Surely "large" is subjective and 1 in 7000+ doesn't seem that large.

However it seems to you Jim, statisticians, the polling industry and their customers all find such a sample gives an accurate indication of the views of a population.

What is the value of your opinion on survey techniques and methodology?
Who the fuck are you?