If you had kept going down the list, you'd see this: > The studies only compare differences among developers who actually complete the task. > Their figures d

If you had kept going down the list, you'd see this: > The studies only compare differences among developers who actually complete the task. > Their figures don’t take into account the people (~10% in some studies) who didn’t even finish. Nor can they take into account the real-world cost of software that is nominally completed but is so buggy, flaky, or hard to maintain that it has to be rewritten by someone else."