You would do well to learn about effect size and p values. Your common sense guess at what might be reliable is incorrect, which means your doing a disservice to your own understanding of the world.