>If you come from a math, statistics, or physics background you may have leaned that a Markov chain is a set of states that are sampled from a probability distr

>If you come from a math, statistics, or physics background you may have leaned that a Markov chain is a set of states that are sampled from a probability distribution. That's a very weird way of putting it. A set of states sampled from a probability distribution is usually just called a 'sample' and has nothing to do with Markov or chains. Really 'Markov chains' are called 'chains' because they represents an infinite sequence of states, which depend on each other. And they're called 'Markov' because they have the 'Markov' property, which roughly states that the probability distribution for each state is completely determined by the previous state.