What if combat wasn't RNG dependent?

TMIT - I can see why this is an important issue for you. In competitive MP games, which spend the majority of the time in pre-medieval eras, skirmishes happen all the time. Choking and skirmishing can constitute the majority of combat, and in these situations the RNG actually impacts the game. For the majority of Civ 4 players, however, the RNG doesn't alter the outcome of games. Yes I know players like to complain about losing some battles at 90% odds, but if you lost the game it's not because of that. When you play a full game, winning conquest in the industrial era or whatnot, and fight hundreds of battles, the RNG won't determine game outcome in over 95% of games, it just won't. So for most of us, whether we're playing with an RNG or not won't alter the game that much. So if game outcomes aren't altered then it does become a matter of preference or flavour. Personally I like the flavour or RNG better than not having one, so if it doesn't alter game outcomes I'd rather have it. Now, if I played lots of competitive MP (I have in the past, btw, but I stopped playing MP because RL got in the way) then I might be more annoyed with the RNG. However, I think Civ 4's combat is horrible as is and there are so many problems with it that the RNG doesn't even top my list. Example - collateral damage overpowers everything else in the game. So if I were to make a mod to improve the combat in Civ 4 I wouldn't think of starting with removing the RNG. I mean think about how stupid combat is in this game. We have unit counters, like horses vs spears, spears vs axes, etc, and yet we can stack defenders. The entire combat system is complete garbage. If you want a good tactical wargame Civ is probably one of the worst out there anyway. It's a good game overall, because of all of the other elements, but actual combat is terrible, imo. I don't know if removing the RNG would improve that, tbh. It might, if it were part of a larger change to combat.
 
Would it not make some promotions a bit useless though? Like barrage and first strike for example?

Barrage would just do more collateral, I'm not sure why it would need to change at all from its current form, except that it sucks consistently even now. First strikes would mean less damage, and retain their ability to be more meaningful when strength differences are already large.

And getting ahead in the tech race for better units would be even more important, while currently you can get by with obsolete units to a larger extent.

Both tech race AND well-promoted units would have far more impact. Especially in the latter, suddenly aggressive and even protective have more teeth.

I'd hate it - It'd just turn combat into a giant number-crunching exercise. Oh look, Pericles is defending that city with 3 CG Longbows. Right, let me get my excel spreadsheet war calculator out...so if I attack with 5 CR2 Maces I'll win and my last Mace will have 4.3 str left. Select units. Attack. Win. Yawn...

You...do realize people already do this, right? And so what if you know your 5 cr II maces will emerge victories with 1 left? Does that guarantee you hold the city next turn? That you can reach the city with 3 longbows before it reinforces?

But what it does do is greatly increase the strategic depth of the game as it forces you to put contingency plans in place in case the unlikely happens.

The reality is that when you're competing against other players either directly or in XOTM/HoF formats, there is no contingency plan. You either go all-out on war or you keep just enough to survive. If you go all out and fail, you're not coming back. You might be able to still beat the AI on weak difficulties, yes, but the competition is effectively over when that rush fails...which makes RNG based rush outcome "attempt scumming" a serious barrier to skill-based HoF competition by the way.

And when the odds go in your favour and you win an unlikely fight, it can be a game changer.
Yes it can be, and someone...possibly someone who outplayed you...will pay the price for it.

You have your opinion, I have mine.

Nope, what you have is the *fact* that you're wrong. The statement I attacked reads:

For the record all war games need a RNG since war is never that predictable.

Despite the fact that there are quite a few war games out there, including a couple I can think of off hand that are *more* popular than civ (starcraft) that do not rely on RNG whatsoever, and then another subset where you rely on it FAR less (HOMM still has it but not gamebreaking for example).

To put it more mathematically: Your fixed combat resolution has a jump discontinuity at the point where both units strengths are equal and the jump height is huge.

Agreed, but I'm wondering how bad this is. By the way, in current civ the jump point at near-even str is also quite large ;).

Are you sure you can mod this into the game?

My friend dug up the code and seems to believe it can be worked with. Knowing him, the code will probably wind up cleaner, not that such is much of a challenge.

I was under the impression that the OP is talking about something like that?

Civ V uses the RNG in battles also, and the result can be disastrous placement of units out of position or failures to kill something. However, my beef about civ V is a separate issue, and largely due to the fact that it's still in beta.

There should be some kind of middle ground.
Carry some cheap cleanup units?

