Let's shoot down the Fastest Swag-Hunter in the West

Question

The Bad Guy

If you have competed in one of the Stack Exchange competitions which run here on Meta from time to time, you know all about the Fastest Gun in the West Problem.

This problem is particularly pronounced in swag-yielding competitions, because entries are neither right nor wrong; their success is partly due to how punny they are, or how they are bursting-at-the-seams-with-memes. Oh yes, there are also sometimes really great answers. The other part is due to how much exposure they get. From personal experience I have found that once there are 50+ answers, there is just about no chance for an answer to reach one of the top 25 spots.

The worst part is that this encourages low-effort speedy entries, ruining the chances for entries which take time to create.

---

An idea which could help somewhat:

Announce the contest a few days before it actually opens

This will have numerous advantages:

• Giving everyone time to find out about the contest before it opens.

• Actively discouraging low-effort entries while actively encouraging great entries; the biggest gain. This will happen because everyone will know that there has been time for others to create great entries. To compete, their entry has to rely on its worth alone, not on speed.

• Giving answers an equal chance. Having lots of entries at a similar score as once will encourage users to look through the answers before just voting on high-scoring ones.

Edit: let me clarify my intent.

I am not trying to solve the Fastest Gun in the West problem. I merely propose an idea which will positively impact these issues, one which can be implemented extremely easily. Yes it has its weaknesses, but don't the benefits outweigh them?

---

How?

The contest question could be posted and then locked to prevent having answers posted...

Your ideas on how to implement this?

Answer 1

I too have noticed something similar to what you are describing, to the point that as some may remember I was already discussing about it in the Tavern chat room some days ago.

Anyway, let me start with a little clarification: I don't think we are fighting time here - the real problem actually seems to be short attention span from the potential voters base.

Swag contest attract two kind of users: the ones who want to try to win something and therefore will actively contribute with an answers and the "watchers", those who only browse the already written content because they hope to find something interesting. Now, as sad as it may sound, I think that the first category is the less likely to actually vote for competing answers (yep, I know, there are some users that will still vote competitors, but I fear most won't), so we should be able to just concentrate on the second category instead.

"Viewers" of the contest are usually attracted to the post soon after the contest is announced. So, what I have actually observed is that the contest will be "forgotten" after a few days: as soon as the initial "oh, shinny, something new and funny" effect is gone, the viewers will start to lose interest. In my humble opinion this is true for every kind of events - not just swag give-away: even during Winter Bash many users lose interest very soon (mostly as soon as all the hats have been "seen", so there is nothing new to see).

So, here is a first element: Surprise. As soon as the surprise is gone, so are most of the voters.

But this still doesn't really reply. So, why would the first answers have an advantage? As we just saw, it may be due to the fact that "came first == had more time to be voted before the question is forgotten", but that probably isn't the full picture.

So, there must be something else: the short attention span. As the post gains more answers, the newer one inevitably fall off the front page. The problem is that many potential voters will just quickly browse the first pages and just skip the rest.

Basically, we have three issues:

1. Interest in the contest will fall rapidly after the first days (unless an external force comes into play, for example a bounty). This advantages answers that came before the drop.
2. The longer the answer, the bigger the chance the reader will stop reading halfway: many potential voters don't want to put much work in reading answers, so post that requires more effort to read (longer posts, post that require running a script etc) will be more easily ignored. This advantages shorter more readable posts, which are often perceived as low effort (and maybe some times actually are). Longer answers that required a lot of work have actually a disadvantage here.
   3. Since all new answers start at 0, answers that started on the first page are advantaged. People will stop reading long before reaching the last page, so the post on the first few pages are advantaged.

All of this brings to the logical conclusion that there are two keys factors: precedence and readability. Often both are related to the First Answer problem, but not necessarily, as some factors may easily twist them (for example, I got many votes on my answer in the last contest after a certain parrot posted a bounty with a link to said answer).

So, after all of this... what can actually be done. I fear that the answer is "very little". As we have seen, the real problem is that the potential voter will (most of the time) get bored before reading all the answers: he/she is there only to look for a quick laugh, not to "work" on an election. As such, it doesn't really matter what you do: if there are a lot of answers, the display order will advantage the first ones.

