State resistance based on number of turns or effect magnitude instead of chance

[DoubleX]

In the default RMVXA and RMMV(I don't know if it's the same for the others as well), the state resistance's in the form of decreasing the chance to inflict a state, but without any other change if the state's still inflicted.
I personally feel that it can be excessively and unnecessarily random, especially for states lasting for several turns and with serious effects.
Therefore, my very naive idea is that, instead setting the state resistance as decreasing the chance to inflict a state, the former should be set as decreasing the number of turns of the inflict state, thus eliminating the randomness completely.

Let's say a state reducing a battler's atk by 50% for 3 turns, which is of course a very serious deal.
But there are some battlers having certain resistance on that state, some may reduce the number of turns to 2, and some may even reduce the number of turns to 1, thus removing the situation of a gamble between having no effect and reducing atk by 50% for 3 turns.

Similarly, state resistance can also be implemented as reducing the state effect magnitude, and in the above case, like reducing it to 25% from 50%.

So what do you think about state resistance? Do you prefer randomness or determinism in this case?

[DerVVulfman]

Well, some states literally cancel out others. You can have a Blink state that totally opposes a Poison state. Or a Stun state that negates the Regen state (if one such Regen state exists).

As towards one state weakening or actually reducing the number of 'turns' of another state sounds interesting/ If they are polar opposites, that might not be fully suitable. But I can see that as a viable option with states that may be considered ... Incompatible or Conflicting.