Well. I would love to see how does it works.
I tried to figure it out with Explosive Arrow due to it's high initial mana cost.
WITH NO MANA REDUCTION
Empirically:
Base Mana Cost: 18
1.4 Multi Mana Cost: 25
1.5 Multi Mana Cost: 27
1.4*1.5 Multi Mana Cost: 37
Calculations: (Base_Mana_Cost*Multiplayer1*Multiplayer2*...)
Base Mana Cost: 18
1.4 Multi Mana Cost: 25.2 -> 25 (round down)
1.5 Multi Mana Cost: 27
1.4*1.5 Multi Mana Cost: 37.8 -> (round down)
Wtf? Are Mana Cost's rounded down?! Always?
NOW WITH 5% MANA COST REDUCED
Empirically:
Base Mana Cost: 18
1.4 Multi Mana Cost: 24
1.5 Multi Mana Cost: 26
1.4*1.5 Multi Mana Cost: 36
Calculations:(Base_Mana_Cost*(1-Reduction)*Multiplayer1*Multiplayer2*...)
Base Mana Cost: 17.1 -> 18 (round up)
1.4 Multi Mana Cost: 23.94 -> 24 (round up)
1.5 Multi Mana Cost: 25.65 -> 26 (round up)
1.4*1.5 Multi Mana Cost: 35.91 -> 36 (round up)
And now it rounds up
So this theory fails. How does real theory look like?
10% MANA COST REDUCTION:
Base Mana Cost: 17
1.4 Multi Mana Cost: 23
1.5 Multi Mana Cost: 25
1.4*1.5 Multi Mana Cost: 34
Theory that seems to work, but look very... ugly is that reduction itself is calculated and rounded separately - and then Inital Mana Cost is reduced by it:
Inital Mana Cost: (Base_Mana_Cost*Multiplayer1*Multiplayer2*...)
Reduction: (Initial_Mana_Cost*(1-Reduction))
Resultant Mana Cost: Inital_Mana_Cost - Reduction
Inital Mana Cost: 18
Reduction: 0.9 -> 0 (round down)
Resultant Mana Cost: 18
Inital Mana Cost: 25
Reduction: 1.25 -> 1 (round down)
Resultant Mana Cost: 24
Inital Mana Cost: 27
Reduction: 1.35 -> 1 (round down)
Resultant Mana Cost: 26
Inital Mana Cost: 37
Reduction: 1.85 -> 1 (round down)
Resultant Mana Cost: 36
