That's not how statistics works. You can have a significant effect with a tiny effect size, or a large effect size that on analysis turns out to be insignificant.
TotallyHuman
joined 2 years ago
How so? The study showed no consistent association between funding and crime rates. That is true verbatim.
If there's no zero in the dataset, then we don't have any zero about data. It could be, for instance, that some police have a large effect, but that you hit diminishing returns incredibly quickly.
It specifies who the protester was. If it was a Yellow Vest, they would have said that.
If they're dead, they can't recover. Yes, fixing people is better than just holding them together, but holding them together buys more time for permanent solutions.
Small comfort to the reporter who got beaten up or SWATted or stalked, or the news organization that gets vandalized or DDoSed. If you're more likely to visit violence on your critics, people are less likely to criticize you. It's not fair, it's not right, but it's true.
There was an attempted coup after the 2022 election. That is was incompetently executed doesn't mean it wasn't an attempted coup. If the insurrectionists were better-organized and less stupid, there could have well been a civil war.
Right-wingers are more likely to beat you up. Changes the calculus for photographers.
In terms of the trust problem, one easy way to solve it would be to just require real names. Instance admins (maybe also moderators) must post an address, a name, and a (redacted) ID. A registered corporation would also work. Then, they would provide escrow, taking the payment but only giving it to the seller once receipt has been confirmed. The concern would be fraud on the part of the purchaser. There's no foolproof way to fix that, but if both buyers and sellers have "reputation" scores it would be pretty easy to tell if someone's lying.
The admin could also skim 1-2% off every transaction, and then put that into a fund to pay buyers in the case of complaints. That way both the seller and buyer are satisfied, and reputation scores can be used to boot probable fraudsters.
Either way, the system would also allow buyers and sellers to arrange payment in-person, in which case there would be no guarantee needed and the admin wouldn't take a cut.
This system centralizes power in a small number of people who can be sued. Everyone else stays anonymous, and if they're bad actors the admins deal with them. If an admin is a bad actor, their name and address is posted publicly for the world to see. Obvious problem here is that fewer people would want to be admins, but maybe it would be possible to set up a corporate structure where the owner's identity is revealed only if they're being sued -- I'm not a lawyer and you'd have to talk to one. Maybe there could also be a way for them to post records of every transaction in a verifiable yet anonymous fashion, to prove they aren't skimming anything off the top (beyond whatever they say they're taking for server fees).
We can have perfectly secure online voting, if you're willing for all votes to be public. Or we can have perfectly secure and anonymous voting, if you're okay with some secret master list. There are very smart people working on cryptographic voting protocols and I think I would love to live in an online-voting-based direct democracy, but as it stands we don't know how to set that system up.
Maybe we could make publicly known votes work. Athens did it, the early US did it. But there are problems with both intimidation and incentivization, and we'd need some sort of framework to prevent that.
The staffer has two jobs. Their first job is to send a useless mollifying email. Their second job is to make a tally mark next to the words "PORN AGE GATE -- OPPOSED". (Or these days, probably they click a button on a spreadsheet.)
Writing your MP is like voting. It's useless individually, but in aggregate it will change their behaviour.
Wesnoth's very fun too!