Bias exists, but people underestimate the effect of noise. For example, a weighing device, if it gives you a weight gain of 4 kilos every time, is biased, but if it gives you a different measurement every time, then it is affected by noise. The combination between the two is the most risk. Ultimately, noise and bias It comes from errors.