High bias means the data is being underfit. The decision boundary is not usually complex enough. High variance happens due to over fitting, the decision boundary is more complex than what it should be.
High bias high variance happens when you fit a complex decision boundary that is also not fitting the training set correctly in several places. See example below :