RB1 - Cuban Isolationists

[EmperorNapoleon]

Blink is an excellent book I'm amazed by the way that one guy can predict relationships!

Anyways, On Topic, hurry up with the turns!!

 

[ChrTh]

Blink is an excellent book I'm amazed by the way that one guy can predict relationships!

Next week, the Civ 4 SG board book club will be reading "Wisdom of the Crowds"

Anyways, On Topic, hurry up with the turns!!

Hear hear!

 

[weakciv]

S&S, you guys have created a frenzied mob here. I even understand people are shirking work or school or (gasp) family, waiting for each and every game update.

Field Reporter: (chants of CUBA! CUBA! in the back ground)
It is utter chaos down here. People camping out, calling in "sick" to work. An older couple even put off traveling the country in their RV just to wait, watch, and listen here.
If something doesnt happen soon, we may have a riot on our hands. Back to you in the studio.

 

[Spatzimaus]

If people continue to receive poor results than they expect, aka "My 10/20 units keep losing to 10/10 units, why, why, why!?!?!" based on erroneous information given out on threads on forums/sites and/or erroneous info, or left out info in the game manual, tutorial, etc, they will not necessarily stick around to become that 'expert'.

There's no happy middle here.
> If you leave it the way it is now, you get people complaining that the numbers are misleading, telling them they have an advantage/disadvantage when in fact the true odds went the other way due to additional factors. (Example: the Rifleman fights that started this discussion) The advantage of the current system is that it's not actually the odds, and people should KNOW it's not the odds (or at least they quickly learn that). It's the chance of winning a single round of combat, no more, no less, so there's no need to consider exactly how many rounds there'd be.
> If you compute the EXACT odds of winning the entire fight, which I personally think would take a lot more effort than it's worth, then you get the "robot" effect. This happened in Alpha Centauri, to some extent. And for some people, it would make the complaints worse; "I had a 99% chance of winning and I still lost! This game is broken!" Since you'd never be the one attacking with only a 1% chance of winning, you'll only ever see the flip side, when the computer pulls a Spearman-beats-Tank upset.

You could try to make a middle option; multiply current strength by health again, to generate a crude "power" rating, then turn it into a win/loss percent. But that'd in many ways combine the worst of both extremes; it'd be close enough to the true odds that new players would come to depend on it at the expense of actual decision-making, but it still wouldn't be the correct odds (for example, it'd leave out the effects of first strikes, or the quantization of health), so you'd have complaints more often when the results didn't match expectations.

 

[MeteorPunch]

The in-game calculator should be 100% accurate (like we all thought it was when we first used it), but those who want to guess should be able to turn it off.

 

[Spatzimaus]

The in-game calculator should be 100% accurate (like we all thought it was when we first used it)

Having played Alpha Centauri, I never, ever thought the calculator was a 100% accurate representation of odds, so subtract one person from your "we all" generalization. Anyone who thought about it for a second would realize that it wasn't including several factors (first strike and withdraw, for instance), even before realizing the health issue.

But again, what's given on the screen is not the "odds", it's the unit's current Strength, and the documentation doesn't say otherwise. It IS "accurate"; the Strength that goes into the damage equations, and that determines which unit takes damage each round, is exactly what's printed on the screen. It's just not what you personally want to see, but that's not the same as saying it's inaccurate; in the most common situation (two units with the same health and no unusual promotions) this mimics the odds fairly well, but that's only a happy coincidence.

Now, you could figure the true odds. It'd take a little work, but it's doable. But, the obvious corollary of this whole discussion is that the AI must have access to the true odds for its decision-making. It probably isn't going solely by the units' current Strength, and is taking things like damage or first strike into account. So, the functionality must be there already, and once we have access to the SDK, we can bring this to the screen without consuming any extra processor time. (Unless it turns out the AI doesn't think this through, in which case we can improve the AI's performance while we're at it!)

 

[jdotmi]

So, I started reading this thread three days ago... Like a previous poster, I assumed the game was done due to the size of the thread...

...



Someone needs to post... one... more... turn...

 

[ChrTh]

Now, you could figure the true odds. It'd take a little work, but it's doable. But, the obvious corollary of this whole discussion is that the AI must have access to the true odds for its decision-making. It probably isn't going solely by the units' current Strength, and is taking things like damage or first strike into account. So, the functionality must be there already, and once we have access to the SDK, we can bring this to the screen without consuming any extra processor time. (Unless it turns out the AI doesn't think this through, in which case we can improve the AI's performance while we're at it!)

