basic mysql query question

I need help formulating a query to return results based on three tables I am using to store data relating to film box office revenue. (I’m not sure it is possible to put this in one query, in which case some alternative will be necessary.)

The three tables and their relevant columns are:

1) films (This table stores US box office gross data for films.)
      a.      film_id
      b.      title
      c.      box_result (the film’s US box office gross)
      d.      close_date (date the box_result was posted)

2) estimates (This table stores US box office gross estimates given by users.)
      a.      user_id
      b.      film_id
      c.      estimate

3) friends (This table stores data regarding who each user’s associates are.)
      a.      fan_id (the user_id of the user indicating someone is his friend)
      b.      star_id      (the user_id of the user so indicated)

I have stored a .sql dump of the test db here: http://infinidex.com/film_sample.sql

For the logged in user $_SESSION[user_id] I wish to return the following data for each film_id for which the user has provided an estimate (ie, there is a match between the user’s user_id and the film’s film_id in the estimates table):

1)      title
2)      box_result
3)      estimate provided by the user for this film
4)      % error of the user’s estimate [user_percent_error]
5)      the total number of estimators for this film [total_estimators]
6)      the average estimate provided by all users for this film [average_estimate]
7)      % error for the average estimate [average_percent_error]
8)      the user’s rank among all estimators for this film [rank]
9)      the user’s percentile rank among all estimators [percentile_rank]
10)      the average estimate provided by the user’s friends [friend_estimate]
11)      the total number of friends who provided an estimate for this film [total_friend_estimates]
12)      % error for the friends estimate [friend_percent_error]
13)      the user’s rank among his friends for this film [friend_rank]
14)      the user’s percentile rank among his friends [friend_percentile_rank]

ordered by close_date descending.

This would be the ideal format for the query, but if it is not possible to do this, or it would result in a very slow query, I am open to suggestions on alternatives. For example, I could store the user’s rank for each estimated film in the estimates table (calculated in advance), and the total number of estimators for a given film in the films table (also calculated in advance), which would mean that the query would require no rank calculations.

Thanks for your help.