I've written before about some of the explainable reasons why lab procedures don't work. I forgot to add that sometimes experiments don't work because of outside forces that have nothing to do with your technique, inability to do algebra, or poor experimental design.
Sometimes, the whole thing is just doomed from the start.
A friend of mine once spent a summer as an intern at a local, now-defunct, biotech company. During this time, he worked for a very demanding technician for whom he developed a strong dislike. Every day, he patiently stood by as she unpacked sample jars filled with pinkish media and small chunks of tumor tissue. The evil tech's lab routine revolved around taking these bits of tumor tissue, grinding them up, and starting primary cell cultures. My friend would maintain these cultures once the cells were happily growing. But every day he suffered as she harangued him for missing this or forgetting that.
He was also a big fan of Chinese food. One day, it occurred to him that the barbecued pork he was eating for lunch bore a strong resemblance to the samples of tumor tissue that his supervisor received in the mail. My friend put a small chunk of barbecued pork into a sample flask, added media, added a label, and set it on his nemesis' bench where all the other samples were congregated.
She never noticed a difference and processed the sample like all the others.
I heard reports that the evil tech agonized over that misplaced chunk of barbecued pork.
But she never did get it to grow.
Doctor: "Well, I just got your biopsy results back from the lab, and I'm afraid I have good news and bad news."