[BTS] Civ IV combat odds are rigged.

[Sinyail]

Civ IV BTS I have been playing for awhile now. I used to just rub off those 98%, 99%, 99.6%, 99.4%, 99.8% odds that were in my favor when I lost, but now I can't stand it anymore. This is total bull crap. You don't understand. This happens every single game that I play. No joke. There seems to be a sequence to it though. When I slaughter say 5 of the ai's units and then go for a sixth, no matter how high the ods are, I always lose. I specifically remember losing at 99.8% odds because I was so mad I almost chunked my keyboard. lol It happens too consistently for it to just be "Oh well you didn't have 100% odds" or "This is just player frustration." No. I'm talking like every 5 battle I will lost a 99.9% battle. Why doesn't Firaxis just add that little 100% battle so I can lose that too? Oh, wait, they did. I would like to be corrected if I'm wrong by some hardcore proof that the civ iv combat is not rigged. Just saying that it's not rigged and that whoever says it is rigged is a conspiracy theorist doesn't do it. Well thx

[ParadigmShifter]

Yay, another teeny trauma thread!

[Sinyail]

You don't have anything useful or relevant to add? Piss off.

[ParadigmShifter]

I have got a degree in maths though with 2 years of probability & statistics in there. Stuff happens, it wouldn't be random if it didn't.

[Sinyail]

This is exactly what I said not to do in my post. "It's not rigged because I say it's not rigged."

[ParadigmShifter]

It uses a published random number function (GetSorenRand I think). It's not rigged. All of the source code is available.

[Sinyail]

This simply does not exclude the possibility that Civ IV combat is rigged.

[Ghpstage]

I would like to be corrected if I'm wrong by some hardcore proof that the civ iv combat is not rigged.

This isn't hard to do yourself.

1. Start a custom game and make sure you turn on the option "new random seed on reload".
2. Open worldbuilder and use it to create 2 stacks, one for yourself, and one for a non barb AI. For simplicity its pretty much necessary to have each stack made up of one unit i.e. all axes vs all swords. Good stack sizes would be 50 or 100 as the numbers are easy to work with, are big enough to get a good sample size fast and won't slow your comp down too much.
3. With your stacks created, save the game.
4. Smash your stack into the AI stack. Its much, much quicker if you enable the stack attack option here, and results for wins/losses can be calculated from remaining troops, or are available in the info tab (F9)
5. Reload and repeat for as many data sets as you feel is necessary.

If you want to use reload to get more results it is vital to have the new random seed option I mentioned in step one enabled, else you will get the same results every time......

There was another thread like this not long ago....
Leading to this one

[Sinyail]

Even if the way that the random numbers are generated is known, this still doesn't exclude Civ Iv combat being rigged. There could exist a hidden element or something that the curious people who peer into the games files do not notice or link to the "random number generator." Just saying that if the company wanted to rig the combat system and keep curious people unaware they could do so easily.

[Sinyail]

That sounds just like my post. Yeah it could be simple user frustration, but losing 99.8% combat rounds extremely frequently makes me suspicious.

[Ghpstage]

It is caused primarily by human bias, the majority of people only really notice a bad odds loss when its an important one. For example losing a great general unit, or not taking a city with their final unit, or losing a hilltop city with a fortified archer to a barb warrior.
Unlike us, the RNG doesn't care how important the battle is!

[oldskald]

Originally Posted by Ghpstage

It is caused primarily by human bias, the majority of people only really notice a bad odds loss when its an important one. For example losing a great general unit, or not taking a city with their final unit, or losing a hilltop city with a fortified archer to a barb warrior.
Unlike us, the RNG doesn't care how important the battle is!

Exactly.

The other things to note with human bias are;

1) How many times has suicide siege actually survived? We don't notice as much because we are not so upset. The AI is forced to fight at very low odds much of the time but the human player isn't. It would be interesting to see a human reaction to fighting dozens or hundreds of times at 0.1% odds.

2) How many combats do we fight? In a large map with active privateers I fight A LOT. Last night I lost a privateer at 99.4% to a caravel. That's the way it goes. I'm playing a marathon game on a huge fractal map and I must have fought 500 or 600 combats so far. At those odds for that number of combats it was only a matter of time before the good ship "Jolly Roger" got sent to the bottom.

The idea that the RNG is rigged just doesn't really stand up. There are lots of things about cIV combat that could be improved but the RNG really isn't something that worries me. Loosing sometimes at 99.9% has to be expected. I'd be more concerned if I was consistently loosing 80% of combats at 50% odds...

[Sinyail]

What I'm specifically talking about though is a pattern. I think that the Civ IV combat may be fixed to make someone who has consistently lost all of the sudden win. Vice versa also. Someone who has consistently won, lose. I mean with every thread that I have read regarding this issue there is no evidence one way or the other, but in the end all of the "regular" people who visit this forum and more or less know each other end up saying "Oh this guy is just venting. It is obviously just the random number generator getting on his nerves." This seems like bias to me and a refusal to actually admit that there is a possibility that the combat could be fixed. Every post that I have read regarding this matter ends exactly like this. "You are wrong. The combat is not fixed." I contend that it may very well be fixed and no one wants to tackle the issue head on because every one is just too comfortable in the mass thinking that is "Of course the games combat is not rigged." Is ignorance bliss? Not for me, it's not.

[Sinyail]

But then again I could just be an idiot

[ParadigmShifter]

So how familiar are you with C++?

[Sinyail]

C++? I'm completely ignorant. I still don't think that negates the possibility of fixed Civ IV combat odds though. I think that this issue has been thrown under the rug so to speak. I know people have done tests and this and that (500 units vs 500 units, etc), but still I think that there may be a specific and situational instance in which the computer makes a person win or lose. This is also the main thing that people I have seen complain about. A situational issue. Something that would be next to impossible to test and figure out and come out with a satisfactory answer in the end. Going on a winning streak and then all of the sudden losing at 99%+ odds is just fishy. Even more so when it happens consistently.

[Sinyail]

Ok after reading many posts from "God-Emperor" from this thread: http://forums.civfanatics.com/showth...=426745&page=2 I have officially changed my thinking. Civ IV is not rigged and I'm just a . lol k all is well now.

[Ghpstage]

If you carry on using the random numbers eventually all streaks will end. If there was something intentionally placed in the results to stop winning streaks i.e. an instant loss after 10 wins, it would be extremely obvious in large scale tests at both very high and very low odds.

People more knowledgable than me have tested it much more thoroughly and have found very minor issues, but those are caused by the limitations of the RNG, not from intentional meddling.

EDIT -

Ok after reading many posts from "God-Emperor" from this thread: http://forums.civfanatics.com/showth...=426745&page=2 I have officially changed my thinking. Civ IV is not rigged and I'm just a . lol k all is well now.

Good to know
Most people who come here making claims about the RNG tend to be immune to reason

[Sinyail]

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

[ParadigmShifter]

Why would they do that? They'd have to call the code that isn't published - the hundreds of modders would notice that.

You could step into the code that wasn't published anyway using a debugger.