As long as almost any ground unit can move at least 6 hexes in one turn and with the current railway movement there is no way a "race to the sea" can happen. As soon as the proposed neutral strip would be unlocked the first side to move would reach the coast in one turn with a garrison/cavalry held in reserve just for this purpose, which can hardly be called a race.
Instead of introducing yet another unit it might be better to significantly raise the cost and effectiveness of the "infiltration tactics" upgrade. Maybe +10 to assault and shock at 20 points cost? Thing is: the AI would rarely apply this since it's constantly stripped for cash by building lines of infantry 5 ranks deep. I rarely see it building tanks at all even after lowering the cost to 30pp.
Btw: the ratio of "special units" to infantry (the default type of unit to produce for the AI) seems to be controlled by this part in ai_economics.lua
Code: Select all
if info.frontlineGaps < 0 and canAfford and unitsInProdQueue < 1 then if protoInfo.name == "fighter" then if IsTechCompleted("armed_fighter", faction) then score = score + 7 + game.turn%3 else score = score + 5 + game.turn%3 end elseif protoInfo.name == "bomber" then score = score + 7 + (game.turn+1)%3 elseif protoInfo.name == "armour" then score = score + 7 + (game.turn+2)%3 end score = score + (faction.pp/1000) end
I guess my current problem with Russia building no units at all might be connected with having changed those values too much or with the AIs inability to define "fronts" when you advance too far and too fast into Russia.