Combat system - statistical stuff from >20 games

[DarkwingGT]

Ok, I was originally going to post something that might be inflammatory. I've decided against that. So I'll sum up and try to avoid a flame war.

"But then, I don't mind the occasional unbelievable result - if I knew ahead of time how everything ended, I wouldn't waste my time playing."

These type of quotes are a little confusing to me. I for one want predictable results, if everytime I attack with a Modern Armor it's up to the whims of fate if I beat a Pikeman, then I can no longer count on overwhelming military might to win battles. If that was the case, then I should never bother to build anything other than Warriors as fate would dictate they win as often as the Modern Armors.

Oh well, it just seems to me that the reason we as a race develop technology is to remove uncertainty. Of course no technology will ever completely remove uncertainty, but that's completely obvious.

 

[Zachriel]

Originally posted by DarkwingGT
Of course no technology will ever completely remove uncertainty, but that's completely obvious.

Right on! History is ripe with the folly of man leading to a unexpected turn of events.

(Your other point as to why we bother to upgrade is obvious. The vast majority of the time superior units win. If anyone thinks you can stop MY tank advance with a bunch of spearmen, you might just want to rethink your strategy.)

 

[eyrei]

I am absolutely amazed that this thread is still going strong, and has not caught fire. Anway, I like the occasional 'bizarre' combat result, as it forces one to plan more carefully for the worst. It may not necessarily be realistic, but it does make for interesting gameplay. All of this has been said before, so I will bow out, still in awe of the longevity of this thread.

 

[DarkwingGT]

I think it's a testimony that we have rational adults posting to this thread that no flame war has broken out.

I guess my point is that occasional flukes are ok, in fact it does make me raise an eyebrow every now and then, but I seem to have more than my fair share of occasional flukes. I might dismiss it as complete bad luck except for the number of people who've posted that they have the same problem.

As a side note, occasional flukes indeed do not add much spice for the dedicated player. For the dedicated player would take the precautions necessary to remove the aforementioned spike. Having said that, that's one of the reason like I said before that I don't understand the "I wouldn't play if everything was predictable" comments. They probably have contingencies for said flukes, thereby turning there strategies into <drumroll please> a predictable outcome.

 

[eyrei]

Originally posted by DarkwingGT

As a side note, occasional flukes indeed do not add much spice for the dedicated player. For the dedicated player would take the precautions necessary to remove the aforementioned spike. Having said that, that's one of the reason like I said before that I don't understand the "I wouldn't play if everything was predictable" comments. They probably have contingencies for said flukes, thereby turning there strategies into <drumroll please> a predictable outcome.

I agree. But these players, myself included, would never have developed these strategies if everything always went as planned. Developing strategies is half the fun of the game. I actually wouldn't mind if the combats were slightly (and I do mean slightly) skewed in favor of the AI. It takes an aweful lot to make up for the fact that the AI cannot learn.

 

[etj4Eagle]

Is this a viable solution? It appears Firaxis is in denial over this, so Civ-ers will have to develop a patch or a solution themselves. Where do we find these files?[/QOUTE]

Load up the editor and go to the "Combat Experience" tab and for each experience level change the number of hit points. Though make sure to change them all, you don't want you base 20 hit point veteran upon becoming a elite to drop down to 5 hit points, oops

Originally posted by eyrei


I agree. But these players, myself included, would never have developed these strategies if everything always went as planned. Developing strategies is half the fun of the game. I actually wouldn't mind if the combats were slightly (and I do mean slightly) skewed in favor of the AI. It takes an aweful lot to make up for the fact that the AI cannot learn.

Agreed, preparing for your military plans going arie is half the fun of the game. I look at combat here just the same as I do in Axis and Allies. There my bombers have an attack value of 4 and the defending infantry a value of 2. But I know when I roll those dice that just because the numbers are in my favor, that is not what is going to happen. There was a wonderful time when a squardron of heavy bombers rolled only 5's and 6's, 12 dies no hits. Oh that hurt, had to call off the attack.

Well time to head back to securing my border in case the Persians decide they want to expand my way.

 

[Marzipan]

Maybe this has come up before, but whether there are statistical flukes in the game or not it might be better to have a certain point where a unit will just overrun others, without the usual back-and-forth "oof oof, oof oof" or "boom boom, boom boom".
For instance if you attack a town with modern armor, and the town's best defender is a spearman, perhaps - regardless of the number of spearman - they should just be considered overrun, be wiped from the game, and your tank moves right in. I dunno, I'm asking

 

[Zachriel]

Originally posted by Marzipan
For instance if you attack a town with modern armor, and the town's best defender is a spearman, perhaps - regardless of the number of spearman - they should just be considered overrun, be wiped from the game, and your tank moves right in. I dunno, I'm asking

I returned, and saw under the sun, that the race is not to the swift, nor the battle to the strong, neither yet bread to the wise, nor yet riches to men of understanding, nor yet favour to men of skill; but time and chance happeneth to them all.

That's the biblical description of a randomizer.

 

[Lt. 'Killer' M.]

hey eyrei! Isn`t that a real bore when you set HP so high? i can barely wait through a fight between an army with 45 HP against a strong defnder, but giving veterans 20HP????? You suggest multiplying by 5 - correct?? That would make my current army..... 4 * (5*5) = 100HP - wow! I`m perfectly happy with trippling (elite=15HP) and bring along that extra 2 to 3 units I need for the flukes....

As I said I´m currently on the roll and it really decides the game whether it`s 40:60 or 60:40 early on! Yesterday I just stormed through the French - now I`m the major power on my continent with lots of elite horsemen - in 170 BC on Monarch, huge & 16 Persia will die next, then germany then I`ll control about 40% of earths surface
What it would be like if the French towns hadn`t each fallen within one turn and the French had had a chance to snap up my retreating Horsemen I don`t want to imagine. That would have forced me to bring slow defensive units along - not good!

