Hello,
I described one option to debug ranklog calculation in my blog - Explain Rank page
http://powersearching.wordpress.com/2013/01/25/explain-rank-in-sharepoint-2013-search
Tuning using query tools- yes, i'm afraid manually so far.
Tuning of ranking model - Sharepoint 2013 uses state-of-art algorithm to calculate document relevancy. Simply put it calculates values for various features (BM25 score, term proximity, click through and others) and then mix them together non linearly using
neural network (configured in ranking model) which has ~70 floating point coefficients to tune.
Now I'm preparing a blog post with in-depth explanation of how scores are computed, how to debug and how understand impact of each component. I will cover the topic of tuning coefficients in ranking model as well. Simply put : all you need is to collect
a lot of so called "Query Judgments" - user feedbacks of search performance, and then "train" your model using machine learning approach.
Subscribe to my blog if you're interested in updates regarding relevancy stuff in Sharepoi