The score boundary can be solved directly in input space. With w=1 and b=-2, setting z=0 gives the exact boundary x=2.
highlighted = computed this step
Start with the score rule
The displayed rule is z = w*x + b. Here w=1 and b=-2, so the score rule is z = 1*x - 2.
z=1⋅x−2
Set the boundary score to zero
The decision boundary is the input where the score z equals 0. Put z=0 into the exact score rule.
0=1⋅x−2
Solve for x
Adding 2 to both sides gives x=2. That is why the plot marks the boundary at x=2.
x=2
Check the displayed rows
The table recomputes scores -2, -1, 0, and 1. At x=2, z=0. This is exact boundary bookkeeping for the displayed score rule only, not a probability or a claim about later rows.