Spoilers for the >!Family Core!< puzzle, diving into the mechanics of >!numeric core!< calculations.
The puzzle deals with the roman numerals from >!Wine Cellar!<, >!MCCXIII!< and >!Blue Tents!< memos instruct you to >!split the number into four parts!< like this:
>!M, CC, XI, II = 1000, 200, 11, 2!<.
If you do the >!numeric coring!< on these parts, you arrive at the game's solution value, >!53!< (achieved with >!1000 / 200 * 11 - 2!<). However...
If we look at the >!parts as 1 long number that we're seeking the numeric core for!<, >!1000200112!<, a >!smaller!< result can be achieved with >!1000 / 200 - 1 * 12 = 48!<, which makes the game's intended result technically wrong according to the explanation in the >!numeric cores!< document from >!the file cabinet!< in >!Foreman's Office!<.
I discovered this through an online calculator tool (not mine). I wanted to see test that it works as expected, put in >!1000200112!< and got 0 (not a valid result). The project is open-source and says that pull requests are welcome, so I fairly simply fixed it, went to test again and sat there pretty taken aback, staring at the "technically correct but not what the game wants" result. And I was even more shocked when I found 0 previous discussions about this - nothing ANYWHERE. Have I made a mistake somewhere along the way that I fail to keep discovering?
And now I'm facing a dilemma. Since this tool comes up as a high result when searching for >!"numeric core calculator"!<, I'd like to improve it with the fix. But when the "correct answer" is actually wrong for the only purpose people will ever want to use it, I guess it's better when it just continues showing the obviously wrong 0...
EDIT: An explanation of what I'm doing so outrageously wrong to warrant downvotes over a discussion would be appreciated. I'm genuinely confused. Of all places, you'd think a community of people who enjoy puzzle games would like things having consistent logic and working towards understanding something.