BtW how would you deal with withdraws, TMIT ?

Questions like this are why I made this thread. It's a tough question. You'd either have to leave them married to RNG (aka chance of getting away in losing attacks, obviously irrelevant in winning attacks) or rework mounted. The latter is the more consistent approach, but it begs the question of "how".

I don't think the jump points are as large a deal as made out to be, especially if you make expected damage fairly high unless the str differential is huge. Yes, a musket could reliably kill a mace, but what if that leaves it in the position that it can't even survive an archer follow-up? You'd still have to be careful vs dated units because both defensive bonuses and using 2 cheaper units (losing 1 to take a 1-1 trade) would both be serious threats to the more advanced attacker, as would collateral.

I don't see a problem with giving fixed damage to nukes or air power. Espionage still has its costs, but it's a bit of a broken system unto itself (so many MP games ban it) that it could probably use a "rework" look too, but for the purposes of our weekend MP games we'd probably just leave it alone for a while.

How RNG dependent is Civ V combat? Not that it should be used as an example, but it seemed less luck driven than Civ 4's.

It's still there, but definitely less. It's also one of the areas where V is a legit improvement over IV (it's not like the AI is good at fighting in either game). It's such a shame that civ V forces players to spend hours per game doing nothing or I'd even advocate it.

@Noto

It's not like I'm proposing we use this system as a S&T standard or something. I'm using GD as a sounding board to come up with reasons for and against doing this for our weekend co-op MP games where almost always someone gets screwed early game. We're also doing things to reduce griefing like removing sistine chapel that have nothing to do with this thread :).

But issues that I wouldn't think of offhand do come up, such as how do deal with withdrawals, flanking damage, and whether collateral damage itself needs a serious look.
 
Possibly, all ranged units should have 2 movement.
This way an archer can attack and then retreat.
This would simulate the archers ability to attack from a distance.

Axeman may be able to kill archers, but only after they have moved past the kill zone of the arrows.

This is one of my main gripes about Civ combat.
 
The more that I think about it, the more I like the idea. I mean, when I'm playing an RPG, for example, I hate the attacks that depend heavily on luck.

I think flanking and withdrawals aren't that great anyway, so no loss there. Collateral isn't as important when other units do set amounts of damage. All I would change there is to let horsies get their attack bonus to work on defense (eg HA's get +50% versus catapults). Either that, or keep siege units from doing anything but collateral when attacking.

Btw, I think that there's still room for compromise even if you don't want to give units twice the hitpoints like I'd envisioned earlier. You could have both a set amount of damage component and a random portion (promo's versus base strengths?). Meh, I guess that would still increase the amount of calculations that the computer has to do.

What if you could combine stacks so that they live or die together? In a real battle, I can imagine armies launching siege and arrows before sending in their cavalry and so on. So, to simulate that, you get to use the first strikes of the archers, then horse archers try to kill all of the siege, then the collateral (which would have to be implemented differently) from your siege or whatever. The point is that with a shared HP pool, you can lower the impact of RNG (both to screw you over and from having to wait while the computer ends one animation to start the next).
 
TMIT: What if combat outcomes between two relative strength ratios were *fixed*? As in, an axe attacking an archer (neither with promotions) on a flatland *always* had the axe win, but also *always* with the same amount of damage taken?

I am about 2/3rds with you in this.
In this example, I agree the axeman should *always* win.
I too agree, it should also *always* take the same amount of base damage. I would simply add +-10% to give it some amout of variability.

For example, lets say that the axe (5) vs an archer (3) will *always* yield a victory for the axe and he will be left as a 3 strength Axeman base. (I am using the ratio 3/5th here or 60%x the 5 base)
Then, modify this by +-10% or +-.3, so our axemen could come out as a 2.7-3.3 strength from this battle.
With a stack of axemen attacking an equal stack of archers, in a real battle, wouldn't have everyone have the exact battle wounds. This variation permitted, keeps some amount of unknown element, but with a stronger element of known common sense victories.
If, your friend can pull this off, it will definitely be an improvement.


Possibly, all ranged units should have 2 movement.
This way an archer can attack and then retreat.
This would simulate the archers ability to attack from a distance.

Axeman may be able to kill archers, but only after they have moved past the kill zone of the arrows.

This is one of my main gripes about Civ combat.

This is best simulated by a guerilla3 archer which has a 50% chance to withdrawl. They newb archers that are less experienced, just can't withdrawl in time.
 
