Longshot Combat

[gen.dragolen]

Dear Firaxis,

Please fix the combat system: it is too random, and ruins the game.

As an example, I was just playing, and ran into another series of garbage results. The last scenario (all units were on plains and grasslands, unfortified) saw 2 vetran knights (4/3/2) attack another vetran knight (4/3/2) that just moved 2 squares, and both attacking knights died, and did only 1 hp to the defending knight, who was promoted to elite for my troubles. I attacked with a 3rd knight, who finally killed the defender, but had only 1 hp left.

According to the polynomial method for resolving the combat, the attacking knight should win the fight 60% of the time. The second knight should win 74% of the time, and the 3 rd knight should win 86% of the time. The chance of success for the 2 knights killing the one knight are about 99%.

The that same turn, I attacked a 2/2/1 Peltast with a Medieval Infantry unit. The peltast went to 1 hp and then did 3 hp to the MI before dying. What's so bad about that ? The next turn the 1 hp MI was attacked by a knight which died leaving a 1 hp MI. According to the same method, the MI should kill the peltast 98% of the time, and the knight should kill the MI 98% of the time.

There is a serious problem with the game when the 1 and 2% wins are happening all of the time. And the problem becomes worse as the dificulty level increases. I am running the 1.21f version of PTW. I have done considerable modding with the unit strengths, but still find too many extreme results, especially in naval combat. And this is after having played several hundred games at all difficulty levels.

I have served my country and am familiar with the abilities of a good leader and high morale so I know what people are capable of in a crisis. And having played wargames since I was old enough to read, I expect more out of a game decendant from Civ I and Civ II, than a fancy version of RISK. I will not be buying the Conquerers expansion with out some changes to the combat resolution model.

D.

 

[Roland Johansen]

The probability of two attacking knights losing to one defending knight on grasland, etc, etc.. is about 6,3% (not even considering the chance of promotion of the defender which increases this probability). And that's not that strange as the attack strength of a knight is only slightly larger than its defence strength. So these things will happen regularly (once every 16 times you try this). It wouldn't be right if two knights almost always defeated one. You were unlucky that you only caused 1 hp of damage though. The probability that your third knight won wasn't that big though. Both knights were at 4hp so you had only a 60,4 chance to win the last battle.

Good luck in your next fights. Use larger numbers of units and the average result will happen (10 knights will almost always win against 5 under the same circumstances).

 

[casual_moose]

i dont think they should change it from random. its fun to watch a bad unit defeat a far superior unit (especially when your the one that beats it!) also i dont remember seeing a peltest unit in civ 3 or ptw.

 

[Xen]

Remember, individual units in civ 3 represent full companies, regiments, phalanxes, Centuries, ect... like how the legionary unit is represented by a centurion, who inturn represents the entire century/cohort, a weak unit defeatin a stronger one is similer to how a good general can over come superior numbers with superior tactics

 

[Bamspeedy]

The Peltast unit is in this month's GOTM.

As usual, the biggest complaints about the combat system are the ones using 'dribbles' of units.

2 knights defeating 1 knight is NOT 99%. You are using the wrong mathematical theory, or something.

 

[Bamspeedy]

According to Sirp's combat advisor:
http://forums.civfanatics.com/showthread.php?s=&threadid=51435

2 veteran knights vs. 1 veteran knight=93% chance of you conquering the position.

MI attacking a peltast (assuming both are veterans), you have a 79% chance of winning, not 98%.

Knight attacking a 1-hp MI defending:
Knight should win 98.4% of the time, so you do have a case here, but this is your only 1 or 2% 'longshot' that you have posted. Unless, the MI was on better terrain than plains, or was on the other side of a river that would improve the MI's odds.

And the problem becomes worse as the dificulty level increases.

Higher difficulty levels will have the AI have more units to offer a more effective counter-attack, but otherwise they don't get any advantage in battle outcomes/better odds. I hope this is what you meant.

 

[Rik Meleet]

What I have experienced is that Elite units with full strength die more often in attack than the same unit-type which is veteran. The defender usually goes to 1 hp (redlined) and than keeps winning until the Elite dies.

Another thing I experienced far too often to be a coincidence is that redlined defenders are remarkable strong compared to yellow-lined or green-lined defenders. They win far more battles than yellow and green.

Anyone else has the same observations?

 

[Xen]

yeah unfortunutley, and for the most a good elevition is this-

A)Try to get a UU as soon as possible, as they seem to generate more GLs'

B)No matter what, use the GL to create an army, make this first army a pure offensive one, although the BEST unit in the game to make an army out of is the Legionary, with heafty attack and defense, useful (by its self too, but VERY useful in an army) all the way until the industrial age

C)Use this army as the first strick force against a city- the reason being is that the strongest defendeer is at the top of the stack, and so the first to die, very rarelly will an army die, but to make sure it dosent, try not to attack when you have 50% hp or less, to have them heal more rapidlly, bring them to a city with the barracks improvement

hope that helps

 

[TerraHero]

but still, 79% of winning and still losing fat ass style, something is mayor wrong. I know ppl beat odds, but not so often, and not so blah....