Does the AI have full access ... and would we want it to? That's a thought to ponder.

I can't help but think that if the AI had full access, it would never lose when it attacked (or it'd be a 1 in 10 chance of losing) ...

actually, maybe that's what dictates an 'aggressive' CIV: lower odds threshold. A non-aggressive CIV only attacks at 75% or better, while an aggressive CIV attacks at 50% or better. Hrmmmm (I looked for a good 'thinking' smiley, but I couldn't find one so instead I'll put up :suicide: )

 

[Spatzimaus]

I can't help but think that if the AI had full access, it would never lose when it attacked (or it'd be a 1 in 10 chance of losing) ...

I was thinking about that; I don't think it's a problem. There are times when a 50-50 fight is worth the risk (for instance, if the town you're attacking is on its last defender and you won't have any reinforcements coming in time), and there are times when you shouldn't accept less than an 80% chance. A human can easily tell the two apart, while an AI wouldn't do so well. This has always been the problem in previous Civ games; either the AI wasn't aggressive at all, or it would throw units into any fight regardless of odds, with very little in between.

Besides, we're not talking about a huge range here. If the true percentage is 63.41%, a human who's played the game long enough would probably know it's in the 60-70% range, i.e., the "more likely than not, but still a good chance of loss" category. Is there really a huge benefit to knowing the exact number? So, having the AI know the exact odds wouldn't really change anything; you'd still be more limited by the algorithm that determines exactly when a risk is justified.

 

[Ozymandous]

Back up a second ... I wasn't. If you look at my post closely, my issue is when the player makes decisions like "I'll only attack if I have a 87% or higher chance of winning" ... a combat calculator is not intrinsically robotic, however, it can be an enabler to robotic behavior. That is my concern.

My other concern is that all of this focus on combat modifiers and strengths and calculators, etc., will result in the following "rule" in an SG:



I want games--including SGs--to allow for the decision to go against the odds. To take a risk when the situation requires. It makes for better games. And better reading.

Yes, LCS' are bad blah blah blah. But I don't want us to overreact so far that all the fun is taken out of the game. If Firaxis adds a true odds calculator to the game, I'm going to turn it off the second it's there. Because if I wanted to play a math game, I wouldn't have purchased Civ.

So tell me soemthing... What's the difference between the 'rule' you fear as posted above if unit health was listed in the probability display and the following rule:

No participant may attack with a unit unless the unit is 25% stronger than the opposing unit. For example, a strength 5 axeman cannot attack until or unless his opponent is < 4 in strength.

Umm, there -isn't- a difference. Such a 'rule' as you fear could already be easily enacted with the way things stand now. That's not relevant. Your same 'fear' of increasing "robotic behavior" could already be in effect due to the simple strength's shown IG already, but that's all beside the point.

The issue is thus: If they are going to show relative strengths, then why not show the effects of all the variables that effect those relative strengths?

As already mentioned it's not like they need to do Nobel science award winning computations to achieve a relative display of the odd's of success. Arathorn has already shown that, and I don't think I have seen you counter my initial argument, which is thus:

Players who see that units seem equal in strength but aren't (really 50/50 on odd's of success), would accept taking a chance and losing as opposed to someone who see's a battle of 10/20 vs 10/10 and not understanding at all why a wounded, but newer/superior unit lost to an older/weaker one, all things considered.

I think if someone saw their odd's of winning were 25/75 with a 10/20 vs 10/10 as opposed to 50/50 would more readily accept the loss than otherwise. At least those players who wouldn't complain at the drop of a pin anyway.

 

[KrikkitTwo]

My other concern is that all of this focus on combat modifiers and strengths and calculators, etc., will result in the following "rule" in an SG:



I want games--including SGs--to allow for the decision to go against the odds. To take a risk when the situation requires. It makes for better games. And better reading.

Yes, LCS' are bad blah blah blah. But I don't want us to overreact so far that all the fun is taken out of the game. If Firaxis adds a true odds calculator to the game, I'm going to turn it off the second it's there. Because if I wanted to play a math game, I wouldn't have purchased Civ.

