CIVPhilzilla
Reagan Republican
It has some flaws but thats what makes it interesting knowing nothing is for sure.
Who hates Civ3's combat system?
- Thread starter Sark6354201
- Start date
Bluemofia
F=ma
yes, flawed it is, but i can live with it. in wwi it is still possible to beat the enemy without tanks, but it would probably take hundreds of millions to overrun germany, and the axis. in civ III, it is possible for just 1 unit to destroy an entire army of units that is about the same helth as it, just very difficult. the combat should have something like a scizors OFTEN beats paper and paper OFTEN beats rock and rock OFTEN beats scizors. it is annoying on how, not too useful subs are. low attack, low defense, and in real life they are anti battleship. if they use it this way, it makes the possibility of snipers who can only damage troops and not hurt tanks, and RPG people who excell at destroying tanks, but scumb to conventional infantry.
Commander Bello
Say No 2 Net Validations
Personally, I believe that most players (those, who are not satisfied with the combat system) feel that random results just happen to often.
Well, as we have discussed over and over again, randomness is inherent to the use of a RNG. So, you cannot avoid to have randomness - and you shouldn't try to avoid it, as one of the very principles of the game.
Nevertheless, sometimes it can literally drive you crazy.
This stands true for me, as I am pretty sure to have noticed a higher number of "streaky" results in C3C than in PTW, for instance. As I've tried to explain in another thread, this may very well happen, just because C3C might have been compiled with another version of the same compiler.
Furthermore, I think it doesn't help just to declare the current combat system (and it's different factors) to be unchangeable, fixed by divine law or whatever.
The combat system for sure is way off from being perfect.
Now, how could one fix this?
First, there would be the hitpoints. I guess, everybody agrees upon that more hitpoints would level out "extreme" results of the RNG. The result would be more "predictability" of the individual fight between 2 units.
Nevertheless, at least a significant fraction of the posters here complains every time about such an proposal. Most times, they refer to early ages' units, when they explain their point of view.
This brings me to the second point of my argumentation.
I think, it just is a principle weakness of the whole way, in which Civ3 has been set up. When we have a look of the game mechanics, we find small integers everywhere. At first glance, this seems to make things easier to understand. It is very easy, to identify the higher defense value of a spearman with D=2, compared with the warrior at a defense of 1.
Nevertheless, it makes the steps just much to high.
At later ages, since the numbers are much more higher, the problem seems to occur less often. At least I've read very few posters complain about the outcome of battles of the modern ages.
Now, since the editor allows to use much higher values for the combat stats, I regard it as a proper idea to change the combat values.
If all combat values would be multiplied by 10, in a first step we would have the warrior with A10, D10 and the spearman with A10, D20. Of course, this wouldn't change anything, since the relation just stays the same.
But, now one could balance it out (what we cannot do with the current numbers). The spearman could get a defense of let's say, 18, and an attack of 13. The swordman could get A27, D22 and so on.
In combination with some higher hitpoints, this could make the "very random" results appear less often, which I feel is the main problem.
And - this would be a positive side-effect - it could help to bring units of a certain type to a more comparable level. UU's replacing a standard unit could cost the same amount of shields, since they would have been levelled by the combat stats. Currently, they just create a class of their own, as I feel it. Furthermore, this system could be enhanced more easily by the use of the "hidden attack". Currently, it makes almost no sense to give the pikeman a hidden attack limited to mounted units, since his attack of 1 makes him loose most times, anyway, even against a torn down unit.
But, let's be honest.. A unit with a rather good defense would be expected to have at least a mediocre attack value as well. This we cannot do with the current numbers...but we easily could with higher numbers.
Of course, all this would require a complete new balancing effort. This for sure is nothing what you could do just at the afternoon. On the other hand, you could quite easily do it with Excel or any other spreadsheet.
I know, that this proposal will not be loved very much by some ..... let's call them "conservative" members of the community. It would cause massive reconsideration of internalized "procedures".
Nevertheless, I assume this to be a manageable way to come to more "logical" combat results, without the need to change inherent functions of the current game engine.
Any comments welcome.
Well, as we have discussed over and over again, randomness is inherent to the use of a RNG. So, you cannot avoid to have randomness - and you shouldn't try to avoid it, as one of the very principles of the game.
Nevertheless, sometimes it can literally drive you crazy.