My friend suggested that units don't necessarily die after a combat, however I don't like that because it would become nigh-impossible to kill anything in large stack battles on offense...you could literally out-heal the damage even if you were redlined with a good supermedic in a city.

What if you could combine stacks so that they live or die together?

Reminds me of civ II and III, except as an option. I do worry about the amount of programming legwork that would take though. Now we're not just talking about changing how the RNG is called, but literally introducing a new mechanic which is much harder/less practical, though it wouldn't be a bad thought experiment.
 
Honestly, try it. It would definitely change the early game a lot. It would really change the feel of the game, though. It would be like going from Risk to Diplomacy. Both are great games.
 
In my example, all of the archers would still die to all of the axeman. I would just adjust the damage done to the axeman, so that not all of them have the same exact wounds. A small amount of variety there isn't a bad thing.


I'm pretty sure we will try it. It's not like a change is unwelcome after doing this for years :).

This reminds me of the modders back on the MOO2 sites waiting for a MOO4.
Everyone wants to see a game that works well. The company seems to get part of it right, but just drops the ball at some point.
 
Yeah, this is definitely worth modding in and testing it out. But I can make some predictions on how this will play out.

I do think that the strategic ability to commit units to 40-60% battles is extremely important to the decision making of the game. I need to weigh the outcomes, some battles are unimportant and I happily take mid-range odds, some are crucial and I will only take high odds, and hopefully with backup.

For instance, if I come across a settler/warrior pair with my own warrior, you bet your ass I will take odds on them. I'd even do it if they were on a hill. It costs me little if I lose, and it costs them alot if I win.

On the other hand, if someone starts choking me with their starting warrior, and it finds itself next to my single C1 warrior, I am not taking the ~60% odds on that fight. Because if I win that fight, great, I can keep improving tiles! If I lose that fight, I am dead, and hence I'm better off sticking my C1 warrior inside my capital and forcing them into <10% odds. If my opponent wants to attack into the city, they're welcome to, otherwise they're denying me worker turns. I'd only take those 60+% odds when I have the opportunity to double-team that warrior with my warriors.

On the other hand, if this completely deterministic system was in place, I will absolutely attack if given a 100% chance of victory. The only downside will be that my warrior will be rooted inside my city for 5 turns healing.

My friend suggested that units don't necessarily die after a combat, however I don't like that because it would become nigh-impossible to kill anything in large stack battles on offense...you could literally out-heal the damage even if you were redlined with a good supermedic in a city.

My personal feeling on healing is that it really shouldn't be free, there should be a real opportunity cost from your entire empire to regain a unit's strength. And this opportunity cost than being out of action for X turns. I know it cost money to heal your units in Advance Wars.
 
I'm not a fan of random combat, but I figure a reason why the designers did it is to force you to take some losses even when you have superior units. So if you fight 4 battles in a row at 75% odds, they want you to spend 1 unit on that. The randomness is an attempt at enforcing that loss. Unfortunately it's not a very good method, because with small sample sizes like that, it's pretty likely that you don't get the expected result and someone gets screwed.

Let's say you make a simple change to the combat formulas so that instead of a 75% chance of winning, each of those attackers would be guaranteed to win and take some deterministic amount of damage. But then you're not spending units any more - unless the enemy has the means to counter attack those wounded units before they heal, your expected loss goes from 1 unit to 0 units. That's a pretty big change to the game's balance, but maybe you're okay with that.

You could compensate by somehow having a hammer cost associated with healing a unit, so that taking damage is a loss of resources in itself. But then you're going beyond simple modding.
 
I'd hate it - It'd just turn combat into a giant number-crunching exercise. Oh look, Pericles is defending that city with 3 CG Longbows. Right, let me get my excel spreadsheet war calculator out...so if I attack with 5 CR2 Maces I'll win and my last Mace will have 4.3 str left. Select units. Attack. Win. Yawn...

You...do realize people already do this, right? And so what if you know your 5 cr II maces will emerge victories with 1 left? Does that guarantee you hold the city next turn? That you can reach the city with 3 longbows before it reinforces?

No, what you get though is a very one-dimensional kiind of warfare. Let's assume in the above example that I've taken his city but the very next turn Pericles comes along with 3 maces of his own, kills my mace, retakes the city and garrisons the maces. I've got 4 additional maces as reinforcement one turn away, but again a quick check of my 'war calculator tells me I can't win. Therefore I'm forced to move these units elsewhere.

