Civ IV combat odds are rigged.

While I generally agree, the glitchiness of the interface tends to say differently. It is really horrific.

(I love the spelling recommendation for glitchiness.)

@rah, what do you mean? It's true that the interface is glitchy in general, but on the topic of combat odd... so you know of any cases where the odds displayed are wrong by more than, say, 0.5%? I'm under the impress that the calculation of those odds is actually quite sound.

I wasn't complaining about the odds. I was agreeing that except for the interface, the game played well. My biggest interface issues are when trying to select units from a stack. When the keys don't always do what they're supposed to so certain things take longer when you're trying to save time, or you just screw up. It's frustrating in MP when time is more of an issue.

Monkey. Typewriter. Infinite time. Shakespeare. Lay person.

Sinyail said:

I still think that the combat might be rigged, though. Firaxis may have put information into the games code that is not possible for modders to see or have access to. /: my 2 cnts

Click to expand...

combat is rigged. I know this from reliable source. In case when your win chance is 99%+, the AI secretly makes a extra roll that has 20% chance of insta-win combat.

BUT the secret code CAN be disabled! here is how: when you start the game you have to press following on your keyboard: Back-Back-UP-Down-Down-Enter-6789, and it will be disabled.
Additionally, you will get to play with Kung Lao as your civ leader.

ChaosSlayer said:

combat is rigged. I know this from reliable source. In case when your win chance is 99%+, the AI secretly makes a extra roll that has 20% chance of insta-win combat.

BUT the secret code CAN be disabled! here is how: when you start the game you have to press following on your keyboard: Back-Back-UP-Down-Down-Enter-6789, and it will be disabled.
Additionally, you will get to play with Kung Lao as your civ leader.

Click to expand...

Ah, I miss those days.

Yeah, it's pretty rigged. I was playing a game on chieftain, and a single barbarian archer just strolled into my country and killed one of my cities that was defended by a longbowman. I figured "well, I guess I was just unlucky", and then I sent a horse archer to kill the barbarian. I lost six horse archers to that one barbarian archer, and all six battles had >99% chance of victory. Then that same archer walked on over, defeated another longbowman-defended city, and razed it. I finally killed the %@#$ with another two horse archers.

Seriously, WTF

Its quite plausible if you are playing Vietnam

The only truly bad thing about civ IV odds is that the impact of any one outcome is too great early in the game. Late game, the impact of any one outcome is small enough that variations in outcomes that different even by several standard devations from what one might expect still wouldn't have a material impact on the outcome of a war.

But early on it means you lose a well-defended city or attain a game-changing advantage. This is part of the reason barbarians are not a well-designed mechanic, though not as bad as huts, events, and spawn locations.

ChaosSlayer said:

combat is rigged. I know this from reliable source. In case when your win chance is 99%+, the AI secretly makes a extra roll that has 20% chance of insta-win combat.

BUT the secret code CAN be disabled! here is how: when you start the game you have to press following on your keyboard: Back-Back-UP-Down-Down-Enter-6789, and it will be disabled.
Additionally, you will get to play with Kung Lao as your civ leader.

Click to expand...

Contra...old dudes know it. "Up Up Down Down Left Right Left Right A B B A Select Start. Infinite lives!!!

On topic, I just had the most ridiculous RNG string.
I sent 10 HA against a hill city, 20% cultural defense with 1 C1 Spear, 1 Axe and 2 Garrison 1 Archers. I lost one HA, and had 3 left with 2 moves to attack an archer on a forested hill on the other side of the city, across a river. I lost one HA!
I had mostly awful odds and won!

I get so peeved when I lose five consecutive battles at 70%. This string was unbelievably good!

SpyPenguin said:

Yeah, it's pretty rigged. I was playing a game on chieftain, and a single barbarian archer just strolled into my country and killed one of my cities that was defended by a longbowman. I figured "well, I guess I was just unlucky", and then I sent a horse archer to kill the barbarian. I lost six horse archers to that one barbarian archer, and all six battles had >99% chance of victory. Then that same archer walked on over, defeated another longbowman-defended city, and razed it. I finally killed the %@#$ with another two horse archers.