I think when you tripple HP it`s ok, but I`d like to see that in an official patch - otherwise I can`t take part in GOTM and we will have a horrible time arguing about which modfile to use in MP!

 

[Hurricane]

Originally posted by Killer
I did this in modern times by blocking iron and rubber and all that and disabling Riflemen. Thus i was able to chek for invisible City walls. They are pretty common; actually almost every time a town with walls grows to size 7 and then goes below 6 the walls dissapear from view in my games. Thus the last outcome was probably heavily influenced by this!!!

Did you all notice this? Anyone else seen this bug? This could perhaps explain why some people see cities that just won´t fall. Maybe they had an invisible city wall?

 

[Zachriel]

Originally posted by Hurricane
Did you all notice this? Anyone else seen this bug? This could perhaps explain why some people see cities that just won´t fall. Maybe they had an invisible city wall?

Bombard, bombard, bombard.

 

[DarkwingGT]

Ok, I think the point is getting lost here.

Let me clarify for everyone: No one has issues with random flukes. Nobody, not a one. So please, no more "But randomness is real" or "I don't want to always win my fights". That's not the issue. Never has been.

The issue is that some of us, not everyone, seem to have massive strings of random flukes. Just because it doesn't happen to you you cannot assume it doesn't happen to anyone. Vice versa is also true, I believe that the people who have no complaints have all fair battles.

I just think the thread needed a little redirection as we have two sides not necessarily arguing about the same thing.

EDIT: What about before you get Catapults?

 

[Sodak]

double post...

 

[Sodak]

Before catapults, your only method is to have your bad guys butt heads with the enemy. That's one reason mathematics is a key tech for conquest plans.

Killer, CivII had HP multiplied by 10, and the combat system worked very well. Ah, how I miss firepower.

Back to the thread topic. While I don't have problems with the combat system, I believe people when they say they have had long strings of bad luck. I think the only real response to this is to try altering strategy. Maybe more bombardment will help you immensely. (Sometimes it doesn't help. I once bombarded a city 50 times, resulting in a whopping 4 hits. But even one weakened defender could tip the scales in your favor.) Plan on using twice as many bad guys per attack. Never use one type of unit exclusively. Things you may have already tried, in other words.

I think one problem is the small number of hit points. More hit points mean less variability in the outcomes. If that can be set higher for all units (or just for modern units), you might have better luck.

At base, Civ3 is Civ3. That some have good luck and others bad is not going to be found in the program code. The number generator may be flawed - maybe because of some incompatible software installed on your pc. For example, my workplace software bails out if our e-mail has been opened before it. Very annoying. I wish I had a solution for you.

 

[Salvor]

Small sidetrack: for the bombard theory, it's unlikely that anyone will have construction and build lots of walls in cities that later grow to size 7 before you get mathematics and build catapults. Not impossible, but unlikely.

Back on track: the code might not reveal differences in "luck", but it might show the true source of the "luck". If it uses something from the OS as the core of the seed, that might have something to do with it. Pieces of the system time string is not an uncommon seed generator.

So the question I would ask next is, what's different about the people experiencing the problem than it is for those who don't? Is it a different version of Windows (95/98 vs. NT-based)? Could it be the system hardware (Intel vs. AMD)? Or maybe something else entirely?

Any ideas?

BTW, I personally haven't experienced severe "luck strings", and am running WinXPHome on an P4/Intel845 MB. And I didn't play before the 1.16f patch came out, so I installed it right after installing the game (before actually playing any games).

 

[Loopy]

Originally posted by Salvor
So the question I would ask next is, what's different about the people experiencing the problem than it is for those who don't?

Those people either remember those unusual outcomes more often or their sense of odds is off kilter. Honestly, I think the software checking for either of those things is as likely as the random code generator being os or hardware dependent. It's much easier pulling an off the shelf random number generator than to coding one yourself.

And to DarkwingGT and your confusing quote, I think that quote's taking it to the other end of the spectrum. Some people would consider a swordsman beating a warrior every time to be a "predictable" result. After all, the swordsman has three times stronger, right? The low number of HP and "hidden" odds adds a wrinkle into the "predictability" of combat. If Civ3 had a pop-up window showing the actual percent chance of winning a matchup, I think all of these "statistics" would be a non-issue.

 

[LaRo]

This is not a pop-up but it may help you. I open it before each civ3 session.
Sometime, I changed my mind on a attack that I think would be fair.

 

[LaRo]

Sorry, this is the website that help me: http://www.civfanatics.com/civ3combatcalc.html

 

[etj4Eagle]

Originally posted by Loopy

If Civ3 had a pop-up window showing the actual percent chance of winning a matchup, I think all of these "statistics" would be a non-issue.

That was a nice item in SMAC, in which your comander would yell at you if you tried to have one of your units engage in combat with unfavorable odds. Of course this was more important in SMAC, as you had the "rush" effect on fighting, where if you had less than a whole move point left for your attack you would suffer a sever penalty.

 

[DarkwingGT]

Confusing quote? Whom did I quote? Not sure what you mean on that one.

I hate to say it but you missed my point. My point is not for completely predictable results. My point is for results to more closely follow the properly predicted outcomes. Take Fantasy General for instance (great game IMHO), when you would decide to attack, they'd show you a Kill and Wound total for each side, showing you what would happen if all the numbers went even for both sides, i.e. if 50% out of 100% were rolled everytime in that particular fight. And when you fought, it generally went as the outcome was predicted. Not always of course, but generally. That's what I'm asking for, my outcomes generally do not go the way they are supposed, based on the combat calculator and what not.