Выбор редакции

Beyond posturing, placebos or belief

Statistics, well done, are astounding.

They tell us, clearly and completely, what is actually happening.

Ignaz Semmelweis saved a million lives (eventually) with his approach to statistics, despite the fact that he was arguing for a significant change and was not at all well liked.

There are volumes of detailed and verified statistics about carbon and other emissions. They’re easy to find if we care to look at them and understand them.

More banal but simpler to visualize, it’s easy to dispense with hype and claims from a running shoe company like Nike, but impossible to dismiss this extraordinary report from the Times. They addressed the possible self-selection and placebo effects and still came up with a massive performance shift in an industry where I thought it was impossible to deliver a massive performance shift.

[If you’ve been putting off stats because the math is intimidating, spend thirty minutes with the Nike article and the Semmelweis story. It’s worth learning what they did, because it will help with your work and the way you see the world when you make decisions.]

Every once in a while, we can see a significant effect in the world, one that’s caused by engineering and can be measured. It’s rare, but it’s worth seeking out. Not everything is simply a matter of belief.

Yes, it’s easy to lie with statistics, but quite gratifying and insightful to tell the truth with them.

Statistics never work as well as we might hope. Since we’re humans, statistics don’t change minds. It’s the story we tell ourselves (and others) that do. Statistics are merely a consistent and reliable way to tell yourself a story that’s actually useful and resilient.