V$LIBRARYCACHE has columns for GETS and GETHITS. You can calculate the number of misses as gets - gethits. Misses always require a load - in the case of cursors, that is a hard parse. Similarly, pin misses always require a reload. Reloads are not necessarily expensive - it depends what is being reloaded. People focus on them because they are in theory largely avoidable. Reloads of cursors are similar to hard parses in that the query execution plan must be generated, but are slightly less expensive in that the handle and cursor definitions are already in the library cache.

I read your answer to the question about reloads. Now, I have questions:

What is the difference between the loads from GETS, and RELOADS from PINS? What makes the RELOAD so critical and expensive? If a cursor is lost and then a reload is nessesary, the parse (hard or soft parse?) isn't as expensive, I think.