There are couple of times in my career as a programmer where I've made a bad decision because I misinterpreted data — analytics, user research, production data samples, etc. Sometimes the data was just plain wrong; sometimes the data was accurate enough but I drew the wrong conclusion from it. I'm surely not the only person to interpret data incorrectly but the topic seems to be rarely discussed in our industry.
That's why I like this article "The time I tried to ruin Halo 2". It is a good reminder to always think critically about the data in front of you.