Take the earthquake in Pakistan. Last night about midnight the death toll was in the 16k-18k range. Midnight is about when the papers go to press. This morning the paper shows up and says the death tool is at 20k. I have to wonder if the papers know they will be read 6 to 8 hours after press time try to do a statistical analysis to try to figure out what the numbers will be at when their audience reads the paper, and then make and educated guess and present it as fact.
This is not like a presidential campaign where you have two options so they can print two sets of papers and release them once they know. And these are fairly hard numbers, which there are a lot of factors influencing.
Basicaly, when a paper cites a number, can you count on that number being accurate, or arethey just making a wild stab?