Hi everyone, I think the root cause is that the IBM Game Control Adapter™ was meant to be versatile.
Originally, maybe, there was a need to replicate paddle support, like with pong consoles.
The C64 had X/Y inputs (via SID), too, for that purpose same (paddle support). They were later repurposed to interfaces mouses, err, mice.
On the original IBM gameport card, there's a provision for custom circuits:
At the end of the card, there's a veroboard-like section..
The attachment 5150_5160_game_adapter.jpg is no longer available
Edit: There's another version with a provision for a second DE15 port.
I'm just mentioning this, because the gameport was the #1 experimental port of my childhood (QB45, VB1 on Windows, Turbo Pascal etc). 🙂
Out of all ports, it was the most plug&play type one. You could remove/add cables on the fly to it without worrying.
Because, the buttons were just switches anyway and the X/Y axises were meant to see varying resistance/conductivity.
This meant that almost everything from zero to max was being accepted on the inputs.
A short (buttons) or sudden power/voltage spike (axes) was essentially part of the design.
"Time, it seems, doesn't flow. For some it's fast, for some it's slow.
In what to one race is no time at all, another race can rise and fall..." - The Minstrel
//My video channel//