This stands true for me, as I am pretty sure to have noticed a higher number of "streaky" results in C3C than in PTW, for instance. As I've tried to explain in another thread, this may very well happen, just because C3C might have been compiled with another version of the same compiler.
Furthermore, I think it doesn't help just to declare the current combat system (and it's different factors) to be unchangeable, fixed by divine law or whatever.
The combat system for sure is way off from being perfect.
Now, how could one fix this?
First, there would be the hitpoints. I guess, everybody agrees upon that more hitpoints would level out "extreme" results of the RNG. The result would be more "predictability" of the individual fight between 2 units.
Nevertheless, at least a significant fraction of the posters here complains every time about such an proposal. Most times, they refer to early ages' units, when they explain their point of view.
This brings me to the second point of my argumentation.
I think, it just is a principle weakness of the whole way, in which Civ3 has been set up. When we have a look of the game mechanics, we find small integers everywhere. At first glance, this seems to make things easier to understand. It is very easy, to identify the higher defense value of a spearman with D=2, compared with the warrior at a defense of 1.
Nevertheless, it makes the steps just much to high.
At later ages, since the numbers are much more higher, the problem seems to occur less often. At least I've read very few posters complain about the outcome of battles of the modern ages.
Now, since the editor allows to use much higher values for the combat stats, I regard it as a proper idea to change the combat values.
If all combat values would be multiplied by 10, in a first step we would have the warrior with A10, D10 and the spearman with A10, D20. Of course, this wouldn't change anything, since the relation just stays the same.
But, now one could balance it out (what we cannot do with the current numbers). The spearman could get a defense of let's say, 18, and an attack of 13. The swordman could get A27, D22 and so on.
In combination with some higher hitpoints, this could make the "very random" results appear less often, which I feel is the main problem.
And - this would be a positive side-effect - it could help to bring units of a certain type to a more comparable level. UU's replacing a standard unit could cost the same amount of shields, since they would have been levelled by the combat stats. Currently, they just create a class of their own, as I feel it. Furthermore, this system could be enhanced more easily by the use of the "hidden attack". Currently, it makes almost no sense to give the pikeman a hidden attack limited to mounted units, since his attack of 1 makes him loose most times, anyway, even against a torn down unit.
But, let's be honest.. A unit with a rather good defense would be expected to have at least a mediocre attack value as well. This we cannot do with the current numbers...but we easily could with higher numbers.
Of course, all this would require a complete new balancing effort. This for sure is nothing what you could do just at the afternoon. On the other hand, you could quite easily do it with Excel or any other spreadsheet.
I know, that this proposal will not be loved very much by some ..... let's call them "conservative" members of the community. It would cause massive reconsideration of internalized "procedures".
Nevertheless, I assume this to be a manageable way to come to more "logical" combat results, without the need to change inherent functions of the current game engine.
Any comments welcome.
Gingerbread Man
Dark Magus
So you are suggesting bigger numbers? Sure, maybe for Civ4, but Civ3 has been completed and released for 2-3 years now. I reckon it is a good idea, but it certainly wont stop random results from occuring. Why, random results occur in 100% of battles!
I reckon the only way that would work universally, and keep everyone happy, is some kind of 'This is good against this' system. The random system would remain, just to help us remember that battles can go wrong, and that the cavalry can get slaughtered by a bunch of impi. You definitely wouldn't have one of these flags set for units of the same era (e.g. spear has advantage over horseman). But for gaps longer than archers attacking Infantry (4 attacking 10) or tanks attacking muskets (12 vs 4) you could give the better unit a 50% bonus. I hope Civ4 may have that kind of thing. But until then we will just have to remember to plan for the worst in battles.
I reckon the only way that would work universally, and keep everyone happy, is some kind of 'This is good against this' system. The random system would remain, just to help us remember that battles can go wrong, and that the cavalry can get slaughtered by a bunch of impi. You definitely wouldn't have one of these flags set for units of the same era (e.g. spear has advantage over horseman). But for gaps longer than archers attacking Infantry (4 attacking 10) or tanks attacking muskets (12 vs 4) you could give the better unit a 50% bonus. I hope Civ4 may have that kind of thing. But until then we will just have to remember to plan for the worst in battles.
Commander Bello
Say No 2 Net Validations
No, I don't suggest bigger numbers.
