I'm trying to find functions $f$ over $\mathbb R$ which satisfy $$ f\left(\frac{x+y+1}{2}\right) = \frac{f(x)+f(y)}{2} $$ for all real $x$ and $y$. One thing I immediately note is that shifting any potential function by a constant preserves the above identity, so I can assume $f(0) = 0$ and so $f(\frac{x+1}{2}) = \frac{f(x)}{2}$ for all real $x$. I can then rewrite the left side of the above equality to deduce $$f(x + y) = f(x) + f(y)$$ for all real $x$ and $y$, which is the well-known Cauchy functional equation. Clearly, non-zero affine solutions do not work, so $f$ is either the zero function or is some highly pathological function.
Can such a pathological solution be compatible with the requirement that $f(\frac{x+1}{2}) = \frac{f(x)}{2}$?


Consider $\mathbb R$ as a vector space over $\mathbb Q$. Let $1$ be a basis element and $\vec v$ another one, and consider projection onto the line through $\vec v$. That satisfies $$f\left( \frac {x+1}2\right)=\frac {f(x+1)}2=\frac {f(x)}2$$
And it's "weakly" linear in the manner of Cauchy's equation. That is to say, it is linear as a map over $\mathbb Q$, though not over $\mathbb R$. Of course, it's not identically $0$ since $f(\vec v)=1$.