Think of 'and' as a filter applied twice. Flip a coin and roll a die. Half of all tries give heads; of those, only one in six also shows a 6. A half of a sixth is 21×61=121. Listing every outcome agrees: there are 12 equally likely coin-and-die pairs, and only one of them is (heads, 6). Each extra 'and' makes the event rarer, so the answer shrinks.
'Or' works the other way: it collects outcomes, so the answer grows. Draw one card and ask for a heart or a king. There are 13 hearts and 4 kings, but the king of hearts is on both lists. Count 13+4=17 and you have counted it twice; the true count is 16, so P=5216=134. That is why the 'or' rule subtracts the overlap: P(A)+P(B)−P(A and B).
Watch whether the first event changes the second. Drawing two cards without putting the first one back looks like two separate events, but it is not: after a heart is drawn, only 12 hearts remain among 51 cards, so the second chance is 5112, not 5213. Two coin flips never affect each other; two draws from the same deck do.
Think of it like this. An 'and' question is like getting through two doors in a row: you must pass the first, then the second, so the chance can only shrink. An 'or' question is like having two doors to choose from: either one lets you in, so the chance grows, but anyone who fits both doors is counted once.
Core idea. Compound probability combines two or more events: P(A and B) multiplies, P(A or B) adds then subtracts the overlap.