sojournTime @sojournTime@social.cleverlibre.org

The 9yo nerdsniped me with a probability puzzle, which I have translated into maths below:

Suppose \(a\) is a random integer between 1 and 20 (inclusive) and \(b\) is a random integer between 21 and 40 (inclusive).

Now let \( c \) be a random integer between \( a \) and \( b \). What's the probability distribution of \(c\)?

Hello to everyone who is showing up because of my Runge phenomenon video! https://youtu.be/F_43oTnTXiw

How the Dunning-Kruger effect is just sugar for the ego (but mathematically wrong):

https://economicsfromthetopdown.com/2022/04/08/the-dunning-kruger-effect-is-autocorrelation/

#science #designofknowledge

@btcprox Let's think about this mathematically. Let \( x \) be the most overloaded symbol in maths and science-- oh wait.

@Pat
> Part of the problem is that people don’t realize that they aren’t voting for president, they never have

A bit pedantic, surely? Your ballot says it's for President, the names on it are those of the candidates... people realise as much as they need to get on with the business of voting. Maybe it's not precise enough for a detailed discussion of electoral mechanics like we're having here, but for casting one's ballot it's a perfectly sufficient mental model. I only made the distinction to try and explain the context to our curious friend :)

> A huge problem with a popular vote is that California and New York will essential pick the president because of their out-sized populations.

No. If we're going to be pedantic, under a popular vote system the states have nothing to do with picking the president; only the individual voters do. The fact that a sizable number of them live in this state or that is incidental, really.

@trinsec @peterdrake @sojournTime

@trinsec @sojournTime Short story (leaving out a few details):

Each state is worth a certain number of "electoral votes". The candidate who wins the most electoral votes becomes President.

Within each state, whoever gets the most votes gets ALL of the electoral votes. So, for example, if 51% of the people in Florida vote for you, you get all of Florida's electoral votes.

One consequence is that all of the political power goes to "swing states" that might go to either party in a given election. Everyone knows that California will go to the Democrats and Texas will go to the Republicans, so they get almost no campaign visits. Ohio, on the other hand, might go either way, so a LOT of campaigning happens there.

(An additional problem is that the number of electoral votes per state is not proportional to the state's population. This is a consequence of consolations made to slavers when the country was founded.)

The National Popular Vote Compact effectively ends this process and gives the Presidency to the candidate who gets the most votes.

@trinsec
To offer a bit more background: the US Constitution was written in the late 1700s. It could take weeks to communicate down the length of the country, which presented challenges we don't see today - organising a single nationwide vote of the general public was logistically impractical. So the framers laid out a multi-step process:
1. The general public casts their votes for Electors
2. Electors meet at the state capital to cast their votes for President
3. Each state sends its Electors' totals to Congress, who adds them up and determines the winner.

Conventional campaigning is about convincing people to vote a certain way at step 1, a "faithless Elector" who votes for someone other than his pledged candidate would do so at step 2, and the events of 6 January a couple years ago were an attempt to influence Congress's determination of the winner in step 3.

Each state has a number of Electors equal to the total number of its Representatives (proportional to population) and Senators (two per state regardless of size), so the framers might have expected the states to use the existing ridings with a couple at-large statewide races for the two Electors corresponding to the Senators, but this isn't codified and the details are left to the states. Two small states still do things this way, but most now award the whole slate of Electors to the candidate with the plurality at the state level.

There's some disagreement as to why most states are winner-take-all. I take the view that it's self-interest: for swing states, this increases the payoff for winning and means candidates spend more effort and money to win its citizens' approval; for stronghold states it means the dominant party assures itself of that many more Electors on its side. But there are others who see this as a purposeful choice by the framers to avoid mob rule, and see this as an invention of great wisdom (or even Divine inspiration) on their part.

