Huh? The likelyhood that an average comes out to divide a population exactly 50/50 seems low. You'd expect it to be off by a point or two, at least. I'm just pointing out that your statement that 50% of X are below average is not usually true. Unless you meant at least 50%. In which case it's a: misleading, B: still wrong.
You've simply conflated symbols (ie, ≈ ≠ = etc).
The original uses shorthand "~=" for coarse equality, in lieu of {option+x}.