I suggest bigger numbers with more diversity, more fine-tuning and in combination with stealth attacks vs. certain units for each individual unit.
And all that in combination with more hitpoints to limit the effects of RNG-inherent streaks.
And the argument, that Civ3 is on the market since 2 or 3 years, doesn't count.. or you would have to erase the TOW infantry as well....
I suggest bigger numbers with more diversity, more fine-tuning and in combination with stealth attacks vs. certain units for each individual unit.
And all that in combination with more hitpoints to limit the effects of RNG-inherent streaks.
And the argument, that Civ3 is on the market since 2 or 3 years, doesn't count.. or you would have to erase the TOW infantry as well....
Sark6354201
Warlord
The most telling sign of the weakness of the CivIII combat system and in past Civs as well has been the Phalanx or Hoplite's attack of Historically the Hoplites/Phalanx were an excellent offfensive weapon as well as defensive. Phalanx in fact means Battering Ram, refering to the mass group tactics employed in charges against the enemy that yield such spectacular results.
What increasing the numbers would do is allow more flexibility in this area. Whereas a 2.3 Hoplite in the normal game would be overpowered, if the stats of the game are reworked to incorporate a 4.7 Hoplite, this could still work because that work make the Roman Legion 7.7.
The Hoplite (with no terrain modifiers) would win 36% of the time attacking the Legion. Attack a Spearman (3.5) and you would win 44% of the time.
Under normal rules, the Hoplite would win 25% of the time against the Legion and 33% of the time against the Spearman.
Now if the Hoplite was fortified on grassland with a river that would give it an effective defense of 12. Under normal rules the defense would be 3.4,
An Infantry attacking that 3.4 Hoplite would win 63% of the time. So at the very best, you would lose 4 out of 10 times (This is also where streaks and bad luck come in to make these numbers even worse).
Now if we revised the Infantry to have an attack of 29 the Infantry would win 70% of the time against the D12 Hoplite or more depending on if any hitpoints are added.
Hmm, wait, 7% difference? That really won't matter with CivIII because the Random factor is too high anyways. No matter what you do, you will never figure out a way to come up with any sort of logical results. This is my point, the engine is flawed, there isn't any way you can fix its deficiencies with higher numbers or more hitpoints or stealth attack or anything else.
And yes I KNOW I used CRUDE win % and I could really care less if theres a 5 or 10% difference from the crude to what a combat calculator might say because it doesnt affect the point I am trying to get across at all.
The more and more I think about it, the more I think they should of just stuck with Civ2's combat system, it was by no means perfect, but firepower and hitpoints were both very well integrated with AD to give a excellent and noticable difference between Armor, Firearms, and Ancient units along with firepower for artillery.
What increasing the numbers would do is allow more flexibility in this area. Whereas a 2.3 Hoplite in the normal game would be overpowered, if the stats of the game are reworked to incorporate a 4.7 Hoplite, this could still work because that work make the Roman Legion 7.7.
The Hoplite (with no terrain modifiers) would win 36% of the time attacking the Legion. Attack a Spearman (3.5) and you would win 44% of the time.
Under normal rules, the Hoplite would win 25% of the time against the Legion and 33% of the time against the Spearman.
Now if the Hoplite was fortified on grassland with a river that would give it an effective defense of 12. Under normal rules the defense would be 3.4,
An Infantry attacking that 3.4 Hoplite would win 63% of the time. So at the very best, you would lose 4 out of 10 times (This is also where streaks and bad luck come in to make these numbers even worse).
Now if we revised the Infantry to have an attack of 29 the Infantry would win 70% of the time against the D12 Hoplite or more depending on if any hitpoints are added.
Hmm, wait, 7% difference? That really won't matter with CivIII because the Random factor is too high anyways. No matter what you do, you will never figure out a way to come up with any sort of logical results. This is my point, the engine is flawed, there isn't any way you can fix its deficiencies with higher numbers or more hitpoints or stealth attack or anything else.
And yes I KNOW I used CRUDE win % and I could really care less if theres a 5 or 10% difference from the crude to what a combat calculator might say because it doesnt affect the point I am trying to get across at all.
The more and more I think about it, the more I think they should of just stuck with Civ2's combat system, it was by no means perfect, but firepower and hitpoints were both very well integrated with AD to give a excellent and noticable difference between Armor, Firearms, and Ancient units along with firepower for artillery.
