Also. The amount of quality skill-based instruction available for free on youtube these days in *wild*, as is what an app, a smartphone with a camera, and a tripod will do for you (mind-blowing to have this kind of shot tracking at home!) Make me wish I was a kid learning to play all over again!

Anyways. For those watching: this isn't about any particular event/outcome for myself. Was just talking with a friend who's struggling to get an idea out with (IMO) obvious value, and it triggered this avalanche of thoughts about all the silly things preventing him from moving forward.

That, after having to convince 3ish people (with varying expertise) that the work is worth doing/we have a good plan for doing it/we are capable of doing it, and, if successful, spending the time/energy/money to do it, to convince another group of 3ish people (with varying expertise)...

Run, run, as far away as you can for your inbox. I have been about 15% as successful at this as I hoped to be, but every bit of that 15% has been the highlight of the year.

The beautiful, unhinged, frivolous, creative, nonsensical result of open-source

So that's out. What's left? Wait until an "in the wild" version of the existential threat plays out at scale, and A) we can either detect it (and when we do, FML); or B) we can't (and unbeknownst to us, FML).

Those three options leave a lot to be desired...

Like, my immediate reaction was, "Would you consider it friendly for me/someone else to attempt to program such a botpool to invade your studies, and later reveal this to you, to show you this could straightforwardly be accomplished?" But that would be awful to convince by sabotage.

My problem with this problem--and the "eh, I doubt this happens frequently enough/we have the current tech for this doomsday scenario to play out..." reactions to it--is that I'm not sure how the frequency/ease of the problem could be convincingly demonstrated in a way that wasn't massively harmful

Andrews gets major edge for beautiful, thoughtful design and intuitive context (the Dems vs. GOP context will immediately click for many)

Felix's, though rougher/less intuitive, has an edge on "QRP realism", as I think cherry-picking outlier obs + "study didn't work: run again" were common QRPs

I also enjoy the p-hacker app by Felix Schönbrodt (@[email protected] on Mastodon): www.shinyapps.org/apps/p-hacker/.

There's also his companion p-checker app: shinyapps.org/apps/p-check..., useful for showing how p-hacked effects do (e.g., TIVA)/don't (e.g., Meta) "pop" in credibility metrics

So much of your career chances are determined by whether you are working with a successful and supportive team.

I.e. Whether you are working with a line manager or mentor who hands down opportunities. Coauthorships on papers. Coapplicants on grant.

If you're not being given these, you MUST move.

That being said, I think some of the same magical patterns could be fit to certain initiatives/efforts in the OS movement. So I say all of this as a rabid supporter of some OS products (e.g., RRs) and a skeptic of others (e.g., preregistration outside of RRs) 🤷‍♂️

And I don't mean that figuratively. I mean people in the world today who think they are reality-bending sorcerers would look at some of how our business is conducted, and think "Those people have some serious $%&*ing magikal capabilities!"

To go a step beyond "storytelling"--and I think I made this argument, unhinged, in a conference room with @rogerthegs.bsky.social a handful of years ago--there is a legitimate and distressing overlap between A) contemporary social science practices and B) contemporary schools of *magical* practices

This looks awesome, though I have a very, very, very fuzzy memory of @minecr.bsky.social once-upon-a-time writing a really persuasive thread or piece (can't remember) to the effect of: we should stop trying to duck R's built-in contrast system (and bad things happen when we do). Ring any bells?