Your criticism is accepted. For my curiosity: which groups have you observed embracing the practice of introducing data on past model performance when presenting a new model? I failed to provide a source, but is it your impression that this isn't an area in which most people perform poorly