But the fact is
1. you can Currently have such a rule (ie no attacking unless your Strength is more than 50% of the opponent, or something like that)

2. If you didn't want to play a math game you Shouldn't have bought Civ.... who wins in a Wonder, Religion, Space, or Culture Race=Math
Who wins a War=Math
When a city flips=Math
What barbarians attack your city=Math
Whether or not you get Iron=Math
All math, although the inputs to that math include player decisions and an RNG in some cases.

The decision is made interesting because of unknown information
in the case of Wonder/Culture races, the Unknown is the whole other player side of the input, you don't know exactly their state. Same for a War.
For a Battle the Only unknown information is the RNG number, and that Should be the only unknown information.

Currnetly however there is information that is known (technically) for the Battle but is not presented clearly.

Now if posting the exact odds of a battle could be dome simply, then it Should be done (perhaps it should be listed as chances of losing, so that the people will remember they have a 1% chance of losing this battle)

If an odds calculator cannot be made simply, then it would be better to at least give all the Known information in an easily usable form. (so something in the Strength that shows hp ratios..even roughly...as well as strength ratios and First strikes, and Retreats would present all the known information in an easily accessible way)

 

[ChrTh]

Deleted to get back on topic.

 

[ChrTh]

Deleted to get back on topic

 

[Zeviz]

Sirian and Sulla, this is a great game and very fun to read.

EDIT: Yes, I probably shouldn't be continuing off topic discussion.

 

[ChrTh]

Hey guys,

I've think we've gone a little too far off-topic. Talking about combat as it applies to the game, that's ok (if off-topic). Discussing personal philosophies about combat resolution? Too far.

So, I'm going to shut up on this topic. If you want to respond to my last few responses, that is fine (I don't want to be accused of trying to get the last word in), but in respect to Sirian and Sullla (and all the readers who I'm sure are getting sick of the email notifications), I won't be saying anything else on this topic.

probably.

maybe.

 

[ChrTh]

Sirian and Sulla, this is a great game and very fun to read.

EDIT: Yes, I probably shouldn't be continuing off topic discussion.

Damn. I liked your post

 

[knupp715]

Originally Posted by ChrTh
that's just me. My end point for everybody is that as long as we don't have an LCS, and one can generate the odds if he or she so chooses, then it's not really broken. It just doesn't do the work for you that you want it to do.

I hate to continue the discussion, but I just wanted to say I agree with everthing you are saying right here. Firaxis shouldn't have to make an extensive in-game calculator. They just need to supply the information that a player needs in order to do what he or she chooses. I for one don't think pulling out a calculator in each and every battle if fun. Especially because during long wars, a turn could take hours just because each time a battle is taking place you are using a formula to calculate the exact odds of winning on your calculator. It's a game, and I intend to have fun. Slaving over math problems, formulas, calculations and such is not fun (for me anyways).

 

[GenericKen]

Markov Chains? Why would such a simple probabilistic function need to be determined by markov chains? You're not localizing a robot or anything; there's no new randomized data coming in through which you determine your probabilities. The only variables that affect the outcome of a battle are the set stats that the progress of the battle thus far, which is also determined by the set stats.

Personally, I think that if a complete combat calculator is implemented, it should include the base str vs str, the overall %s, and the standard deviation on the %s. Even then, the decisions wouldn't be automatic, as the given value of the units being matched up against each other isn't just determined by shield cost vs shield cost (I'm sorry, hammers), but is determined by a large number of intangibles (whether you need a medic upgrade on a 4/5 exp guy, whether he's your last spearman, the vulnerability of your overall force if you lose a relatively safe gamble, etc.).

As is, I do think just the raw strength values is misleading, especially compared to the relative comprehensiveness of the other mouseover help screens.

 

[Sirian]

Haha! You guys are making me start to wonder...

What would be the most effective way to pull the thread back on topic?

A. Threaten not to post the rest of my turn results until 48 hours AFTER the last word is had on the Combat Calculator

B. I don't need Plan B! PLAN A !@#$%# ROCKS, MAN!




Serioiusly. Enough on that, please. Thanks.


- Sirian

 

[Sirian]

i totally agree with you!!! Free Chinese Checkers!! Hail Chinese Checkers!! Long live Chines Checkers!!


(sorry, but...What are chinese checkers?)

That's from Hoyle's Puzzle and Board Games 2005


- Sirian