I'm confused with the time-shifting operation post the time-reversal operation when performing convolution. Let's say we were to convolve $x(t)$ and $h(t)$, so I would have the term $x(k)$ and $h(t-k)$ to solve for. In $h(t-k)$, we first reverse $h(k)$ to $h(-k)$ and then shift it to $t$ units. Ideally, we should shift t units to the left but I find examples where they have moved $t$ units to the right instead.
I'm confused because, for any $x(t+n)$, we shift $n$ units to the left of $t$, so for $x(-t+n)$, why we are moving n units to the right of $-t$?