Seriously, WTF

Click to expand...

Lol! As someone earlier on mentioned in this thread that the results are randomized, it doesn't care how important the battle is, this can cause human player to rage quit in certain situations, when they are expecting to win. I've seen it myself but my catapults have also survived many time with odds like 30%, so in short it can go both ways.
If you are too irritated by this issue you may try ciV because it has FAR superior combat system.

Sent from my HTC One V using Tapatalk 2

Deterministic combat IMO *would* be better like Babri said above. Civ's replayability for me comes from different maps and AI arrangements each time. I like predictability in a lot of other stuff, like my % increase from libraries etc. Combats with predictable results would make each stack battle a fun little subgame for me, not the slightly vague "bring rather too many assorted units and hope for the best" it currently is.

If barbarians never attacked player units that might be better. The barb sub-game would be much the same but we would not be subject to occasional game-losing bad combat rolls in 1 on 1 fights against a low-odds spawning archer in the 1 unbusted tile near our culture.

AI DOW mechanics for me are one area I don't know whether I would like less randomness or not. It's obviously such an important area. Perhaps if there were periods of turns where DOWs were not possible, and other periods where... I don't know, they were subject to some known rules or something, ... that would be an improvement gameplay wise, if not realism wise! I mean the AI is already coded not to DOW us for a while at the start, why not formalize that and take it a bit further maybe :

I do realize different people like different amounts of wildness in their game experience though, as shown by the number of random events lovers Next version perhaps they could make deterministic combat an option too.

Oh yes, and once again let's be clear, the Civ 4 combat odds are not rigged They're just odds - the main question is how much we like our game of civ to be a rollercoaster ride with lady luck or something more .... scientific ...!

the funny thing is, the OP is basically right.

the game does not display the combat odds, but an estimation of them. so they are rigged. The code for the combat odds and the estimation is different. In particular the estimation uses the average of firststrikes. The result is that if the game shows you 99.99% fightings odds and you have a couple firststrikes, your true combat odds are probably only about 96%. In unmodded civ its not a big issue but if you play mods like ffh it becomes more of an issue.

Sephi said:

the funny thing is, the OP is basically right.

the game does not display the combat odds, but an estimation of them. so they are rigged. The code for the combat odds and the estimation is different. In particular the estimation uses the average of firststrikes. The result is that if the game shows you 99.99% fightings odds and you have a couple firststrikes, your true combat odds are probably only about 96%. In unmodded civ its not a big issue but if you play mods like ffh it becomes more of an issue.

Click to expand...

Do you have any hard evidence to back up your claim that odds are inaccurate, or is your claim just based on anecdotal evidence and intuition?

Obviously the code for calculating the odds is different to the code for the actual combat, because those are two completely different tasks.

I doubt he has evidence for that, although sample sizes are small enough in most wars where you care about odds at all that unexpected outcomes can really screw you, and the investment to make it nigh-impossible is cost-prohibitive. That's bad game design, because ideally you'd want players to trade between "I can win this guaranteed", "I will usually win this", "I might win this", "I probably won't win this", and "I can't win this" based on plausible degrees of investment.

In civ IV, in the early game it is absolutely impossible to get "guaranteed" while in the mid-late game the difference between guaranteed-win investment and "might win" investment is trivial; that's why everyone will say things like "I know this game is won."...and be correct despite victory being 50-100 or more turns away.

But why the divergence? It doesn't make sense that the outcome of very long games can be decided at nigh-random in the opening turn sets, only to stabilize to the point of every non-broken-event factor being *extremely* minimal on a per-rng basis. Probably the biggest offender is spawn location (god is this awful in civ IV, and barely any better in V), but barbs, "early rush" situations, and huts/events if one leaves them on all follow the same pattern...RNG is potentially GAME BREAKING early and completely irrelevant later...with only the most asinine and ludicrous events (bermuda triangle, forced DoW) being exceptions.

