_portrait.jpg){kind=link}
Text from article:
David Rice, a disabled Army paratrooper who has been on probation since joining the U.S. Department of Energy in September, also learned Thursday night that he had lost his job.
Rice, who has been working as a foreign affairs specialist on health matters relating to radiation exposure, said he’d been led to believe that his job would likely be safe. But on Thursday night, when he logged into his computer for a meeting with Japanese representatives, he saw an email saying he’d been fired.
“It’s just been chaos,” said Rice, 50, who had just bought a house in Melbourne, Florida, after he got the job.
Rice said he agrees with the Trump administration’s goal of making the government more efficient, but objects to the random, scattershot approach being taken.
Originally linked here:
This is interesting, how did they get those calculations?
The person you're replying to is describing (without giving proper context except for "game theory") an algorithm that's fairly successful at the "iterated prisoners dilemma": https://en.m.wikipedia.org/wiki/Tit_for_tat
I don't know where this particular graph came from, but Richard Dawkins has a whole chapter about strategies for the prisoner's dilemma in his book "The Selfish Gene".
Veritasium did a nice video covering the research and explaining the sources. It was an academic competition of sorts
There are a variety of ways. One way is to run a computer program that executes each strategy and then just have them all go against each other some number of times like a tournament, or sometimes just "random matchings". Super fast to do so it's easy to try different scenarios and make a lot of different strategies.
They've also done tournaments with actual people, and then compared the different people's behavior to the different "pure" strategies that they made. This helps them validate that the behaviors carry over.
https://en.wikipedia.org/wiki/Prisoner%27s_dilemma
It's worth noting that nation states don't always behave the same as individuals, but often closer to the game theory ideal. Additionally, there are circumstances where tit for tat isn't actually the dominant strategy, specifically when you know that the game is going to end.