Why do most modern zero knowledge proof systems convert the computations to polynomial equations ?

Let's get started : Assume we have a problem with five computations, now to verify these computations, the verifier will have to execute all those 5 computations and compare the computed value with th