Last Week Tonight host John Oliver finds himself in the midst of controversy as he brought the name of God and the Old Testament to debunk clickbait myths that included caffeine.

During the May 8 episode of his show, the host laid down his own study claiming that most of the researches being done is nothing but myth.

As an example, John Oliver made use of the study on coffee. It has undoubtedly become one the most popular subjects of research in the recent years.

The host went on to enumerate the supposed health benefits of drinking caffeine. This includes helping in the prevention of colon cancer, reserving effects of damage on liver and decreasing risk of endometrial cancer. He also presented the flip side of caffeine like its link to increase in the risk of miscarriage.

Also Read: Coffee Good for Cancer Cure?

However, he invited controversy when the talk show host dragged the name of God in presenting his case.

“Coffee today is like God in the old testament,” John Oliver said per Zap2It.  “It will either save you or kill you, depending on how much you believe in its magic powers.”

He went on to say that the real danger lies in people looking for a study that would accommodate their own liking.

“If you start thinking that science is a la carte and if you don’t like it, another study will be along soon, that is what leads people to believe man-made climate change isn’t real,” he said as shared by Uproxx.

“Or that vaccines cause autism — both of which the scientific consensus is pretty clear on.”

Also Read: Coffee Linked to Miscarriage 

The Last Week Tonight host laid down the reasons why there are so many “bad research” being regularly published. One is that replication, which would have verified finding, is almost never done.

He claimed that “replication studies are so rarely funded” thus they are “underappreciated.” They also never get published.

John Oliver also pointed out that findings are often blown out of proportion by the media. He said that reporters do not usually go through a study’s methodology, and instead go directly to the results.