The general idea here was to take on the challenge of only using Excel cell equations to simulate Go. In Go, a set of stones on the board is considered a group if all stones are path-connected to each other along horizontal and vertical lines (not diagonals). While conditionally coloring the cells was not hard, determining which stones formed what groups was non-trivial. I had to implement breadth-first search in Excel by duplicating the board state a number of times equal to the maximum stack size, and having each copy of the board state iterate/depend on the last. A lot of cells were used :)
Also looking for feedback from experts as to how I could have made this better/more efficient! Thanks!