The entropy is low; using only the last two numeric digits provides only 100 possible outcomes. This limited range increases the likelihood of participants predicting the result, especially on a large scale