Gingerbread Man
Dark Magus
The TOW is a small addition to the game. It fills in a small gap, is nothing drastic. Totally reworking the game mechanics, essentially creating a completely different game, is not something an expansion pack should do.
May I suggest that if you dont like the unpredictability factor, find a different game. Because a predictable game of civ is a boring one. "Wow, I'm guaranteed to win this war. Well, I know I'll win this game, no need to finish it." The great leaders of the world never knew what was coming next. Neither should we.
nihilistvoid
Chieftain
- Joined
- Feb 15, 2004
- Messages
- 29
Maybe in civ4 they could give us the option of which combat system we want. We could pick one with straight math, or one with the current randomness, or maybe one somewhere inbetween (where random events occur only with closely matched units)
SOMETHING needs to be done though. There's no point in having an elite unite and one with a standard 2 "hitpoints" in Civ3; there's practically no noticeable difference in the combat.
I think the thing Firaxis et al dont see is that a unit can serve it's purpose even if it's destined to die no matter what roll it gets. If we had a better hp system a weak attacking unit could partially damage a stronger defending unit. This is rarely played upon in Civ3... you pretty much just wait for a high roll in which the attacking unit takes out the defending unit without a scratch. And the rolls arent random... it fluctuates in a sin wave, I've noticed, with the trough of the wave being longer than the high rolls. All this RPG talk discloses too much of the current combat system...
SOMETHING needs to be done though. There's no point in having an elite unite and one with a standard 2 "hitpoints" in Civ3; there's practically no noticeable difference in the combat.
I think the thing Firaxis et al dont see is that a unit can serve it's purpose even if it's destined to die no matter what roll it gets. If we had a better hp system a weak attacking unit could partially damage a stronger defending unit. This is rarely played upon in Civ3... you pretty much just wait for a high roll in which the attacking unit takes out the defending unit without a scratch. And the rolls arent random... it fluctuates in a sin wave, I've noticed, with the trough of the wave being longer than the high rolls. All this RPG talk discloses too much of the current combat system...
Gingerbread Man
Dark Magus
It goes in a sine wave, eh? I'll test that out tomorrow, and tell you what I get. You're most probably right.
The Last Conformist
Irresistibly Attractive
Commander Bello does have a point with the higher numbers, and so does Sark6354201 with the low attack values of Pikemen and the like, but that's really a basically different question.
I've made my case for the current combat system being fine sufficiently many times I won't repeat it here. But I've never liked the "defensive infantry" concept. On this point I much prefered the CivII system with roughly equal A and D for gunpowder infantry.
nihilistvoid: Are you saying Elites preform no better than Conscripts? You must be playing a different game from me ...
And I look forward to seeing well-documented tests showing there's anything like a sine wave in combat rolls. I've never seen anything to suggest it ...
I've made my case for the current combat system being fine sufficiently many times I won't repeat it here. But I've never liked the "defensive infantry" concept. On this point I much prefered the CivII system with roughly equal A and D for gunpowder infantry.
nihilistvoid: Are you saying Elites preform no better than Conscripts? You must be playing a different game from me ...
And I look forward to seeing well-documented tests showing there's anything like a sine wave in combat rolls. I've never seen anything to suggest it ...
nihilistvoid
Chieftain
- Joined
- Feb 15, 2004
- Messages
- 29
The elites get caught under the same sine wave scheme as everything else, so when it's on the low end of the wave the elite unit is going to die just the same as the conscript unless it's on the curve out of the trough. This has come from hours of reloading games in order to play out different tactics with a variety of different units all in the same turn. The roll frequency becomes all too obvious in this situation (though I realize most people dont play like this... but for us perfectionists who try to plan each move, it's frustrating as hell).
The Last Conformist
Irresistibly Attractive
Even if there is a sine pattern, Elites would have a higher chance of "catching" the rise.
You're not havin Preserve Random Seed on by any chance?
You're not havin Preserve Random Seed on by any chance?
Commander Bello
Say No 2 Net Validations
Gingerbread Man,
I apologize, if this should sound bold, but I think that you either haven't read my statement carefully or you just want to keep the current system, only because it's the current system.
In my statement I was referring to weaknesses of the combat system. I also pointed out, that "strange/random/whatever" results are inherent to the use of the RNG.
