Made some minor changes to make it easier to diagnose the problem.
- added hex digits A-F and revised the score to show the paddle reading on the left and the converted value on the right.
- Disabled the movement of the human/celery
In order to do this I had to bump up the ROM size to 4K.
Highest paddle reading is $A1 which is converted to $30
"Lowest" paddle reading is $1E which is converted to $5A
"Lowest" because turn the paddle a little further and the reading jumps to $31 with a converted value of $0B
Looking in the comments I see:
; Paddle1 will have a value from 49-164
; We need to convert it to 11-49
49-164 is $31-$A4 in hex.
11-49 is $0B-$31 in hex
So something happened that changed the range of paddle readings from $31-$A4 to $1E-$A1. Plus there's a "hiccup" that occurs causing the reading to jump from $1E to $31.
What you need to do now is a binary search
to figure out in which version of Celery the problem first appeared. To do a binary search establish your MIN and MAX versions, then test the one in the middle. With 50 versions the MIN=1 and MAX=50.
(MIN+MAX)/2 = 25 - so the first version to test is 25.
Does 25 have the problem?
- Yes - your new MAX is 24
next search is (1+24)/2 = 12
- No - your new MIN is 26
next search is (26+50)/2 = 38
Repeat the process into you find the version the problem first appeared. Then compare that source with the one before it - so if the problem is in version 40, compare versions 39 and 40. Use a tool like WinMerge
to make it easier to figure out what changed.