Finally, this poll was a "select the candidate" poll, where people are given a list of potential candidates, and they pick the one individual for whom they'd prefer to vote for for POTUS. Now, given that there were likely 5 or 6 names on there, and given that it's VERY early in the election cycle, it's not unexpected that the numbers would be fairly spread out, without on individual far in front. Thus, 36/100 people chose Clinton, 29/100 people chose Obama, and 35/100 people chose one of the other people on the list. It's not an approve/disapprove poll. Do you see it now?