A third factor should be added to ascertain quality; whether or not the algorithm was designed for profit. If yes, then bias exists, such that only data deemed useful for maximising profit is considered valid. All other data is discarded.