actually for 20 TCR (A) states that they forecast rain more often, it doesnt state correct.
the rason it is the correct choice is because otherwise, what if they claimed they were wrong only one day? sounds great right? what if they only forecasted 4 days and got 3 right? doesnt really bolster the argument that they are more useful and reliable. if they forecasted 190/200 (95%) and the other stations forecasted less then them AND were less reliable, that definitely stengthens the argument (example 80/100 for 80%) so a higher reliability AND more forecasts equal stengthened argument that they are more useful/reliable
as for 21.
if the first group responded to the questions with greater accuracy to begin with, but changed none of their answers, they had the same percentage of accuracy/inaccuracy. however, if group B were very inaccurate then flipflopped, they could then be MORE accurate. anothernumber example
if there were 10 questions, and A were wrong on 4, right on 6, thats 60% on both lines of questioning. if the B group flip flopped, they would be say 20% then 80% from A to B, right in line with the choice C.