Forcing the answers to appear at the same time could help, but would require a lot of work since not even the moderator election workflow seems to work that way (basically, you would need a question that accepts answers but keep them invisible to others before a certain date). Also, even if all the answers would be posted at the same time... you would still need to randomly order them differently for each potential voter so that there is no "first/last answer in the list" (in order to compensate for voters only reading the first posts and not bothering to read **all* answers).
I doubt that any of the existing workflows can be reworked to behave this way, and I also doubt it will be worth the effort to develop a new custom one just to handle swags giveout.

So, in conclusion, while I kinda agree with the general idea that an issue exist I fear that unless someone finds a better solution we will have to live with the current rules and try to use that at our advantage. The last two contest taught me that is often better to provide a fast-to-read answer than a longer one (or one that requires you to return each day, like an advent calendar...), so I suggest you keep that in mind whenever the next opportunity to win a Stack Exchange memento arises...

Answer 2

From personal experience I have found that once there are 50+ answers, there is just about no chance for an answer to reach one of the top 25 spots.

I've seen the contrary happen a lot too, especially to quality answers. In last year's contest, I posted my answer as the 34th entry. I ended up well in the top 10. This answer was posted around the 50 answer mark you mention, and ended up second. I know the knitting post from this year isn't a 'contest', but it's a fun post too. I uploaded my ugly sweater(s) as the 64th answer. They're now 'in fourth place'.

The current nr. 1 entry for the 'Time for more swag' contest was posted a day after it had started, as the 60-something answer. The current 4th answer was posted another day later (I'm going to spare myself the effort of counting through the timeline on that post again), and the answer now in 11th place was also posted after the 50 answer mark you mention.

The worst part is that this encourages low-effort speedy entries, ruining the chances for entries which take time to create.

First of all, low-effort doesn't mean bad. These contests are supposed to be fun, and saying someone shouldn't win because their answer was 'made' and posted within a few hours is not really nice towards the creativity and thoughts these people put into their entries. What does 'low-effort' mean? Sometimes, these entries can be just as fun, and fun is what the contest is about, not the amount of hours you spent creating something.

While there are quite a few speedy entries on the latest contest, I feel most quick entries are just as creative or fun as some later posted entries. I like watching plushies time-travel in a cardboard box, even if that didn't take days to create. In fact, when it came the 'Time for more swag' contest, I felt most of the good entries were made quickly, and there was a flood of one-liners, meme's, similar ideas, etc. only after those posts were already made.

I agree that those do clutter up the contest and that it's probably frustrating for that one person that does put time and effort in their answer, that after a few days still manages to upload an original, well-executed idea, only to have it drowned out by 4 pages of those kinds of answers. So there might be something to gain from e.g. random sorting, or showing the newest answers first (not active, newest. Edits don't count for more visibility).

Actively discouraging low-effort entries while actively encouraging great entries; the biggest gain. This will happen because everyone will know that there has been time for others to create great entries. To compete, their entry has to rely on its worth alone, not on speed.

Given the above, and my general faith in humanity, I don't really see how announcing the contest beforehand will stop answers from pouring in as soon as it becomes known there will be a contest. You're always going to have people in other time zones, are they going to have to stay up late to upload their answer ASAP?

An announcement isn't going to solve this FGITW problem you see, it only gives people a few days to prepare stuff, to prepare the same post they would've posted anyways. This includes meme's or one-liner jokes, or whatever else people consider 'low-effort, low quality'.

An announcement won't stop answers from cluttering things up, and I don't think it will lead to higher quality entries. A fun post that can be made in a few hours, is a fun post that can be made in a few hours. And someone that wants to upload a meme or one-liner joke or photoshopped whatever is going to do it anyways.

---

Giving answers an equal chance. Having lots of entries at a similar score as once will encourage users to look through the answers before just voting on high-scoring ones.

If you ask me, an announcement won't solve this enough. Having 80 answers in 2 hours is just as bad as having 80 answers over the course of two days: People that aren't interested aren't going to sift through it to see if there's something they like there. They'll see a few answers on the front-page, drop a few votes, and drop out again.

Even e.g. delaying the voting on answers until a contest is 'opened' (a bit like elections, only voting after having all nominations) and then displaying posts in random order isn't going to guarantee each post equal attention, in my opinion. 150 answers to a single swag contest is just too much, not many people will have the time and dedication to spend that much time reviewing each answer and judging it for how creative and fun it is.

Answer 3

I really like this idea, opening the contest a few days after announcing it.

I would propose that the answers should be sorted randomly every time a user opens the page (like elections) .

Optional: If they'd like they can choose to sort by time/votes etc.