By taking away RNG you've taken away the option of attacking anyway. There are times I will attack at 20% odds, because the strategic value of the win outweighs the fact I'll probably lose. You attack enough times at low odds you will win.. I've taken a strategic decision that the prize of victory outweighs the loss of defeat for the given odds.

And when the odds go in your favour and you win an unlikely fight, it can be a game changer.

Yes it can be, and someone...possibly someone who outplayed you...will pay the price for it.

Possibly, but see my original point about having contingency plans. I would argue that a 'superior' player isn't actually superior at all if he has lost a battle without a contingency, he/she is just a gambler who plays the odds and lost. How is that superior play?
 
In general I like this idea, and would be very interested in reading more about its outcome once play-tested. The only argument I can find against it is that there has, historically, been unpredictability in war. With this kind of system (which could work very well), would we see the redcoats lose to Shaka's impis at Isandlhwana, or the 7th wiped out at Little Big Horn, etc? There have been cases where the side that really should have won, has lost. But its only a small argument against.

As regards mounted in general (a point referred to up-thread) I can only feel that it is basically broken. If mounted was divided into melee and missile it would go some way to fixing it, but this is only really an argument in favour of proper ranged combat, and this isn't available in infantry combat either, and most particularly with siege...
 
I just thought of something last night that maybe you haven't considered yet. When units fight, they damage each other by percentages--each unit has 100 hp and they do around 20% per round. The game balances this out by shifting the odds of winning to favor the stronger unit, but if you get rid of the odds, you'll have to have them do set damage to the unit strength, rather than HP. Alternatively, set the HP and (base) strength equal to one another.

So, what did you think about the idea of catapults only being able to collateral? There should probably be some way of killing them, but I kind of like the idea. You'd probably have to tweak the system a bit so that siege bonuses would increase collateral instead, but I think that it would be interesting.
 
Possibly, but see my original point about having contingency plans. I would argue that a 'superior' player isn't actually superior at all if he has lost a battle without a contingency

Statements like this lack a fundamental understanding of the game's reality. You can only invest so much in any one aspect of developing a civ. When you choose something that puts you at a large advantage, only to lose to someone who made an inferior choice (aka not defending himself properly) on sheer chance, the luck**** person is not somehow superior. The guy who outplayed him and got screwed is still a better player.

I'm sick of games overly catering to skill equalization. I'm also sick of people who routinely benefit from skill equalization equating that benefit to skill.

would we see the redcoats lose to Shaka's impis at Isandlhwana, or the 7th wiped out at Little Big Horn, etc

I thought the redcoats were *quite* outnumbered by those impi, though I might be mistaken.

each unit has 100 hp and they do around 20% per round.

HP damage is a function of strength ratios, even currently. A much stronger unit might only need 2-3 rounds to win instead of 5. Even a marginally stronger unit will win using one less round than its opponent, which is why you see break points in odds.
 
It would really affect the way barbarians are fought in the early game. If you always win the battle if you have >50% odds, you could always defend against archers by putting warriors in the forest. So barbarians would be less of a threat, but this could be considered a good thing.

Yep, and you would also always win when AIs make silly moves, big advantage for human players who make less mistakes. And if you program barbs to never attack with losing odds..or AIs..what will happen with their stacks if they cannot get winning odds anywhere? ;)
So i think this would not work, either it'd turn combat into "wait and see" boredom, or human players could use silly AI moves for their advantage big time ~~
 
And if you program barbs to never attack with losing odds

Barbs are a joke mechanic in the first place, especially with how their spawning/attack routines currently work.

So i think this would not work, either it'd turn combat into "wait and see" boredom, or human players could use silly AI moves for their advantage big time ~~

You're not going to win "wait and see" against an empire with a better productivity setup (or bonuses), so you better hope you can bait out the AI somehow...possibly masterfully baiting it ;)
 
"I thought the redcoats were *quite* outnumbered by those impi, though I might be mistaken."

Of course the redcoats were outnumbered, however in numerous circumstances during the scramble for Africa horrendously outnumbered European troops defeated massed technologically inferior opponents - see Ulundi, Rorke's Drift, Uganda, Namibia, Tel-el-Kabir - or in situations in the US "pacification" of native Americans where the same thing happened. It was commonplace, to the extent that when the "natives" actually got a small victory back it was earth-shattering. In short, it was unpredictable. It shouldn't have happened. Hence my reference to the unpredictability of war...
 
Back
Top Bottom