A cost is "sunk" once it's already been spent and can't be recovered, no matter what decision gets made from here — and the sunk cost fallacy is the tendency to let that unrecoverable past spending improperly influence a decision that should really be based only on current facts and future prospects. "I've already put so much money into this, I can't stop now" is the classic form the fallacy takes — but the money already spent is exactly as gone whether you continue or stop. The only question that should actually matter is whether continuing, starting fresh from today, is genuinely the best use of whatever resources remain.
This shows up constantly in investing: continuing to add money to a clearly failing position specifically because of how much has already been invested, rather than because the current facts support it. It shows up well beyond investing too — staying in an unfulfilling career path, or continuing a failing business project, partly because of the years already invested rather than because continuing is genuinely the best choice looking forward. The pattern is the same in every version: past, unrecoverable spending is quietly treated as if it were still relevant information about the right decision now, when it genuinely isn't.
Correct moves you up, wrong moves you down — reach 100 to master this lesson.