Performance impact of using single join vs using multiple joins to get same results in SQL

Question

I have two SQL queries (using Postgresql as backend ) which essentially retrieves the same results.

The two tables structures:

values[id(int), attr(int), v