A lot of stats or insights I see for things I often wonder how important they really are. Are they all really moving the needle?
E.g. lets say 50% of gun owners accidentally shoot themselves. Now lets say Im talking to someone about how I want to buy a gun for home defense. They tell me "well 50% of gun owners accidentally shoot themselves, so you're safer if you dont have a gun"
This is a gross misunderstanding if statistics. There is no way whatsoever to take the observed general trend, and apply it to a future possibioity in a single instance. What this interpretation misses is that there are variables well within your control. A gun owner who follows all safety rules will virtually never shoot themselves accidentally. Turns out 50% of gun owners are stupid as hell, and you are in control of yiur own outcomes.
I see this problem CONSTANTLY in debates online, on the news, etc. It is very frustrating and contributes greatly to propogating dumb ideas.
It is true that fifty years ago many people in Denver (and no doubt not only Denver) regarded New York as the embodiment of The Other and were very quick to pick up on a New York accent. But this was silly.
You go from a stable basin to an astable mountaintop basin where you are responsible for the maintenance of the banks.
And they're both used and useful much of the time. The field of risk management and industry of insurance are based on it.
Your frustrations are not uncommon. They are, however, rather poorly-founded.
They're an excellent illustration failing to get a handle on this safety thing (where "safety" is the inverse of "risk", as one of yesterday's submissions addresses:
<https://news.ycombinator.com/item?id=34036978>
There's also the philosophical question around the "sea battle" question, in which the truth of a statement about a future possibility ("there will be a sea battle tomorrow") is assessed.
<https://en.wikipedia.org/wiki/Problem_of_future_contingents#...>
They have shown us a graph showing more people are using the site.
I have asked for the same chart to include a metric showing that more people are using the site and there are less overdoses or medical interventions or another metric showing more people using the site results in some sort of quantifiable metric of community benefit.
The public health authority who provides the graphs is unable or unwilling to provide the information and charts I want.
I thus worry that data is selectively used to prove a predetermined narrative.
On the other hand, you can use stats for personal gain. For instance, prediction markets. That might have an impact.