After the announcement of national basketball championship you bought two tickets to the final game, hoping that your favourite team will be there. Tickets sold out quickly, and you felt lucky to have them. Unfortunately, your team didn’t make it, so your expectation of having a great evening was ruined. What should you do?
Going to the game anyway doesn’t bring a lot of enthusiasm. Selling or giving away tickets is not an option, because they have your name on them. It’s painful to think about losing money (tickets to the finals are expensive!), so the only option is to go. Or is it?
No one wants to waste anything when it comes from our own pocket. This is what sunk cost fallacy is—the tendency to go with the plan, even if it’s irrational to do so.
Stock traders have to deal with this everyday. If stock they invested in does poorly, it shouldn’t affect their future decisions about that stock. What matters is the potential of that stock, not its past performance.
When building software, we face decisions like this as well. We spend an enormous amount of time on a new feature, but not always we’re happy with the result. If we can’t find ways to improve that, the rational thing would be to scrape it off and do something else, but we don’t want to feel we wasted so much time for nothing.
Whatever we do, it brings us here. Our experiences, including failures, leads us to this day. Ignoring sunk costs means being rational with the information we have now, without remorse of past choices.