Now that we have the ability to conduct a single nationwide election, it would make a certain amount of sense to implement this - but amending the Constitution is really hard. There's a sizable cohort of Americans who revere the framers as religious prophets and consider any suggestion of modifying their writings akin to sacrilege. On top of that, the current system gives a small but durable advantage to less-populous states, who would actively oppose any attempt to weaken their position.

So instead the workaround is that if a group of states which collectively comprise more than half the Electoral College all sign onto this compact, they'll each award their slates to whichever candidate wins the *national* rather than statewide vote, guaranteeing him a majority. As @Pat points out, the legality of the plan to have Congress approve the compact after the states agree is unclear, and circumventing the normal amendment process this way is controversial.

@peterdrake @sojournTime

The National Popular Vote compact has passed house and senate committees in Minnesota!

https://www.nationalpopularvote.com/state/mn

If you live in Minnesota, please urge your legislators to pass the bill:

https://www.national-popular-vote.com/Campaign/MN/MN0

If Minnesota joins, we'll have 205/270 = 76% of the electoral votes required for the compact to take effect!

#uspol #minnesota #NationalPopularVote #EveryVoteEqual #democracy

Back page of the Science Times in today’s New York Times.

PyTorch is not having a very happy end to the year

"... and the kiloseconds..."
"Do you mean milliseconds, Fred?"
"Oh yeah, maybe."
"A kilosecond would be a thousand seconds. I wonder how long that is?"
Pause. Mutter. "Three hundred...". Some nodding. "Sixteen minutes and 40 seconds."
My eyebrows shoot up. "Very good, bud! How did you work that out?"
"I know that five minutes is 300 seconds, and fifteen minutes is as close as I can get. Then I added one minute and 40 seconds."
Genuinely very impressed! #tmwyk

"Is 2023 a prime number, Bill?"
Very brief pause. "No, you can divide it by 7."
Honestly, he worked that out more quickly than I would have. "Can you?"
"Yes. If you add 7, you get 2030. Divide that by 10, you get 203. Add 7, you get 210. Divide it by 10, you get 21, which is 7 times 3."

#tmwyk

IPA

I bet you’ve heard of Galileo & Hubble, but what about Henrietta Swan Leavitt?

Leavitt changed astronomy. She figured out new ways to measure a star’s distance from Earth & her work helped determine the universe is expanding.

Her boss, Edward Pickering, published her findings UNDER HIS NAME. Later, Shapley used her findings to determine distances around the Milky Way w/o credit.

Leavitt’s work is still used today. So next time you hear about famous men in #science share her remarkable story.

@freemo I have no idea what this is, but I have to plug framework. Linux works great (barring very minor and easily fixable hardware accel issues).

They're repairable, thunderbolt 4 capable (works great with the Caldigit TS4 dock), and they're planning on working with coreboot and fully open-sourcing their schematics (for component level repair) soon! I love mine, and I just migrated from a thinkpad, but of course, YMMV.

@peterdrake @chris @s_gruppetta

My #PyConUS2022 talk was on #Python oddities and had this #pythonoddity as the punchline

http://trey.io/oddities

@peterdrake @chris As @s_gruppetta said, it's all about how += works on lists.

+= on mutable objects is a combination of __iadd__ and =,

The __iadd__ call succeeds but the assignment fails

Interestingly += on tuples doesn't work this way! The += operator falls back to __add__ followed by an assignment on immutable objects.

So while

a += b

Is the same as

a = a + b

With a tuple or a string.

Those two lines aren't the same on a list!

This oddity is in the docs too: https://docs.python.org/3.5/faq/programming.html#faq-augmented-assignment-tuple-error

Still one of the oldest #Python oddities I've ever seen 🐍🤔

>>> a = ([],)
>>> a[0] += ["what"]
Traceback (most recent call last):
File "<stdin>", line 1, in <module>
TypeError: 'tuple' object does not support item assignment
>>> a
(['what'],)

Note that an exception was raised but the operation also worked! 😳 #pythonoddity