They will be at least as accurate as the climate change algorithms.
Sounds like a great tool to pre-identify potential soldier material, as in those most likely to be willing to fight, and win!
“They will be at least as accurate as the climate change algorithms.”
No. The numbers reported are a lot better than the Climate Change algos. The problem with this kind of study is the false positives. Even though a young man falls in the high risk group, the probability he commit a violent crime is probably still quite low.
So people get labeled as “likely violent offenders” and the gvt spends a bunch of money counseling folks who don’t actually need help.
We do this kind of assessment all the time. If you’re walking down the street and encounter 1) a young man with tats in a hoodie smoking a cigarette and doing the pimp roll or 2) a 55 year old man wearing a suit and carrying a briefcase, you correctly assess the odds you will be mugged from 1) as many times higher than 2). So 2) is in the high risk group and you cross the street.
Notwithstanding that, the probability that 1) is a mugger is actually quite low. You have mostly false positives even though your informal algorithm is quite correct.