Thanks for detailed explanation. I see what you are saying. The question I ask myself though - is it really a problem? Defensive artillery was kinda designed this way: it cannot just be ignored, you must do something about it. You can bypass the defenders and attack it, or suppress it with your own artillery, or attack it from the air, or force it out of ammo, but you must do something. Also, you are right that for this particular attack defending unit is irrelevant. But there is a big difference between 10-snt infantry and 1-strength Panther after you have dealt with artillery. So, the first option still gives you double defense, while the second - only single defense.deducter wrote: The defending unit becomes completely irrelevant at this point. I could garrison a 1-strength Panzer I in the city and the attack is worthless, since it is completely broken up by the defending artillery.
It can be argued that artillery in general is too strong against infantry, both on offense and defense. From all the evidence we have accumulated, it might well be the case. But then, shall not we try to solve this problem? Even if we have cover, the attacking unit is in the open terrain, and so according to general rule it should not be subject to cover? Or do you suggest to assume that the attacker enters defender's hex first, and then artillery shoots at both units (but somehow damages only attacker)?
This is true. With base entrenchment removed, units will retreat/surrender more often. It can be argued though that a unit which has just entered a hex must be made to retreat from there relatively easy too. So, this change might actually benefit gameplay.deducter wrote: That's a very good point. However, entrenchment does serve another purpose in the game: surrender/retreat mechanics. If you completely change the entrenchment system, you'll also have to tweak surrender/retreat. Adding a new variable has the advantage of being an easily tweaked independent variable. But it does have the disadvantage of added complexity. Changing an existing variable is fine too, but I imagine it'd take quite a bit more work, and you'd have to carefully adjust all other values which depend upon this variable.
Anyway, the change which you propose is clearly safer, so this is a good point.