Greg Dinsmore

Sports, Politics, Philosophy, Media, Data... stuff like that.

[Kareem Carr, Ph.D. @kareem_carr]

A (short) thread on why algorithms like deep learning are almost completely useless for statistics:

[Greg Dinsmore @gdinsmore]

@Noahpinion @japoushter I would love to see this by how people get their news. I suspect that those who follow the news on TV would have a more negative reaction to Russian action and a more positive response to Zelenskyy. TV would make the contrast clearer.

[Patrick Wyman @Patrick_Wyman]

For @TheAtlantic, I wrote about the salt-of-the-earth millionaires who make up America's local gentry, and how their economic and political clout shapes American politics and society. https://t.co/sOHTPclh7W

[Greg Dinsmore @gdinsmore]

@colbycosh @KarlBelanger If that is in fact the most relevant and interesting finding in the study, then I would dismiss it. Because I don't think it's a finding at all. I'm assuming there's other stuff there. Including it in the chart is a judgment call (its absence would also raise questions).

[Greg Dinsmore @gdinsmore]

@colbycosh @KarlBelanger I honestly don't know. I don't know if there are other data points that support that. I could imagine that there is a fair bit of movement between NDP and the LPC/GPC/BQ. The point is that I wouldn't dismiss an entire survey based on a questionable division of a very small number

[Greg Dinsmore @gdinsmore]

@colbycosh @KarlBelanger I would think the detailed results would be wonky for the smaller parties, since he's slicing a tiny portion of the sample. That doesn't mean there isn't anything to learn about the LPC/CPC/NDP vote.

[Jennifer C. Martin @notreallyjcm]

Alex Trebek should have named an heir to continue the royal line so we wouldn't have to deal with this war of succession.

[Greg Dinsmore @gdinsmore]

@IanGraph @ShutdownLine I think there's also a "first do no harm" principle at play. If they're happy with that pairing, don't mess with it. This allows them to focus on the parts they want to improve and not have to rebuild their defense.

[Greg Dinsmore @gdinsmore]

@IanGraph @ShutdownLine Could it just be that it's harder to replace a #4 RD @ 2M than a #3 C @ 3-3.5M? Even if McCann is better, the replacement cost might be lower.

[Scott Galloway @profgalloway]

via @NewYorker

[Greg Dinsmore @gdinsmore]

@bruce_arthur I like this. But slightly less radical... Take the penalty kicks before extra time. Then only count it in the event of a tie.

[Greg Dinsmore @gdinsmore]

@cjpberry @johncutlefish They're the same organisation. B is how it's drawn up. A is the best existing perspective of how it really works. There is another view, A', which is how it actually works and is many times more complex but there is no one position in the org that can see all the connections.

[XKCD Comic @xkcdComic]

Types of Scientific Paper https://t.co/dKRewc3zdQ https://t.co/JtjfOwJUzn

[Soner Cagaptay @SonerCagaptay]

1,000 years of national borders in Europe across the Middle East, overlaid on one map: Czech Republic and England have the oldest permanent borders, followed by large parts of Denmark and France

[Soner Cagaptay @SonerCagaptay]

Where the Romans lived: Every Roman settlement and nothing else

[Greg Dinsmore @gdinsmore]

@JB_Goldstein It does give us less physically gifted athletes a fighting chance. One downside is that it can lead to skill elimination through coaching, which makes for a bad show. Team sports tend to make for better theatre, but there are still bad directors.

[Greg Dinsmore @gdinsmore]

@JB_Goldstein Team Sports are so much more complex just because of the number of players playing at once. There are many games within the game both between teammates and opponents. It's much more interesting for both player and spectator. There are so many ways to win or lose.

[Greg Dinsmore @gdinsmore]

@kareem_carr I think the point (jargon aside) is that knowledge is probabilistic. I don't think that's trivial or widespread to ask not "is this true?" but "how much do I believe this to be true?". I take Bayesian thinking as a shorthand for probabilistic rather than deterministic thinking.

[Greg Dinsmore @gdinsmore]

@DavidLSmith_Iam @OlufemiOTaiwo Aren't they more like biologists talking about medical science? Their concern may be genuine but not immediately useful to the problem at hand. One may criticise them for focusing on abstractions but I'm not sure that's a uniquely philosophical problem.