Most people are wired to underestimate personal risk

Sharot's surveys and lab studies find a consistent pattern: when people estimate their own chances of divorce, disease, unemployment, or being a crime victim, they guess lower than the actual population statistics — even when they know those statistics. Ask the same person about a stranger's odds and the estimate creeps back up toward reality. The gap is specifically self-directed.

This isn't limited to any one demographic; it shows up across income levels, education levels, and cultures, though the specific things people feel optimistic about vary. Smokers underestimate their own lung cancer risk while accurately citing national smoking-related mortality rates. Newlyweds who know the divorce rate is roughly 40-50% will still tell you their own marriage has almost no chance of ending that way.

The bias isn't corrected by better information because it isn't primarily an information problem — it's a feature of how the brain processes self-relevant predictions. Takeaway: knowing the statistics rarely fixes optimism about your own case, because the bias operates beneath the level of the facts you consciously hold.