 

[Jeff_ATARI]

Originally posted by gen.dragolen
Dear Firaxis,

Please fix the combat system: it is too random, and ruins the game.

Dear gen.dragolen,

Civ Conquests may include a solution for you. Sorry to be a tease, but I can't say more about it just yet. Stay tuned.

Best,
Jeff Foley
Atari

 

[ipris]

i like the combat system now. i hope the change is a selectable option... an either/or kinda thing.
That said for those that dont like it, i'm glad to hear a possible alternative is being implemented

 

[alpha wolf 64]

Jeff,
That was evil, just plain EVIL!!!....LOL

I've been hoping that more experienced units would also get the retreat option. An elite spearman would be much more likely to know when to disengage than either a veteren or regular spearman.

 

[Portuguese]

Originally posted by Jeff_ATARI
Dear gen.dragolen,

Civ Conquests may include a solution for you. Sorry to be a tease, but I can't say more about it just yet. Stay tuned.

Best,
Jeff Foley
Atari

Good Job!

 

[Speaker]

Originally posted by Xen
B)No matter what, use the GL to create an army

I would much rather build my FP in a corrupt area and be able to produce more troops than having one army....

 

[warpstorm]

But producing one army allows you to build the small wonders that allows you to build armies whenever you want and have a greater chance of more GLs. When I'm warmongering a few armies as 'nutcrackers' are great. I prefer using units with a move of 2 or better so that you get the free 'blitz' attack and the chance to retreat.

 

[Speaker]

True you can build the small wonder that lets you build armies, but at 400 shields a pop, they're not too feasibile until pretty late. I'd personally rather have 6 knights at 420 shields than an army at 400.

 

[grasu]

Originally posted by gen.dragolen
Dear Firaxis,

Please fix the combat system: it is too random, and ruins the game.

As an example, I was just playing, and ran into another series of garbage results. The last scenario (all units were on plains and grasslands, unfortified) saw 2 vetran knights (4/3/2) attack another vetran knight (4/3/2) that just moved 2 squares, and both attacking knights died, and did only 1 hp to the defending knight, who was promoted to elite for my troubles. I attacked with a 3rd knight, who finally killed the defender, but had only 1 hp left.

According to the polynomial method for resolving the combat, the attacking knight should win the fight 60% of the time. The second knight should win 74% of the time, and the 3 rd knight should win 86% of the time. The chance of success for the 2 knights killing the one knight are about 99%.

The that same turn, I attacked a 2/2/1 Peltast with a Medieval Infantry unit. The peltast went to 1 hp and then did 3 hp to the MI before dying. What's so bad about that ? The next turn the 1 hp MI was attacked by a knight which died leaving a 1 hp MI. According to the same method, the MI should kill the peltast 98% of the time, and the knight should kill the MI 98% of the time.

There is a serious problem with the game when the 1 and 2% wins are happening all of the time. And the problem becomes worse as the dificulty level increases. I am running the 1.21f version of PTW. I have done considerable modding with the unit strengths, but still find too many extreme results, especially in naval combat. And this is after having played several hundred games at all difficulty levels.

I have served my country and am familiar with the abilities of a good leader and high morale so I know what people are capable of in a crisis. And having played wargames since I was old enough to read, I expect more out of a game decendant from Civ I and Civ II, than a fancy version of RISK. I will not be buying the Conquerers expansion with out some changes to the combat resolution model.

D.

Oh my God! You read my mind, I hate the combat system, today I was playing a Regent game and (again) I got attaked (I had a Medieval Infatry on a hill) by a regular medieval infatry and the totaly kicked my ass, God I HATE that. After a few turns in that war I attaked their city (pop 7) with 4 Veteran Knights, Catapult and Pikeman, and they kicked my ass with 2 regular spear man, then they attaked one of my cities (3 Regular Pikeman in it, 8 pop) with ONE regular Longbow man and they lost 1 life, and conquered my city in 3 turns... they really have to fix this (but appernetly they will (I hope).

 

[WindSpirit]

I completely agree with gen.dragolen. The combat in Civ 3 is way to random and this is mostly why I still prefer to play Civ 2 instead of Civ 3 (gasp!).

The event that turned me off to Civ 3 is when I spent about 7 knights to take out an A.I. controled knight on a hill only to have a single A.I. knight on the next turn take out one of my fully healthy knights on that same hill. This excesive randomness completely takes out the role of geographic strategy in combat.

So hopefully Conquests will get me back into the Civ games.

 

[Rhye]

The solution is adding 2 hp to units in the editor (5/6/7 instead of 3/4/5). Random effects and veteran/elite influence will be reduced

 

[alpha wolf 64]

I went 4/5/7 and some units with heavy armor get an additional HP, and some very lightly equipped ones get a -1 hp. Altho upsets are still possible, its much more tolerable rate now. I still think the RNG ISN'T random enough. Too often hits come in bunches. Like the rifleman that defended against 3 tanks without losing a HP, but the 4th tank killed the rifleman losing only one HP.