During a password entry, a matrix is traversed to select a position. The matrix includes several levels, each level in the several levels including at least one position where data can be entered, where a second level in the matrix forms a sub-level of a first level, and where the second level is reachable only from a particular position in the first level. In response to an input, a mode of the selected position is changed such that the position becomes unchangeable and unselectable during a remainder of the password entry. The selected position is encoded in an auth-step. In response to an indication of an end of the password entry, an auth-code is transmitted. The auth-code includes a set of auth-steps, and the set of auth-steps includes the auth-step.