What I tried was to show a way of how to improve that system with the currently available means. That wouldn't cause the need for any change in the inner mechanics of combat, since everything would be done by changing the variables.
Furthermore, as I see it, this change would allow to make use of some new features (the stealth attack) which are currently almost meaningless.
I would kindly ask you to read my article once again.
About your suggestion I should go for another game, since "I don't like the unpredictability"...
With that statement any suggestion to make changes in the game can be made obsolete. It's a killer argument. Not for the meaning or content of a given suggestion, but for the suggestion itself.
And, if I may put your attention upon the fact that never in history some godlike leader of a state/nation knew in advance: "Let me put x shields in building the ToE, or let me put y shields in the search for tech XYZ and I will have effect ABC."
This is very predictable. Since I read from your lines that you exactly don't like this, may I suggest you stop playing Civ3 at once?
Of course, the last suggestion is not meant seriously, but should show you how weak your argument for keeping a currently weak combat system sounds...
nihilistvoid
Chieftain
- Joined
- Feb 15, 2004
- Messages
- 29
Nice
wilbill
That Old Time Religion
You really need to tell us the numerical results from you "study" of randomness. Otherwise it's just another player's observations.
Read up on randomness a little. There's no "sine wave" pattern to the streakiness. There are streaks of results inherent in a string of random numbers. If there are no streaks, the numbers aren't random. The streaks also appear randomly. They don't follow a pattern and aren't predictable. Since Civ 3's combat system first came under scrutiny, dozens of posters have set up experiments to prove that the games RNG isn't really random. So far nobody has shown this to be the case.
Try playing a dice game like craps or Yahtzee. You'll see the same kind of streakiness in the dice rolls as you see in the Civ 3 combat system.
Gingerbread Man
Dark Magus
Yeah, sorry, the harshness on my part was unnecessary.
Really, all I want to say is that whatever happens, I dont want the game to run like maths, e.g. 4 vs 9 means 9 will win. That, to me, is just not how war works. I'm not sure if that is what you want or not. But that is certainly not what I want.
However, when Civ4 comes out, I can only expect that the stats will be more accurate, e.g. Spearman defence = 1.8, Tank attack = 14.2. I think you are suggesting that also. And I would like that.
So, I agree with half of your suggestions, and not the rest. I'll bow out of this argument... rather, debate, and just agree to differ.
nihilistvoid
Chieftain
- Joined
- Feb 15, 2004
- Messages
- 29
Considering I'm neither a Civ3 programmer nor a computer program, I don't have any way to translate my experiences into numbers. All I can say that is time after time after time, the sine wave has always been apparent in my gameplay. Try reloading the same savegame that has multiple battle options in the same turn. You'll notice that it's all precalculated and follows a curve; the trick is to simply find the best combination that loses the least amount of units until you reach the small top of the curve. Rarely does something like this occur in craps or Yahtzee, which are actually random. If you need any more proof than that (ie, your own gaming results given the above constituents), I might recommend a book or two in phenomenology.
akots
Poet
Right. Or a simple book on dice statistics. Or play some dice in spare time. Gambling is fun, do not take it from the game. Besides, players like it. Just educate yourself, people. Or watch Rain Man casino scenes again. Or listen to some country station.
Every gambler knows
The secret of survival...
Commander Bello
Say No 2 Net Validations
The problem with the RNG - as I see it - is that it just *IS NOT* random. After all, any RNG is nothing less than a little computer program.
So, as soon as you know about the algorithm (A) used, the seed used (B) and the number of calculations already made (C), you will be able to calculate the next "random" number by yourself.
Since we don't know about the factors A, B and C, it just *seems* that those numbers would be random.
Since the results of the RNG are modified by the combat stats and the hitpoints of the two units involved in a given battle, it would take a good mathematician to investigate whether the RNG algorithm used by the game engine really is according to the expectations.
So, as soon as you know about the algorithm (A) used, the seed used (B) and the number of calculations already made (C), you will be able to calculate the next "random" number by yourself.
Since we don't know about the factors A, B and C, it just *seems* that those numbers would be random.
Since the results of the RNG are modified by the combat stats and the hitpoints of the two units involved in a given battle, it would take a good mathematician to investigate whether the RNG algorithm used by the game engine really is according to the expectations.
Similar threads
