Smirkley (@Smirkley) header
Smirkley (@Smirkley) profile picture

Politics + data + charts + randomness

Joined September 2023
Photo by Smirkley (@Smirkley)Photo by Smirkley (@Smirkley)Photo by Smirkley (@Smirkley)Photo by Smirkley (@Smirkley)Photo by Smirkley (@Smirkley)Photo by Smirkley (@Smirkley)Photo by Smirkley (@Smirkley)Photo by Smirkley (@Smirkley)Photo by Smirkley (@Smirkley)Photo by Smirkley (@Smirkley)
Moderate Rs were more likely to lose than conservative ones. We can safely assume moderates aren't randomly distributed but sit in swing seats. Control for Trump share, a proxy for how conservative a district is, and ideology explains nothing. Conservatives did slightly worse.
The blue wave hurt moderate House Republicans the most by far. The most moderate third of Republicans lost 41% of their races, compared to only 7% of the most conservative third of the party.
14
61
5,739
Imagine looking at this chart and thinking the sensible thing to do is to increase immigration to a record level. JFC 🤦‍♀️
24
77
2
335
6,219
Smirkley retweeted
The post compares polling-station data from recent elections. It shows Russia's results have a strong positive correlation between turnout and the winner's share, thousands of stations with extreme 90%+ turnout and 80%+ support for the ruling party, and far more whole-percentage results than chance would produce. These patterns are widely used as statistical red flags for ballot stuffing or fabricated numbers. Canada's 2025 data shows none of that: near-zero correlation, only 11 extreme stations out of ~66,000, and no excess whole percentages. Taiwan and North Carolina look the same. On these metrics, Canada's results match fair elections, not the anomalous Russian pattern.
1
1
227
Imagine thinking Russia has fair elections.
Replying to @DonJBacon
Putin was just reelected while zelensky refuses elections, maybe cia shouldn’t have led regime change in 2014.
27
89
6
2,668
151,631
Some say correlation alone isn't evidence of rigging. It is highly suspicious, though, especially when the correlation climbs from 0.24 in 2000 to ~0.7 since 2011, and held at 0.68 during a largely unsuccessful war with high inflation and interest rates. What about other methods to detect fraud? You can plot each station by turnout and the winner’s share, and a fair vote will form a single blob. Russia goes to the upper-right corner. There are 5,221 stations reporting 90%+ turnout & 80%+ for United Russia. Canada: 11 of 66,000 polls. Taiwan & North Carolina: 0. If that doesn't convince you, there's the integer grid. When officials make up results, they tend to pick round numbers, so results pile up at whole percentages. You can simulate each station's actual vote counts to show how many whole-percent results chance alone would produce. Russia had 1,600 more stations than chance for vote share and 1,200 more for turnout. By comparison, Canada, Taiwan & North Carolina had none above chance.
2
12
989
Since 2025, wages are up 6.4% and groceries 3.8%. So down after wage growth. Yes, Trump promised to bring grocery prices down, implying they would be down in nominal terms. That was an unrealistic promise. Grocery prices fell nominally in only 2 years since 2000.
Replying to @PpollingNumbers
Only 19% approve on cost of living? I'd honestly love to meet one of them and ask what they're paying for groceries.
5
11
54
1,507
Wages have generally grown faster than grocery prices. The big exception is after COVID:
1
6
310
Proud to say I don’t get paid to shill for issues or causes
🚨NEW: We just uncovered a covert influence operation using right-wing creators to promote Nvidia chip sales to China. Earlier this year, we reported on similar campaigns, and Republican lawmakers went after Nvidia. But the story didn’t end there. 🧵
5
6
102
9,276
Young people started trying to cut back on their social media use around 2021/2022 And then young peoples' mental health started improving SHOCKING STUFF HERE
Replying to @adamboxer1
you might want to look into what young people are actually doing today before making that claim
3
6
1
86
3,148
THE WIFE-STABBING AFGHAN GERMANY JUST CAN'T QUIT In 2012, an Afghan man entered Germany with his family. In 2013, the family returned to Afghanistan with a substantial cash bonus and prepaid flights thanks to a "voluntary return" program organized by the state. In 2014, the authorities rejected his asylum claim for good measure. However, the family decided life in Afghanistan was no longer to their taste, so they returned to Germany in 2016. The man's asylum claim was rejected again in 2017. What this means is that the man received a letter in incomprehensible German legalese informing him he had no right to remain in the country and would he please return to Afghanistan. Of course, nothing was done to actually implement the "notice of deportability". In 2019, the man stabbed his wife 18 times. For this, he was sentenced to 6 years in prison, a typically lax German sentence. The court found he had tried to murder his wife because he believed she was integrating into German society too well and thus might be tempted to have an affair. At this point, Germany decided it would be insane to allow a man to stay in Germany when his asylum claims had been rejected twice, he had already been ordered to leave the country, and he had been convicted of a violent crime on German soil. Therefore, they decided to do that insane thing. Upon his release from prison, he was permitted to remain in Germany under "toleration" (Duldung), a bureaucratic loophole that allows deportable people to stay in Germany indefinitely. Yet with its other hand, the German state (in this case, the foreigner agency of the city of Trier) ordered the man to be deported. The city argued that even though there is a general policy of not returning people to Afghanistan, this presumption can be overturned in cases of people who have been sentenced to a crime with a prison sentence of over two years (which is the equivalent of a ten-year sentence in normal countries). Trier's arduous efforts to rid itself of this foreign predator were recently rewarded by the Administrative Court of Trier, which on November 6 rejected the man's appeal against the deportation order. The court noted the man had not expressed remorse, continued to believe in the backward ideology which prompted his attack, had never attended a German-language course, and not sought psychotherapy (which the German legal system quaintly still believes in) to help him deal with his wife-killing tendencies. Of course, the man can appeal this decision and almost certainly will. He has also filed a separate legal action demanding that Germany implement an absolute ban on deporting people to Afghanistan. So sometime in 2027, it is just possible that this foreign violent criminal will, after hundreds of thousands of Euros in legal proceedings, actually be deported to Afghanistan. Of course, he will still have many further opportunities to fight the deportation, and even if all those fail, can easily disappear into the underground. But perhaps Germany will get lucky this time. And then, the guy will return once again. welt.de/vermischtes/kriminal…
6
16
4
82
2,612
Someone doubted my Spotify claim. Spotify revenue outside the US is basically the audiovisual exports of Sweden in the original chart. It doesn’t mean Sweden is a cultural powerhouse (they may be). It means they have the world’s number one music streaming platform.
