On other Stack Exchange/Stack Overflow sites, voting is used as a form of quality control. High quality questions get lots of votes and are seen by many, low quality questions get negative scores and are pushed away from answerers.
But what is the purpose on Code Review, where some of the point of the site is to improve lower quality code?
The same happens ...
As you surmise, the quality of the description plays a large part in deciding how to vote on a question:
Is the requirement (problem statement) clear and concise, with appropriate mention of any unusual constraints?
Are there any examples showing how it's used?
Did the asker explain why this algorithm and language were chosen, and what other approaches were ...