4 ms·

It's based on the number of whales killed by whalers http://onlinelibrary.wiley.com/doi/10.1111/j.1748-7692.2004.tb01190.x/abstract;jsessionid=EC6D0DEF83EF0D2B

• LeifCarrotson 9y ago
  > Pre-exploitation population estimate of 239,000 (95% interval 202,000–311,000). > Estimated 1996 population size, 1,700 (860–2,900) Don't use three significant figures when your confidence interval spans 50% of your estimate. "200,000 to 300,000" would more clearly express the data. And be honest about accuracy: if modern satellite, aircraft, and naval technology can't pin the 1996 value down more tightly than "half this number to double it", you can hardly say with a straight face that your confidence interval on the 1800s data is "minus 15% to plus 25%".
  • accountyaccount 9y ago
    Sometimes shitty data is all you've got — I was making a simple point, not submitting a paper to a journal. Humans killed these things by the tens to hundreds of thousands and the population is no where near that today.
    • LeifCarrotson 9y ago
      Sorry, I was commenting on the numbers in the paper and not on your simple point, with which I agree
      • accountyaccount 9y ago
        No worries. Thanks for being cordial. It's easy to get frustrated when conversations turn this way online.