You leave your bike outside a shop. There is a 1 in 50 chance it gets stolen, and replacing it would cost $300. Is that a big risk? The chance sounds small and the loss sounds big, so neither number settles it alone. Multiply them: 501×300=6. On average, leaving the bike unlocked costs you $6 each time. Do it 50 times a year and you should expect to lose about one bike.
Comparing risks means comparing these products. Option A, a 10% chance of losing $200, has an expected loss of $20. Option B, a 50% chance of losing $30, has $15. B's loss is five times as likely, yet A is riskier because its loss is so much bigger. The same arithmetic prices protection: a lock that cuts the theft chance to 1 in 500 lowers the expected loss to $0.60 a stop, a saving of $5.40 each time, so a $20 lock pays for itself within 4 stops.
Risk is not the same as probability, which ignores how bad the outcome is, and it is not the same as expected value, which adds up gains as well as losses. Expected loss also leaves something out. A 1 in 1,000 chance of losing $100,000 has the same $100 expected loss as a sure loss of $100, but most people rightly treat the first as more dangerous, because they could not afford it if it happened.
Think of it like this. Risk is the amount a careful saver should set aside. If your phone has a 1 in 10 chance of needing a $200 repair this year, setting aside $20 covers an average year, and a bigger chance or a bigger repair raises the amount. Where the picture stops: $20 is enough on average, not in the one year the repair actually happens.
Core idea. Risk weighs the probability of a loss against the size of that loss: Expected Loss =P(loss)× amount of loss.