geneiusxie
05-16-2014, 10:51
Hey guys,
(I'm new to the forum, but been playing this game since it came out 7 years ago)
I have a pretty good feeling that this is different from combat from games like Age of Empires (where each attack does the same amount of damage, every time, no randomness involved) because I see, e.g., soldiers can die in one hit from a longbowmen at times, but the guy next to him (who should have identical stats, but not sure) might take 2 or 3.
I'm not sure, but I thought individual soldiers had hit points (which would have to be decimal numbers), so once they get non-fatally hit, they would be more likely to die from the next hit. But when I think about it more, when playing with elephants, which (I assume) have to receive 6 fatal hits to actually die, or Hospitaller Marshall, which have 2 hit points, when they are under missile attack, (or when all their soldiers are somewhat equally engaged in melee combat,) they rarely lose any men at first, but after their clothes turn bloody and stuff, they start dying really fast. (Easily exploitable by hit and run attacks for minimal losses)
Normal soldiers can start dying under the first missile barrage. So I guess most wounded soldiers (those that look bloody) aren't really any weaker (of course, their attack may be decreased, but I'm not sure how to really test that)
I've always assumed that how powerful an unit's attacks are scales with the square of their attack points, so a 10 damage arrow does 4 times as powerful as a 5 damage one. But according to
forums.totalwar.org/vb/showthread.php?75233-How-the-Combat-System-Works
it looks like it's actually a logarithmic scale so a 10 damage arrow would be 1.2^(10-5) = 2.488 x as powerful
BUT that's for Shogun 2. AFAIK, that uses the engine from Empire, which might have changed that stuff.
The reason I think this would be wrong is because that would mean a retinue longbowman with 8 missile attack that hits a dismounted chivalric knight with 8 defense (2 from armor, 6 from shield) would only have a 1.9% chance of killing it, and if soldiers can only be either dead or alive (not wounded with partial hp left) then they would be able to walk up to the longbowmen with only a few losses, (obviously not true, at 120 unit size, one unit can easily lose 10 or 20+ men per volley at close range)
Also, this would make units almost impossible to kill, unless they're attacking over 10 times/second.
This also means that the shield still provides most of its protection, even from the sides or back while the in-game advisor kept telling me (before I shut him up :beam:) that the shield does nothing (although this wouldn't be the first time the in-game advisor has lied to me before, not just from this game)
This would also mean that the "ap" or armor piercing attribute means (target armor-1)/2 instead of just target armor/2, which is what export_descr_unit.txt told me in the description (and I think the in-game advisor may also have told me that at some point)
I think this is really important for deciding which soldiers to recruit, especially for deciding whether to use 1200 cheap ones or 600 really expensive ones (in battles I have seen my expensive knights and heavy inf get creamed by hordes of peasants, even without ap weapons. So why even bother to build expensive stables when mounted sergeants would do?
Also important for mods - this would make the Realistic Archery mod unbalanced, since he multiplied the attack points of all arrows by 3, so longbows do 18 attack instead of 6, peasant archers do 15 instead of 5, etc. but he really should've added 6 to everything (since we're actually dealing with a logarithmic scale) to triple every arrow's damage so all archers get equally boosted.
-Gene
(I'm new to the forum, but been playing this game since it came out 7 years ago)
I have a pretty good feeling that this is different from combat from games like Age of Empires (where each attack does the same amount of damage, every time, no randomness involved) because I see, e.g., soldiers can die in one hit from a longbowmen at times, but the guy next to him (who should have identical stats, but not sure) might take 2 or 3.
I'm not sure, but I thought individual soldiers had hit points (which would have to be decimal numbers), so once they get non-fatally hit, they would be more likely to die from the next hit. But when I think about it more, when playing with elephants, which (I assume) have to receive 6 fatal hits to actually die, or Hospitaller Marshall, which have 2 hit points, when they are under missile attack, (or when all their soldiers are somewhat equally engaged in melee combat,) they rarely lose any men at first, but after their clothes turn bloody and stuff, they start dying really fast. (Easily exploitable by hit and run attacks for minimal losses)
Normal soldiers can start dying under the first missile barrage. So I guess most wounded soldiers (those that look bloody) aren't really any weaker (of course, their attack may be decreased, but I'm not sure how to really test that)
I've always assumed that how powerful an unit's attacks are scales with the square of their attack points, so a 10 damage arrow does 4 times as powerful as a 5 damage one. But according to
forums.totalwar.org/vb/showthread.php?75233-How-the-Combat-System-Works
it looks like it's actually a logarithmic scale so a 10 damage arrow would be 1.2^(10-5) = 2.488 x as powerful
BUT that's for Shogun 2. AFAIK, that uses the engine from Empire, which might have changed that stuff.
The reason I think this would be wrong is because that would mean a retinue longbowman with 8 missile attack that hits a dismounted chivalric knight with 8 defense (2 from armor, 6 from shield) would only have a 1.9% chance of killing it, and if soldiers can only be either dead or alive (not wounded with partial hp left) then they would be able to walk up to the longbowmen with only a few losses, (obviously not true, at 120 unit size, one unit can easily lose 10 or 20+ men per volley at close range)
Also, this would make units almost impossible to kill, unless they're attacking over 10 times/second.
This also means that the shield still provides most of its protection, even from the sides or back while the in-game advisor kept telling me (before I shut him up :beam:) that the shield does nothing (although this wouldn't be the first time the in-game advisor has lied to me before, not just from this game)
This would also mean that the "ap" or armor piercing attribute means (target armor-1)/2 instead of just target armor/2, which is what export_descr_unit.txt told me in the description (and I think the in-game advisor may also have told me that at some point)
I think this is really important for deciding which soldiers to recruit, especially for deciding whether to use 1200 cheap ones or 600 really expensive ones (in battles I have seen my expensive knights and heavy inf get creamed by hordes of peasants, even without ap weapons. So why even bother to build expensive stables when mounted sergeants would do?
Also important for mods - this would make the Realistic Archery mod unbalanced, since he multiplied the attack points of all arrows by 3, so longbows do 18 attack instead of 6, peasant archers do 15 instead of 5, etc. but he really should've added 6 to everything (since we're actually dealing with a logarithmic scale) to triple every arrow's damage so all archers get equally boosted.
-Gene