Did ABBA get back together? Spotify, which charges subscribers from Stockholm, streams Taylor Swift in Germany or Australia or Canada, and this counts as a Swedish export. If a US studio makes a movie in the UK, because of in insane tax credits (corporate welfare), who gets the export? UK. Who gets the profits? If Netflix sells to viewers overseas, which tax haven do they use? The Netherlands. Its Dutch subsidiary took in €17.5 billion in 2024, more than all US film and TV exports ($16.4 billion). None of the above counts as a US export.
1
27
1,779
If you can't see what's wrong, Putin’s or his party’ share climbs with turnout. You can literally see the made-up roundish numbers forming grid lines. 100% turnout stations vote 100% Putin. What Canada’s results show is what an honest country looks like. You can see this mostly in earlier Russian elections. Note the lib shape and lack of grid lines.
2
1
212
8,510
Autocorrect. Lib = blob
23
7,645
Naming kids after recent popular culture references is nothing new, of course. Here's naming patterns for boys around the Civil War! Many prominent figures from the time saw their names suddenly rise in popularity during and just after the war!
Restricting to rare names to ensure we're measuring actual Biblical knowledge instead of just "Paul" and "Mary," etc, I find it interesting to see which Testament is driving naming more in America. Currently, Old Testament names are very IN for Christian babies!
3
15
1,903
Did ABBA get back together? Spotify, which charges subscribers from Stockholm, streams Taylor Swift in Germany or Australia or Canada, and this counts as a Swedish export. If a US studio makes a movie in the UK, because of in insane tax credits (corporate welfare), who gets the export? UK. Who gets the profits? If Netflix sells to viewers overseas, which tax haven do they use? The Netherlands. Its Dutch subsidiary took in €17.5 billion in 2024, more than all US film and TV exports ($16.4 billion). None of the above counts as a US export.
One of the biggest stories in the world today is how American global cultural influence has evaporated Not good!
8
8
2
125
9,407
The correlation between vehicles per person and GDP per person across 138 countries is 0.81, or a staggering 0.90 weighted by population:
4
7
57
2,605
Smirkley retweeted
Take two: Why were moderate Republicans more likely to lose than conservatives ones? The most moderate 1/5th fell 14.3 points from Trump 2024 and the most conservative fell 14.8. Or no difference. Moderates lost more seats because they had less of a margin. They were in Trump +16 districts, vs Trump +27 for the most conservative.
3
7
39
3,098
Smirkley retweeted
Replying to @saraaa7447
The US has nothing to do with this. Remove the US from the chart, and the correlation between cars per person and income per person goes from 0.81 to 0.81.
1
6
1
35
757
Come on, son. Don’t go by your own taste but look at revealed preferences. US households earning under $15k average 0.9 vehicles. Those earning over $200k average 2.7, three times as many. When people have more money, they choose to own more cars. Poor people in poor countries don’t choose, on average, to own fewer cars. They just can’t afford them.
Why is cars per people there and why is more better? 😐
9
6
1
66
2,327
Elbridge Colby says strikes on “civilian infrastructure” (because none of these are dual purpose) have probably hardened the resolve of the Russian public. This is, of course, false: Also, it’s hilarious that Colby is taking this stance on Russian warehouses and refineries. I mean, Russia, the invading country, has been attacking Ukrainian civilian infrastructure, schools, and hospitals, many with no dual purpose, from day one. The asymmetry of attacks here speaks volumes. There have been 3,000+ attacks on health care alone.
❗️US Deputy Secretary of Defense Elbridge Colby criticized Ukrainian deep-strikes at the Ramstein meeting, specifically targeting logistics centers like Wildberries that were actively supplying drone components and ammunition. He stated that such strikes fail to achieve the goal of inflicting suffering on Russians and instead provoke greater escalation from Moscow, so allies should focus on strengthening the front line and developing capabilities for strikes in operational depth rather than deep strikes.
4
17
2
70
2,925
Take two: Why were moderate Republicans more likely to lose than conservatives ones? The most moderate 1/5th fell 14.3 points from Trump 2024 and the most conservative fell 14.8. Or no difference. Moderates lost more seats because they had less of a margin. They were in Trump +16 districts, vs Trump +27 for the most conservative.
3
7
39
3,098