Fiction is a model of human behavior (among many other things, but this is where I'm going today). This means that no matter how a writer tries, real life and real people are always more complicated than whatever is in the model. Nevertheless, we do everything we can to make stories as "real" as we possibly can.
That leaves writers in a cleft stick. There's only so much room in any given story (even in one of those 200,000-plus word things), and showing complexities, whether they're complexities of situation or of character, takes a lot of space. Often, this means that the story in the writer's head has to flatten out a bit in order to fit…and the writer doesn't always notice. After all, we already know all sorts of things that never make it onto the page. The question is what to leave out.
Most writers, especially early in their careers, choose what to leave out by instinct. Usually, that works fairly well - most writers have read enough fiction that they've developed a reasonably decent eye for what does and doesn't have to be in a story.
Sometimes, though, working by instinct alone causes difficulties, and when it does, it's usually due to one of four things: either the writer leaves out really important things because they are so obvious to the writer that he/she forgets that they won't be nearly as obvious to the reader, or else the left-our bits are what I call "word-processing errors" (things that happen because word-processors make it so easy to snip a bit here and a bit there, without ever realizing that the snipped bit was the only mention of a particularly critical piece of information), or else the writer is strongly plot- or character-centered, and tends to leave out important parts of whichever one they're not centered on, or else the writer has overestimated just how common some critical piece of information is, or how good his/her readers will be at picking up on assorted hints.
About the only thing one can do to avoid the first two problems - leaving out things because they are so obvious to the writer that they seem not to need mentioning, or snipping something critical and not realizing it - is to have a bunch of first readers who are not afraid to say "I have no idea where the hero got hold of that sword he's suddenly using in Chapter 6″ so that the writer can look and realize that he never mentioned the hero picking it up during his visit with his uncle in Chapter 4. Once the writer's attention has been called to something like this, she/he can make a conscious decision how much of a mention to add.
First readers also help with the third sort of problem - being so strongly plot- or character-centered that important bits get left out because they're not plot- or character-related. So does experience of the sort that leads to self-knowledge. Writers who know going into a manuscript that they're plot-centered (or character-centered) can often make conscious adjustments, reminding themselves during the first draft that they need to put in more plot, or making an extra revision pass specifically to add some of the missing character or plot bits.
And then there's that last one: overestimating the breadth of knowledge available to one's readers, or the sort of hints they're likely to pick up. The thing about this is that unlike the other three, it's not necessarily a writing problem. It's not really about the story, or the clarity of the writing; it's about the readers themselves.
Writers control their words, not their readers. Some stories need to be told by implication; sometimes the most effective approach is the indirect one. When this happens, the writer simply has to accept the fact that a choice must be made: either the writer tries to be clear and obvious for the maximum number of readers, regardless of what that does to the story, or else the writer tells the story as it needs to be told, and accepts the fact that not all readers will appreciate it or get it.
Of course, if the writer simply has a horror of being too obvious, then being aware of this and trying to push one's limits a bit can prevent a story from sinking in clouds of needless obscurity. I feel, however, that it is generally better to overestimate one's readers a little bit than to underestimate them. A writer whose work gives people the impression that he/she thinks all readers are more than a little dim is unlikely to be terribly popular.