![Reed hastings](https://static-ssl.businessinsider.com/image/5931c885b74af4063e8b554b-2254/gettyimages-503631370.jpg)
Getty/Ethan Miller
Netflix CEO Reed Hastings
Netflix recently cancelled a pair of its high-profile original shows. If CEO Reed Hastings gets his way, there will be a lot more cancellations to come.Both "Sense8" and "The Get Down" got the axe after limited runs. That brought the number of original shows Netflix has dropped to somewhere between six and eight, depending on how you count them.
"Our hit ratio is way too high right now," Hastings said this week at the Code Conference. "I'm always pushing the content team. We have to take more risk.
"You have to try more crazy things, because we should have a higher cancel rate overall."
Hastings might just be trying to prep Wall Street investors for more cancellations, and spin them as a good thing. But his statement does bring up an important point about the investments Netflix and rival Amazon are making in original content.
Both Netflix and Amazon pride themselves on the vast troves of information they have about their customers they can mine for competitive advantage. Netflix famously said it used data to determine that "House of Cards" had a good shot of being a hit. Meanwhile, the company's decision to fund a string of reboots - "Full House," "Gilmore Girls," and so on - was almost certainly data driven.
But there are limits to what data can tell you, according to Amazon Studios boss Roy Price.
"You can look at what people watch but you can't be too deterministic about it," Price said earlier this year. "The show that will be a real game changer will be a rule breaker, not what people are watching today."
Another way of putting that: Viewer data likely can't tell you which shows will be breakout hits.
Hastings echoed those sentiments this week.
By taking risks, Netflix puts itself in position to have shows that are "just unbelievable winners," he said. He pointed at "13 Reasons Why," a show that elicited controversy for its depiction of teen suicide but was a huge hit.
"It surprised us too," Hastings said of how popular the show became. "I mean it was a great show, but we didn't realize how it would catch on."
As Netflix's strategy evolves, it's clear that while data will inform what shows its keeps in its catalog - and may even lead to reboot after reboot - it can't and won't be the sole defining factor for its decisions.
Show business, after all, needs a bit of magic, even if sometimes the tricks don't work out.