Oh, the K/Ar method that would be completely useless if the tested igneous level is less than 2 million years old?
Depends on half lives of the isotopes. You have to have an appreciable amount of “daughter” elements to account for the background noise, etc.. There will be an amount of argon in the sample (you test for that) so you have to be able to measure the daughter argon additional to the background. Frequently two means of radio metric dating are used to verify the date.