3 months late but whatever
 
on the explicit v safe argument, you have to remember, the only way to explicit material to be useable is to be reasonably good. i just did a quick search, the first page of explicit, has only 4 objectively bad images, not content wise, but how bad it was drawn or colored, and i believe that is 50 images. you also had 5 images that were recolors of the same base image, and 4 more that were variations on one image, if you want to count them as something so be it.  
now safe, i’m going to discount the comic that just got dumped and ended up there as it will skew the results, there are 14… but on top of that, you also have 10 fairly generic vectors, 3 where you have to ask “why did you chose to just draw that?” you have more than a few that out of the context of why they were created don’t make sense, and a massive random compilation image.
 
here lets add one more filter to it, it must have 20 upvotes, ill do the same for both safe and explicit, still filting out the comic dump.
 
for safe, you get 8 bad images, 6 generic vectors, 1 color swatch and lets do this one, 33 of them are ones i honestly think i have seen before, they are just that generic. of the first page that i think will break 100 upvotes there are 15, with one passing 100
 
now on the explicit side, there are again 4 bad ones, 10 minor variation images, and 5 color swap, with 11 being generic to the point i thought i saw them before.  
now there are 15 images that have broke 100 upvotes, and 20 more that given a day or so will reach over 100,
 
here is also how it breaks down, no kid is going to draw an explicit image and be good at it, the rules to get an explicit image generally also skew toward porn, so to even be worth someone’s time of finding it and uploading it, you had to either draw it yourself or you had to like it, so its a built in filter.  
same with suggestive and questionable, each one has a filter that safe just doesn’t have.  
a crappy drawn safe image is better than a crappy explicit, its why explicit tends to hit higher highs and lower lows than safe typically does.

How’d that happen?

@Bhazor  
lol XD

@Mildgyth  
I don’t think it would matter anyway, mainly because the other pictures were also upvoted since then.

Proof this site needs more porn. We can’t let this happen again.

@Background Pony #FF26
 
I upvoted the Dash image after the screencap was taken. what you gonna do now?

We can also discuss the fact that anthro and humanization images are more likely to end up in 3 day top scoring (as well as have a higher score) than most pony pictures, even more so if we are talking about realistically drawn pony pictures without impossibly sized parts or hyper fetish…
 
Sometimes people from this site make me feel like only I think this fandom is and should be based around actual ponies =\

I have downvoted all images he faved and faved the one he didn’t upvote :D

@atalarikt
 
Around 33,000 images if you count grimdark, semi-grimdark and grotesque. if you exclude those that overlap with questionable, suggestive and explicit, you end up with around 23,000 images.

@Mildgyth  
But how about grimdark and/or grotesque?

@untraceablelarry
 
I tried to count. you have almost 700000 safe images against almost 200000 explicit, questionable and suggestive images. i don’t know how this translates in terms of percentages (it might be somwhere around 75% safe), but it is a fact that the the more well-drawn explicit-questionable-suggestive images get to “upvotes galore” more easily than other images.

I like it when the mods try to act as if this site was not largely a porn site. “Only 20% are explicit”… The fact this kind of thing is a special occasion is a better indication, I would say, of what at least a large portion of the site likes. Heck, just look at the all time top scoring list without filters; the first page is 80% porn. That isn’t it’s only purpose, for certain, but I swear that the mods downplay it a little too much, though I can easily see why they would… Though I still think that, if you are counting porn, you should probably count the questionable images as well; just look at the detailed description for that tag.

Huh, well that doesn’t happen too often.

Still counts.