For which integers $n$ does $$(t-n)^2+(t-(n-1))^2+(t-(n-2))^2+...+(t-1)^2=\sum_{i=1}^ni$$ have solutions?
I have managed to find an answer to this, but it is probably wrong (example later). Note that this is not a textbook question; rather one I've created.
Attempt
I will give a very brief answer to this.
The LHS can be written as $$\small(t^2-2nt+n^2)+((t^2-2nt+n^2)+2t-2n+1)+((t^2-2nt+n^2)+4t-4n+4)+...+((t^2-2nt+n^2)+2(n-1)t-2n(n-1)+(n-1)^2)$$ which can be simplified as $$n(t^2-2nt+n^2)+n(n-1)t-\frac16n(4n^2+9n-1)=n(t^2-(n-1)t+\frac16(2n^2-9n+1))$$ using the sum of squares. So our equation can be expressed as $$n(t^2-(n-1)t+\frac16(2n^2-9n+1))=\frac12n(n+1)$$ giving $$t^2-(n-1)t+\frac13(n^2-6n-1)=0$$ The equation has solutions if the discriminant is not less than $0$: $$b^2-4ac=(n-1)^2-\frac43n(n^2-6n-1)>0$$ when $n<7$ by Desmos.
(Counter)example
Let $n=7$. Then the equation becomes $$7t^2-56t+140=28\implies t=4$$ which is a solution. This contradicts my attempt and the result obtained by Desmos.
Can anyone spot the error in my answer?