Since the question is nonspecific, I can only guess that what instigated your complaint here is the response to a recent answer you gave: http://mathoverflow.net/a/194916/2926 I have temporarily undeleted the post so that others can see what is going on here, and I have placed a lock on the question so that no further voting can take place while we discuss this. (However, this action is as I say temporary: the action of the community members who voted for deletion should not be reversed without strong arguments.)
I agree with the commenters who say they do not see how your post answers the question as asked. The problem statement asks whether one can prove a statement of the form $\forall n \; \text{(under some conditions)}\; \exists D \in \mathbb{Z}[x] \ldots$. Giving an example (in a comment) about a special case $n = 2$ does not constitute a proof, and the body of the answer doesn't give a proof either: it seems only to give a (well-known) method for generating a set of solutions. But to give a proper answer, you have to give an argument which applies to all $n$, covering all the (infinitely many) cases.
It's my observation that many of the answers you give at MO do not go into such proofs at all, but rather exhibit sets of solutions to Diophantine problems in parametric form, which may or may not be the complete solution sets. Just to pluck one such example, there's this thread: Does the Diophantine equation $(x^2+ay^2)(u^2+bv^2) = p^2+cq^2$ admit a complete solution? Generally your posts seem to involve a series of algebraic or symbolic manipulations whose correctness is perhaps best tested using software which handles symbolic algebra, but in all the examples I've looked at, I've not seen a proof from you which is guaranteed to handle all possibilities. Meanwhile, the formulations of the questions make it clear that that's what is wanted and expected. Thus, your posts are often considered to be unsatisfactory according to the mathematical standards here. This is frustrating both to you and to others.
Joonas has asked about your professional relation to mathematics. Certainly your answers give (me, anyway) the strong impression of someone who is not a professional but someone who likes Diophantine equations or number theory as something of a hobby. Obviously there is nothing wrong with that, but the standards of proof at MathOverflow are very high, and it seems to be the opinion of many that such answers do not meet the standard.