okay so here is something fun about language.
if you want to learn how to read a language, like really learn it, one of the best things to do is read something you know in your target language. (for me, that's often stuff like dr. seuss, stuff i know by heart, with repeating themes.) after children's books, basically progress in the way children do while learning to read. try to use context clues instead of looking things up.
but tumblr loves to recycle a meme. and this is good because you know the format. you know the pattern of Spiders Georg, so your brain can take the context and figure out what it says using the patterning.
add to this that spanish has a huge number of cognates (see: every post wherein an english speaker goes, "wow i understand this spanish!" to the response of "angloparlantes cuando hay cognados")..
so in this, promedio is similar to medium, so we know that is average. en realidad is easily in reality, so we know that is actually. the phrase error estadístico is easily statistical error, especially knowing that spanish puts descriptions like that after the noun.
and we KNOW the meme. we know Spiders Georg who lives in a cave, so of course Leches Georg quien vive en una cueva. we know Spiders Georg eats 10,000 spiders a day, so of course Leches Georg absorbe 10,000 leches al día. the words make sense to english-speakers: we know what absorption is, and can infer ingestion; we know that día is day.
so by the time we get to the hardest phrase with the least cognates, we are already looking for "is a statistical outlier adn should not be counted," and can infer that "es un valor atípico qeu no debería haberse contado" should mean just that.
breaking it down, we can recognize a few cognates: atípico is atypical, so un valor atípico is an atypical case, an outlier. contado is close enough to counted and the negation in that phrase tells us that no debería haberse contado should be the conjugated form, just like should not be counted is.
if you speak english as a first language, especially as a monolingual, this is a COOL trick that your brain does automatically, and it can feel super mysterious if you don't realize that your brain has so much dedicated to Language. cognates allow your brain to use the natural ability to find similarities in word roots to understand!
you can often use this to figure out what things mean in spanish, especially since many of us have both english and spanish on many products, signs, etc. so if you look around at all, you are picking it up. many people know at least SOME spanish or french in the US (where I am), and we have enough loan words from both of these to easily see what many words or phrases mean in context.
anyway i have been thinking a lot about human linguistic development and now it's tumblr's problem :)