Roll one die. Let A be "even" (2, 4, 6) and B be "less than 3" (1, 2). "A and B" needs both at once: only the 2 qualifies, so P(A and B) = 1/6. "A or B" accepts either: 1, 2, 4 and 6 qualify, so P(A or B) = 4/6. The 'and' event is smaller than either event alone, and the 'or' event is bigger.
The shortcuts come from that picture. For 'or', adding P(A) + P(B) = 3/6 + 2/6 = 5/6 counts the 2 twice, once as even and once as small, so subtract it once: 5/6 − 1/6 = 4/6. For 'and' across two separate actions, like flipping a coin and rolling a die, the coin cannot affect the die, so P(heads and 6) = 1/2 × 1/6 = 1/12: half the time you get heads, and in a sixth of those you also roll a 6.
Multiplying the plain probabilities only works when the events do not affect each other. Draw two cards from a deck without putting the first back: P(both aces) is not 4/52 × 4/52, because once one ace is gone only 3 aces remain among 51 cards, giving 4/52 × 3/51 = 1/221. Ask whether the first result changes the chances for the second; if it does, update the second factor.
Think of it like this. 'And' is one door with two locks: you get through only if both open, so it is harder than opening either lock. 'Or' is a room with two doors: either one lets you in, so it is easier than using one particular door. The picture tells you which way the chance moves, not by exactly how much.
Core idea. Joining events with 'and' keeps only the outcomes where both happen, and joining them with 'or' keeps the outcomes where at least one happens; multiply for independent 'and' events, and for 'or' add the two probabilities and subtract the overlap once.