You will definitely have to achieve a style and quality of paraphrase which demonstrates your understanding of the subject, but you should also be aware that the approval you're striving for is given by a human whose social prejudices and cognitive biases will dictate a great proportion of their action. My advice (in addition to general good style);
Find out which philosophers your professors like and quote them lots. People tend to avoid cognitive dissonance (see here for an explanation), your professors will be more inclined to encourage behaviour which agrees with their world view than behaviour which challenges it.
Dedicate yourself to writing one really good paper, and attend class regularly. Prejudice plays an important part in interpretation especially where you have not communicated your meaning precisely, and your professors are more likely to interpret errors favourably if they think you are a 'good' student generally as research at NatCen has shown here. For a more wide ranging proof of the expectation bias of academics see here,and most recently here, or here or any of the hundreds of examples like them in well-respected journals of all disciplines.
Take steps to determine what the Zeitgeist is around the subject. Your study will include lots of standard texts, but most professors will be more inclined to see the current paradigm as correct than any previous (a combination of availability cascade and confirmation bias), so bring in a wide range of sources, but focus your conclusions on the current paradigm.
Finally, be white and male (see here), or at least get used to being judged less favourably if you're not.
Based on the comment below, and my presumption in the absence of explanation as to the reasons for the down-votes, I should clarify that the question asked for methods for avoiding degrading comments on essays despite a poor writing style. Assuming that "improve your writing style" is so obvious an answer as to be tautologous, exploiting biases which seem, on the balance of evidence, likely to be present (or developing a thick skin where they cannot be exploited) is an entirely rational course of action. It does not require one to reach a conclusion about whether such biases are proven to exist beyond reasonable doubt, only that they may, and that a low-cost strategy exists that would take advantage of them.