I've been looking around trying to find any papers/articles/forums on this topic, but haven't found much. How strongly would you say performance in the first year (course grades and prelims) is correlated with research quality and performance in later years? How strongly does first year predict success in the research phase of the PhD?
I suspect there is at least some correlation - but are people at the very bottom of their class (in first year, who do pass prelims) more or less doomed to stay there? Are people at the very top 'very' likely to stay at the top of their class later on, or just 'slightly' more likely? How about your average Joes?
Thoughts on this, especially from current/past faculty?