Corvette: uh oh, I spoke to soon

Remember how I said the problem is solved?  Yeah I lied.
Played a game yesterday, and the balls were being ejected in to the shooter lane due to the route 66 kickout switch triggering the trough eject switch erroneously.

But what had  happened?  Open up the playfield, test the switch, and sudenly NOT triggering the trough eject.

So obviously this has something to do with the other normally closed switches on the switch matrix.

Which is here:

highlight are the two switches we focus on. #57 is hit, #37 fires erroneously as well

At this point, I want to point out this great video explaining some theory of a shorted diode that can cause switch matrix issues:

Highly recommended.

So we are forming a box, consisting of:
The switch hit
The switch erroneously firing
a normally closed switch
a bad diode that is allowing current to travel both ways

I wanted to check my box theory and try activating Skid Route 66 Exit (switch 58,) but quickly learned that IT DOESN'T EXIST.
It exists in the manual all right:

But that switch does not physically exist on the game, nor on the prototype photo up at imdb.org.


So great, we have an issue that involves both columns of "normally closed" switches.  I played the game and got it to the point where it was ejecting the troughs due to the ball being up in the route 66 kickout.
Turned it off, went in to switch test, and yes, pressing that switch was now triggering the Trough Eject.
What had changed?

I took the balls out of the trough, their being there or not did not change anything.
I unplugged the race track (switches 51, 52, 55 and 57) and voila, switch 57 was no longer triggering 37.

So then I put the machine in race test and started experimenting with going back to switch test with the cars at various positions, and that's where things confused me more.
There is a switch for each car to say whether they are at the starting line. (51 / 52)
Then there is 55 and 56 which have indicators, but also a number associated with position.
Sometimes the cars return to start and also have the box checked, sometimes not.  Sometimes they return to 0, sometimes to position 1....

In this position, all switches reporting closed, the matrix misfire was NOT HAPPENING:

All switches... closed... solves the problem how?

I am unsure how POS relates to the number beside it.

Since we can have POS > 0  and checked, and POS > 0 and not checked!

And we can have POS not checked at zero, and POS checked at zero.

Gah.  But anyways, I imagine when the box is checked it means the opto is being triggered, which would actually mean the switch is OPEN.
So if all 4 optos are triggered, then the switch matrix box can't be completed.

I will run through the logical combinations another day when I'm not so bleary eyed from lack of sleep.  Needless to say, I'm closer to tracing the bad diode down.  I do not think it is on the racetrack.  My money is on 35 or 36.