Most Unlikely Couples definition question

In post #803888, an argument came up about the definition of "unlikely". (read the comments for the individual standpoints)

I believe that a couple that is canon, or a fanon couple that is drawn often enough (example: fujiwara_no_mokou mystia_lorelei) is a likely couple.
Azuria believes that this should be based entirely on canon.

BTW, should all posts in the pool at least qualify for one of these tags: couple, yuri, yaoi or something similar for the futanari. I see a lot of posts of two people simply standing next to each other (posing for the camera):
post #797017 - if I didn't know better, they could be mother and daughter taking a bath.
post #635254 - just friends?
post #551558 - just 2 people looking at the camera
etc.

Personally I agree with Azuria. Strictly speaking it should be based on canon.

However, Danbooru being what it is, even though it isn't very likely, Mokou/Mystia is fine the way it is now. So, even though it doesn't fit my definition, some consideration can be given to how popular it is. Besides, I don't think anyone wants to go through and mark 20-30 pages of Mokou/Mystia.

Some leeway should be given the other way, too. Just because a pairing like Meiling/Marisa, Meiling/Reimu or Sanae/Utsuho isn't that popular or common, none are unlikely by any stretch. Or, as I recall, one picture that was marked as unlikely despite the two being from the same house. Not that it'll ever stop someone who thinks otherwise or doesn't know any better.

Frankly, "a fanon couple that is drawn often enough is a likely couple" gives fans justification for any shitty and unlikely pairing they want, but that's for another day.

TLDR time. Personally speaking, canon determines what is and isn't an unlikely pairing. Normally those are called crack pairings in fandom and crack pairings are, by definition, unlikely. That's the way I see it and always took the pool to mean.

As to what makes a couple likely, that depends on who you ask. Coming into contact in canon or at least having the chance to do so is a start, along with possible frequency of encounters, and having something in common.

That's a problem I have with Flandre/Koishi and Remilia/Satori. Both are popular in their own right, but they would never even have a chance to meet anyway. And as I mentioned before, Mokou/Mystia very likely wouldn't work due to a few different things.

As for Mokou/Wriggle, I forgot that firefly may not translate over there, but they're at least in the same game, which is more than most crack pairings get.

On post #551558, the shoujo-y style, the pink glow around them... it all gives the impression of them being a couple.

Another day, another subjective pool gets its argument.

If people want the pool to be of any worth, only the rarest coupling instances should be included in it. The ones that you really didn't expect people would drawn (e.g. post #697576, post #737429), get a surprise, and worth your time checking out the latest addition.

Pairing involving main characters is standard procedure, and I wouldn't add them into the pool in most cases. "Unlikely" they may be, but not "the most".

UnderneathTheWaves said:
As to what makes a couple likely [in canon], that depends on who you ask.

Which is precisely what makes it a bad criterion for pool inclusion. You clearly have strong feelings about what constitutes "likely" with regard to canon evidence. Other people are also going to have strong feelings about that, and their opinions are going to differ from yours. Rather than set up a situation in which you must constantly butt heads with one another, why not use an objective criterion for pool inclusion which allows no room for disagreement?

In my view of things, the purpose of the pool is to help people find pictures of couples who aren't often drawn together. Mokou and Mystia are often drawn together (seven pages, about half of which are romantic/friendly), so they shouldn't be included; Sanae and Utsuho aren't often drawn together (five pictures, of which one or two is romantic/friendly), so they should be included.

S1eth said:
BTW, should all posts in the pool at least qualify for one of these tags: couple, yuri, yaoi or something similar for the futanari.

I wonder. If not, then I suppose we need a "Most Unlikely Friendships" pool for the rare non-romantic pairings, but really, I'm fine either way.

glasnost said:
Which is precisely what makes it a bad criterion for pool inclusion.

No, I do agree to some extent. Above all else I'd rather have one less ambiguous, poorly defined pool.

I would put forth that inclusion in this pool would almost have to necessitate cross-copyright pairings.

Granted, this would nuke whatever outrageous percentage of the pool is made up of Touhou characters, but it would make things much simpler to define.

Also, at this point no pairing between Touhou characters could be considered "unlikely"

Well, this could be fixed by changing the pool to be "Rare Couples" instead.

Or give Touhou pairings its own pool.

reese:

This is problematic, as this action would open the can of worms called "Touhou specific pools for all pools"

reese said:
Or give Touhou pairings its own pool.

touhou yuri and that "everyone is straight" pool ought to be close enough.

While every unlikely couple is not canon, not every not canon couple is unlikely. Enough negatives in there?

As Fence says, this should really be extremely, highly improbable and nonsensical couples. It should be couples where the two characters have minimal or no interaction in the source material and no reason to expect they'd develop a relationship with each other. After all, two people who hate each other are often common fodder for coupling, and thus not rare or unlikely.

Favoring crossover couples helps. If from the same franchise they should be seriously removed in terms of possibility of interacting. Like, uh... Shima Katsuki x Okazaki Naoyuki for Clannad (how's that for a mental image?).