@user21820HATESSMOKING-HATS I am trying to find a small enough x. Here are my naive attempts:

Test values for x:
x = 1/(1+min(m,δ)),
x = 1/(2+min(m,δ))  ,
x = 1/(1+min(|m|,δ)) ,
x = 1/(1+max(|m|,δ)) ,
x = 1/(n+max(|m|,δ)) ,
x = 1/(1+m) ,
x = δ/(1+mδ) ,
x = m/(1+mδ) ,
x = 1/m ,
x = 1/(δ+|m|)
x = 1/(1+min(δ,|m|,1))
x = 1/(1+min(1,δ))
x = 1/(1+min(1,m))
x = 1/(δ+min(1,m))
x = 1/(m+min(1,δ))
x = 1/(p+min(1,δ)) [where p ∈ ℕ]

@Prithubiswas Why do you even try so many? Why not just see what you need in order to prove what you want?

What do you want to prove, first of all?

@barlop Well, in true mathematics nobody asks such a question, which is why I had to check wikipedia for what non-mathematicians mean by "percentage". According to that article, a percentage includes the "%", so the answer to your question ought to be 50%.

It is simply not mathematics, in the same way that asking "What is the IEEE floating-point representation of a half?" is not mathematics. As long as you know the mathematics, it's good enough. Also, as you guessed there may be potential ambiguity due to the flaws in natural language, just like "How many apples did you eat?" could be answered by "7" or by "7 apples".