It is a statistical impossibility to score a zero without knowing that you are purposefully choosing nearly every wrong answer. Perhaps the “Average Min” should be “Absolute Min” at Zero.
If there is a large group writing this test and getting lower than a 33.3%, I just don’t understand what happened to them. Apparently their midpoint is 25%, but that is less than randomly filling in the answers. Perhaps that lower slope should flatten out, a sort of “glass floor”. Start with 40 as your midpoint and work your way up from there, see if that helps explain things similarly but with a little more cushion towards the 33.3%.