It is 1925 and you need to simulate data from a normal distribution. What do you do? Obviously have somebody (probably some poor student) draw 5000 cards out of a bag. From Tippett (1925, https://doi.org/10.1093/biomet/17.3-4.364). #statistics #history
ON THE EXTREME INDIVIDUALS AND THE RANGE OF SAMPLES TAKEN FROM A NORMAL POPULATION

L. H. C. TIPPETT, B.Sc.; ON THE EXTREME INDIVIDUALS AND THE RANGE OF SAMPLES TAKEN FROM A NORMAL POPULATION, Biometrika, Volume 17, Issue 3-4, 1 December 1925,

OUP Academic
Now we just do rnorm(5000) and call it a day. But what will people say in 100 years about us?!?
@wviechtb "Back in the early 2000s, they didn't realize that 1234 is an extremely atypical seed"

@sharoz

grep --include=\*.r -r -F "set.seed(1234)" | wc -l
298

grep --include=\*.r -r -F "set.seed(1)" | wc -l
73

grep --include=\*.r -r -F "set.seed(42)" | wc -l
5

Oh no ... 😳

@sharoz @wviechtb I'm intrigued ... what is atypical about that seed?
@lukaswallrich @wviechtb nothing that I know of. It's just very popular. So if something about it is especially weird, it'd be very impactful.
@sharoz @lukaswallrich @wviechtb they might find out that the seeds themselves were not following a normal distribution
Back then it would apparently be another 30 years until you'd at least be able to simple look up random numbers in a book...: https://en.wikipedia.org/wiki/A_Million_Random_Digits
A Million Random Digits with 100,000 Normal Deviates - Wikipedia

@croever @wviechtb
It is 1927 and you need to randomise. What do you? Tippett has come to the rescue by poring through Parish records...
@wviechtb my undergrad advisor talked about a simulation study he did in which he genetated draws from a binomial by shaking pennies in a shoe box and counting the number of heads πŸ˜‚
@collinedwards Life was so much easier before computers!