Should have been a plot to begin with

Easy! Forbes says implied win prob for top candidate, Fabiano Caruana was 36% this year. He feels this is too high, so let's go w/ 30%. Assuming one player could maintain such a form for 3 years, his chance for 3 consecutive wins = 30%^3 = 2.7%. forbes.com/betting/chess/…

I support any message that encourages analysts to regularize their artificial machine learnings. The social obstacle I've found is that better estimators -> less variance -> harder to farm significance. So we need diverse messaging and some peer pressure probably too

Fantastic advice, especially bc it contradicts so much other fantastic advice (writing is thinking; shitty first draft-principle). The moral: see what others do, then figure out what works for you (today) ((for this task)).

Thread of general #BM100 thoughts from this year. To start, how were there so many finishers? Not intentionally being a smartass, but, well, it's because there were so many finishers. Had we all been working alone, there might have been 1-2 finishers if even that. (1/10)

Many challenges and problems w/misinformation, but some #fightingmisinformationonline participants asked if there are also reasons for optimism?

Here 3 from Reuters Institute research

1) news outlets rated 'trustworthy' see much more engagement on web & social than untrustworthy 1/3

A sample of university professors miraculously all managed to overcome a series of grad school rejections.

The most important and fun to follow sports event of the year is happening atm. Follow the Barkley Marathons here ->>

If you think academese is better, it is not. Academese is terrible. You'll think better and sound better if you write like a normal human being.

On writing for normal human beings: amazon.com/Sense-Style-Th…

Someone should write 'Titles,' a paper about how economists come up with paper titles.