It is easy to look back and say, "I should have made this modular," or "I should have isolated that external dependency." It is easier to see how another level

It is easy to look back and say, "I should have made this modular," or "I should have isolated that external dependency." It is easier to see how another level of indirection would have saved you work after the fact, but maybe YAGNI was the right call at the time. That's the trap that Magic and poker players call results-oriented thinking: You got burned once by making the best decision available given the information you had, and you give that experience a disproportionate weight, like the player that lost big with a pair of kings in his hand, and undervalues a pair of kings after that.