Let’s go ahead and ask R core to make namespacing case insensitive so pKg:: and pkg:: are the same thing. What could go wrong,,,

I actually reached the point a while ago where I think package names should only be 3-6 characters long because I prefer explicit pkg::function namespacing. Having R (or lowercase) or py in the library names is just irrelevant additional characters in my code.

Please note that "0%" doesn't equal "eliminated". It just means that these teams didn't make the playoffs once in my 50k simulations.

Here's #1 seed leverage going into week 11 for some teams. No complete chart because it's not ready to get published and I don't have time to fix it

I do prepare this chart, but usually not that early in the season

I'm glad this lines up pretty well with my numbers bsky.app/profile/mrca...

Thank you!

Full transparency: I wasn't the first to come up with such a circle. They date back to the early '10s I believe.

I tried 3-4 past seasons but didn't invest much time. I think it usually takes about 140-160 games for a circle to complete. Obviously requires all teams to at least win one game and lose one game.

I used to say "Multipass" in my head whenever I pulled my credit card to pay for something

Haha awesome. I noticed you did quite a lot in dplyr/vctrs recently. Great to see those improvements. Thank you!

Do you think there’s a better way to do that?

Since you are thinking about verbs. I had this usecase I would like to overwrite the variable "b" to the value "D" where "a" == 2. Otherwise don't do anything.

I think the only dplyr solution is using if_else including the "missing" arg.

The data.table solution was so much more straightforward

Yeah that’s great. Definitely a great improvement to the syntax

Would it be possible to combine when_any with mutate?

The fact that you did that instead of some random #rstats nerd makes it 1000 times funnier to me.

Agree, but also this from my latest PR in nflreadpy

72% average series conversion rate is crazy. Defense is so hard.