Skip directly to site content Skip directly to page options Skip directly to A-Z link Skip directly to A-Z link Skip directly to A-Z link
Volume 25, Number 12—December 2019

Etymologia: Markov Chain Monte Carlo

Ronnie HenryComments to Author 

Cite This Article

Markov Chain Monte Carlo

A Markov chain Monte Carlo (MCMC) simulation is a method of estimating an unknown probability distribution for the outcome of a complex process (a posterior distribution). Prior (capturing the concept prior to seeing any data) distributions are used to simulate sampling from variables that have known or closely approximated distributions in the complex process. Thus, the prior distributions are known probability distributions that represent uncertainty about a particular attribute of a population prior to data sampling, and the posterior distribution represents estimated uncertainty about a population attribute after data sampling and is conditional on the observed data.

Monte Carlo (named for the casino in Monaco) methods estimate a distribution by random sampling. Many samples of the prior distributions must be obtained (e.g., many rolls of the dice) to obtain a stable and accurate posterior distribution. The modern version of the Monte Carlo was invented by Stanislaw Ulam and developed early on by John von Neumann and Nicholas Metropolis, the latter of whom suggested the name, as part of follow, on work to the Manhattan Project. Ulam was trying to calculate the probability of laying out a winning game of solitaire from a shuffled deck of 52 cards. Because of the complexity of the calculations, he decided it would be easier to play 100 games of solitaire and count the percentage that won.


Thumbnail of Andrey Markov (1856–1922), photographer unknown, public domain,

Figure. Andrey Markov (1856–1922), photographer unknown, public domain,

In a Markov chain (named for Russian mathematician Andrey Markov [Figure]), the probability of the next computed estimated outcome depends only on the current estimate and not on prior estimates. For example, if you shuffle a deck of cards 3 times, the outcome of the third shuffle depends only on the state of the cards at the second shuffle, not at the first shuffle. Markov chain Monte Carlo simulations allow researchers to approximate posterior distributions that cannot be directly calculated.



  1. Eckhardt  R. Stan Ulam, John von Neumann, and the Monte Carlo method. Los Alamos Sci. 1987;15:1317.
  2. Hamra  G, MacLehose  R, Richardson  D. Markov chain Monte Carlo: an introduction for epidemiologists. Int J Epidemiol. 2013;42:62734. DOIPubMedGoogle Scholar
  3. van Ravenzwaaij  D, Cassey  P, Brown  SD. A simple introduction to Markov Chain Monte-Carlo sampling. Psychon Bull Rev. 2018;25:14354. DOIPubMedGoogle Scholar




Cite This Article

DOI: 10.3201/eid2512.et2512

Original Publication Date: November 06, 2019

Related Links


Table of Contents – Volume 25, Number 12—December 2019


Please use the form below to submit correspondence to the authors or contact them at the following address:

Ronnie Henry, Centers for Disease Control and Prevention, 1600 Clifton Rd NE, Mailstop E28, Atlanta, GA 30329-4027, USA

Send To

10000 character(s) remaining.


Page created: November 18, 2019
Page updated: November 18, 2019
Page reviewed: November 18, 2019
The conclusions, findings, and opinions expressed by authors contributing to this journal do not necessarily reflect the official position of the U.S. Department of Health and Human Services, the Public Health Service, the Centers for Disease Control and Prevention, or the authors' affiliated institutions. Use of trade names is for identification only and does not imply endorsement by any of the groups named above.