Given the tiles are algorithmically rather randomly (as in normal Tetris) chosen, is there an optimal play? A trendy computational approach will be to utilize a

Given the tiles are algorithmically rather randomly (as in normal Tetris) chosen, is there an optimal play? A trendy computational approach will be to utilize a neural network ala AlphaZero.