Compare the axe rush to the cannon attack at tech parity. The former is absolute RNG hell, and yet could easily be an optimal choice since it's the only way you can win a given map (very boxed in). The latter is FUNCTIONALLY deterministic combat, in large part due to collateral. This is part of the reason I always found the people who STRONGLY argued against deterministic combat to be using questionable logic. With the way wars after the early game go, collateral initiative, positioning, and very late game nukes ARE the wars...RNG is non-factor.

Show me the last time the RNG screwed someone in a Rifle-cannon war . That's deterministic combat, or more accurately mathematically so close that the difference is negligible. Why, then, can't we have it at the beginning of the game also? This *is* a strategy game after all...and the vast majority of it actually works off of strategy, go figure! But not early on. Nope. Somehow, people hold that sacred too lol.

The one redeeming factor is that I would much rather be a RNG victim early than late in the game. So while the probability is higher at the beginning the impact is lower. Roll a new start and move on, little time and effort is lost.

The only real RNG problem late in the game is whether your nukes will make it through the AI's SDI and whether his will make it through yours. Granted, you can, and should, always just build a lot so that some should get through, but if yours don't and the AI's do, it can be like an axe rush gone bad.

So while the probability is higher at the beginning the impact is lower. Roll a new start and move on, little time and effort is lost.

Click to expand...

That is operating on certain assumptions. Regardless, it is valid to question why a game that is almost entirely deterministic insists on having pattern-breaking outcome-altering RNG elements at all?

The only real RNG problem late in the game is whether your nukes will make it through the AI's SDI and whether his will make it through yours. Granted, you can, and should, always just build a lot so that some should get through, but if yours don't and the AI's do, it can be like an axe rush gone bad.

Click to expand...

I'm not a fan of this either, but you do have other late game options available, whereas in early axe rushes sometimes it's literally your only option (peninsula start with room for 2 cities and can do nothing else).

TheMeInTeam said:

Regardless, it is valid to question why a game that is almost entirely deterministic insists on having pattern-breaking outcome-altering RNG elements at all?

Click to expand...

It is most certainly valid to pose the question.

Clearly some level of unknown increases the suspense and the fun, but the extreme rolls can negate that in a hurry.

Mec AntiKythera said:

It is most certainly valid to pose the question.

Clearly some level of unknown increases the suspense and the fun, but the extreme rolls can negate that in a hurry.

Click to expand...

But why make the unknown almost unique to combat while the vast majority of decisions are otherwise deterministic? It doesn't make sense to single out this one particular mechanic. If combat adds suspense and fun this way, why not tile yields? Why not randomize the yield of every tile every single turn too? You'd *never* know what you'd get for sure, but over time a farm would be worth more food, mines more hammers and so forth.

It's the same concept applied to different systems...but the effect would be the same regardless of where you put your RNG. This opens a rather valid question of why use RNG for outcomes/planning at all? It would make the most sense for unexpected outcomes to be sourced based on the decisions of the opposing players.

karadoc said:

Do you have any hard evidence to back up your claim that odds are inaccurate, or is your claim just based on anecdotal evidence and intuition?

Click to expand...

getCombatOdds uses the average of chanceFirstStrikes, while CvUnit::setCombatUnit uses a Distribution GC.getGameINLINE().getSorenRandNum(chanceFirstStrikes() + 1, "First Strike"). If anything, it's up to the people who blindly believe anything they see ("oh look combat odds, they must be right") to come up with some hard evidence.

karadoc said:

Obviously the code for calculating the odds is different to the code for the actual combat, because those are two completely different tasks.

Click to expand...

I think if you wanted to create combat odds that aren't rigged, you code only one combatodds function and use it both to display odds and calculate combat, instead of coding two different functions.