Nos, I think you still don't quite get how the matching process works, and Daz will probably just flip out because he's explained it so many times, so I'll beat him to the point.
Oversimplified, it works like this:
Chronologically:
1. A random number is generated for the trial, let's call it TRVA-0123, set for let's say May 10, 2010
2. Today, the remote viewer targets the (future) target TRVA-0123
3. The report is encrypted and basically forgotten about for a few weeks
4. On May 10th, someone else blind to anything the RVers have done, picks a target, it could be a random event that occurred that day, or sometimes an algorithm is used to pick a specific event to ensure random selection, such as a computer will indicate "Pick an event on the 3rd page of the B section of the New York Times for May 10, 2010"
5. The target is completely selected at this point and recorded somewhere as "This is the target for TRVA-0123"
6. The TRVA-0123 file is then de-encrypted (remember, no one has seen it except for the RVer who wrote it, until now)
7. TRVA-0123 is compared against the actual target, and determined how close a match is.
It's not bad as double-blind goes, but there needs to be a more objective way of rating the hits vs misses. This is pretty much the entirety of the AIR report debates.