News for hackers. Stuff that matters.

A comment on Browser-based climate change simulation. Parent comment

Re:Browser-based climate change simulation by Retric on Tuesday December 17, 2019 @11:19AM (#21815053)
It’s interesting how you described that as exactly 0.1F.

I have a very different way of looking at data. When I see a single significant digit I assume minimal precision. So, 0.1 is something like 95% chance of 0.1 +\- 0.05 would be optimistic, and 0.01 to 1 is pessimistic.

I suspect this is a common issue with both scientific and technological reporting for a general audience.

[ Parent | on HN ]
Re:Browser-based climate change simulation by bogwog on Tuesday December 17, 2019 @12:26PM (#21815937)
Increasing precision here doesn't increase confidence in these estimates.

The phrase "highly subsidized" says basically nothing, yet it's enough to estimate global temperatures 80 years from now with a margin of error of 0.05?

[ Parent | on HN ]
Re:Browser-based climate change simulation by zamfi on Tuesday December 17, 2019 @01:54PM (#21817191)
Click the three vertical dots to see what "highly subsidized" means -- it's a specific level of subsidy, which you can actually modify in greater detail in the "detail" view.

I absolutely agree that this thing needs a margin of error on the visual output!

[ Parent | on HN ]