but additionally using to instruct people how to be effective with their material is an insight I’d not considered, but seems like a fine idea.

I do like the idea of encouraging the community who engages with the socials to share their favourite entries.

My thinking was to have these kinds of posts/conversations not necessarily in the midst of the posting/voting period, but even before. During announcements, having follow-up posts to instruct casual fans on engaging with and rating every entry they enjoy — sort of a pre-emptive conditioning to encourage people to not think too much on the weight of their votes and dig into the breadth of writing the authorbase has to offer.

Another issue is these posts might end up getting comments featuring for the most part works that already have a significant amount of attention.

Ah, yea of course. One possibility could be encouraging people to share their favorite *underrated* entries, but admittedly trying to get people discussing entrants has the same issue of further propping up those already getting the most exposure. Drumming up excitement for something already under the radar is no easy task.

ETA: Going over my own promos seems to track well with what you said. Post interactions seem to correlate with an interesting, unique visual. Those with the most interactions seemed to be both my own, the one I did for queerious, and a retweet of parenthesis (fascinatingly, this one outpaced a dozen I either drafted myself for others, or those that I asked people to give me their own promos — despite parenthesis having a tenth of my followers. I’d have to count this due to their incredible artwork, composed of four separate images that made a singular image, but could be expanded to reveal full images with additional detail. This goes to show that a little creative touch and careful planning can pay off in a big way even considering the limitations of the platform.).

Additionally, far and away what seemed the most successful aside from posts about our own entry was what I’d shared for Dysadron’s entry: consisting of a screencap of my on-site comment and a simple order for people to read it.

Diminishing returns, as a same-day sharing of another entry in the same fashion did not see quite the same attention as the first. Only the promo for Queerious’ entry matched/surpassed Dys’ in terms of retweets and likes respectively, but the hers was already a top-performing submission at the time of promo, while the Dys’ had only been up for roughly a day if I’m not mistaken.

TLDR: in lieu of an author needing to construct an eye-catching visual, a strongly-worded recommendation may suffice. Unsure how the logistics of that should work (both because this style of “this is what people are saying” promo was disallowed in certain areas, and because to *get* a strong recommendation, people obviously need to have read it to begin with), but perhaps useful to note as not everyone has the same flair for image creation.

Last edited on 20 Oct 2025 18:18 by S D Locke Show more

Re: [Discussion] A 9kcon Retrospective

Sinking_Otter 21 Oct 2025 08:18

I wonder if it might be worth it to try out a different manner of sharing: to limit promos to the first week (or so) of the voting period, but to promote different articles on different platforms. Rather than say, 160 insta posts, 160 twitter posts, etc, 40 entries get shared on insta, a different 40 on twitter, yet a third batch shared on Facebook.

I think this would be interesting to attempt to see the results and if it's viable, though as you mention there is a difference in engagement levels between platforms. In fact, I worry the difference may be too great. Looking at Instagram, the best performing promo received 41.6k views and 6.2k interactions. That same promo received only 3.4k views and 32 interactions on Facebook. The second best performing promo faired a little better, with 24.6k views and 3.1k interactions on Instagram versus 11.6k views and 241 interactions on Facebook. This is a trend that all the other promos followed.

Given this, if we do post some promos on Facebook, some on Instagram, etc, we may end up with a scenario were some promos will never get the same number of views/interactions as others purely because of the platform they're posted on. So too might it seem unfair to entrants whose promos end up on platforms that see fewer views/interactions, even if where promos are selected to go is completely randomised.

A standardized guide would be quite useful to have handy for contests moving forward, hopefully easing a bit of a burden for IO having to explain each time how authors should construct promos/who to send to, etc. New contests could have a simple bullet point linking the guide as a reminder, and naming who would be the recipient of promos when authors are ready to send.

I'd need to discuss it with the IO team, but I'd be happy to make such a guide. Like you said, it can be used for contests moving forward and it can also be used by authors outside of contests to construct their own promos to post on their own socials and placed like the SCP Wiki reddit page. It could also potentially open the door to a system were authors can request promos for their works on the socials outside of contests. Though such a system would require a lot of discussion before it's even close to getting off the ground.

My thinking was to have these kinds of posts/conversations not necessarily in the midst of the posting/voting period, but even before. During announcements, having follow-up posts to instruct casual fans on engaging with and rating every entry they enjoy — sort of a pre-emptive conditioning to encourage people to not think too much on the weight of their votes and dig into the breadth of writing the authorbase has to offer.

That's something that is certainly doable and I think could definitely get discussion going. I do think as well that posts along this vein will help with what I believe is a clear disconnect that a subset of people who engage with the socials have with how the Wiki operates. For example, on the announcement that Trench won the SCP-9000 contest a number of people commented expressing disappointment that articles they liked from the contest would now be deleted because they didn't win. Of course this wasn't the case at all, but it shows that there is a gap in communication that imo needs to be addressed. Posts such as the one you suggested can help close that gap, as can things like contest-specific faqs posts and encouraging users to sign up to the Wiki, or at least read things like contests pages so they understand what's happening.

Additionally, far and away what seemed the most successful aside from posts about our own entry was what I’d shared for Dysadron’s entry: consisting of a screencap of my on-site comment and a simple order for people to read it.

I wonder if part of it's success came from the fact you posted it on your own social page, and it's a recommendation from you and if the same success would be seen on the Wiki socials if the same post had been published there. I'd of course be willing to try it on the Wiki socials (so long as it's allowed within contest rules as you say), but outside of recommendations from very popular/well-known authors, I feel like their success would be limited. But I cannot be sure of that without some test posts, and I do agree it'd be a good solution for those with limited design experience or without access to resources to create custom promo images.