M-estimators are methods used to estimate something known as a parameter. Parameters are numbers that help to describe a certain mathematical or statistical model. M-estimators help to estimate these parameters using a process that balances out mistakes that are made in estimations. M-estimators look at past estimations and decide what the best guess is for the parameter based on all of the previous estimations. For example, if you were trying to guess how tall your friend is, you might make estimates and use M-estimator to figure out the best guess.