Taking selfies is more dangerous than swimming with sharks, and there are numbers to back it up. So far in 2015, eight people have been killed by sharks, whereas 12 have died, usually from falling or being struck by a vehicle while in the act of immortalising themselves on Facebook, Twitter and Instagram.Most media outlets concluded that "taking selfies is statistically more dangerous than swimming with sharks".Not so fast. As the astute among you may have figured out already you're being misled by being shown only part of the picture.What is missing is exposure. You can't simply compare the number of deaths when one activity (swimming in the same water as sharks) is extremely rare and the other is, well, way too common.For every person exposed to a shark, there are perhaps 10million who are staring at themselves in a phone screen.So, each person's risk of dying while taking a selfie is actually incredibly low, in the order of many millions to one. The total numbers may be of concern to some authority body, but that's separate from saying that selfies are more risky than swimming with sharks.What does this have to do with sport? Well, it struck me that many in the media make the same mistake of not appreciating context when analysing statistics in sport.Over the course of last week I read and heard numerous pundits offering thoughts and predictions for the quarter-finals based on their presumably detailed and insightful look at the statistics from the pool stages.You'd be surprised at how many make this very same selfie vs shark error in thinking.One example was their evaluation of New Zealand in the pool stages. Much was made of their sloppiness and ineffectiveness, backed up by statistics like 44 knock-ons and only 29 turnovers won.This is not, they proclaimed, the New Zealand team we expected to see.Aside from the fact that teams can manage their focus and intensity through a tournament (not every match must be a die-for-your-country battle) as New Zealand so destructively showed this past weekend, the experts were also missing the crucial hidden insight of exposure.Take knock-ons: in all their matches New Zealand utterly dominated possession, and their style of play involves more passing (and at pace) than any other team.So, while 44 knock-ons would have been too high for Steve Hansen's satisfaction, the number alone can't be compared to South Africa or Scotland, who passed the ball fewest in the pool stages.When your tactical approach is to use a big, ball-carrying forward as flyhalf, you're passing once per ruck and the risk of knock-ons is extremely low. Like the selfie analogy, New Zealand's exposure to passes is greatest, so you expect to see more knock-ons, and by simply comparing numbers without appreciating that context you'll arrive at an incomplete conclusion.The same is true for turnovers won. By definition you can only win a turnover when the opposition has the ball, and so the fact that New Zealand ranked relatively low on turnovers won is not an indication of their inability, but rather the result of context (their opponents had so little of the ball).New Zealand did, in fact, win more turnovers per tackle made than any other team. But that's a level analysts don't explore.Overall, New Zealand were not nearly as poor as people thought, but emotion trumped evidence to create a narrative that was summarily erased on Saturday night.I appreciate that sport is about passion and emotion, above all else. Some don't care for technical analysis at all. But I do sometimes feel we're in danger of devolving into a situation where soon an analyst on TV is going to hit us with this pearl; "The key today for the Springboks will be to score more points than New Zealand."Let's hope the brains trust of the Boks are at least aware of sharks, selfies and context when they prepare for the All Blacks. Emotion and bad evidence can only go so far...