Whoever coined that headline phrase was certainly right about how data can be manipulated, but perhaps surprised at just how much statistic-twisting is going on in the social media age (and the age of not trusting anyone, from doctors to presidents to insurance companies -- well, that last one is just smart).
The former president claimed that the U.S. has left about $85 billion in military equipment in Afghanistan. Even the quickest analysis of that claim reveals that the U.S. sent $24 billion in equipment to the country we officially left yesterday... since 2002.
Turns out that much of it is out of commission or soon will be without constant supply replenishment. And much of the weaponry has been demilitarized -- think "blown up" -- by our own troops.
You never know if the ex-president's lies are due to sheer ignorance, evil intent, or hyperbole that excites the mob, but you can be sure that the figure he cited will recur on radical right wing radio and TV for some time.
I was thinking that citing the correct figure of $24 billion might be sufficient to make the point, but in the age of unsupported claims why not go for the biggest conceivable number?
I was also thinking that Covid statistics are very difficult to think clearly about. I have no idea how this number was compiled, but assuming good faith data collection, about 40 million Americans have had Covid. For some diseases, those who have survived a virus are now basically "immune." But the Delta variant (well, not a variant anymore, but you get the idea) seems to zoom right by whatever your white blood cells have built up.
So is spending much time on how many people have gotten sick worth much?
Lots of Covid tests are administered each week -- 1.3 million last week in the U.S. -- and if you divide those into the tests that came back positive for the virus, you get a percentage. It was about 10 percent across the nation last week. That 10 percent is an average, and the range is all over the place, depending on location.
It would be easy to start thinking that 10 percent of the population has the virus. In fact, it might be a tiny percentage or much larger. It all depends on the number of people bothering to get tested, wouldn't you think? It also seems related to who people voted for in the last presidential election. Correlation or causation? Tough to tell.
Schools in Georgia where students wear face masks (likely a small sample) during classes report about a 34 percent decrease in Covid cases among students and staff. That sounds great, but there could be many things that affect that number (more vaccinated students... ventilation within buildings... changes in scheduling... etc.).
And if we began with relatively small numbers of infected students, a 34 percent drop might mean that 1-2 fewer students got sick in an entire school.
Confused yet? Me too.