Congrads on the italics :) yes a sample of 4,000 + makes it a reliable poll (if the sample was random).
Samples are not done randomly in polling. They are carefully constructed to capture the demographics of the country, state, region, etc.