20.5 A volcano ejects lava completely randomly. In long-term statistics, the volcano was, on average, active one day in ten days and inactive the remaining days. In the most recent 30 days it was active 25 days. This recently increased activity was taken into account by experts in the forecast for the next day. The radio news said that the probability of the volcano being active the next day was at least 30 percent. The prediction on TV quoted the probability between 50 and 60 percent. Both experts, on radio and TV, were known to use the m-estimate method. How can the difference in their estimates be explained?