Lichess seems to apply the castling rules of Chess360 also in this situation: the result of O-O is the always the same with the king on g1 and the rook on f1. Unsurprisingly, the players didn't do so since the game predates Chess360. They seem to have treated it as a 'long castles on the king side'.
@Magdiragdag
yes, either Lichess is using chess360 rules, or the positions of K and R after castling (Q- and K-side) are just hard-coded somewhere.
By the way, I found the castling rules of that 1935 "displacement chess" tournament - they allowed 5 types of castling actually:
One Step K-Side Castling
One Step Q-side Castling
Two Step K-Side Castling
Two Step Q-Side Castling
Three Step Q-side Castling
In tschaturanga, rajas were set up opposite of mantris. So that is an early variant of the op's set-up :)
This topic has been archived and can no longer be replied to.