RB1 - Cuban Isolationists

[ChrTh]

I dont even turn off the browser only do a refresh from time to time and in the morning (ohhh... same page opend 24/7 at work and at home)

Oh yeah? Well I have the updates beamed DIRECTLY INTO MY BRAIN so I know what happens even before it appears on screen!

Of course, not too many updates worth beaming lately ...

 

[thedaian]

Someone better update soon, or Sirian and Sullla will have an angry mob on their hands.

 

[LandedGentry]

I know this sounds horrible, but after reading this thread, it seems almost unbearable reading other, less eloquent and less experienced player threads. Not to disrespect the other SG players, (as I am not good enough to even think about SG), but it really is almost like putting down a Pulitzer winning Novel and going to comic books.

 

[Ozymandous]

I think it's time to quote Han Solo



In my opinion, hopefully the person would have enough experience with the game to quickly glance at the matchup and determine whether its risky or not. I don't care if my odds of success (or even failure) are 55% or 65%. It's when the odds are 80% (in either direction) that it becomes an issue, and I should be able to tell at quick glance whether I have an 80% or more chance of failure/success. And I think showing the strength plus the modifier icons is enough for me to make a "blink" judgement.

Plus it's fun not to know the odds too closely. Look at the original issue: Arathorn (of all people) forgot about the +25% Infantry get against gunpowder units. If I'm told the odds every time, I don't need to learn info like that: the computer will calculate it for me. But that +25% strength could be the difference between victory and defeat in a close battle, and I prefer a game that rewards intelligent decision-making based on my own knowledge rather than relying on the game to make sure I'm fully informed. Same with Hills, terrain, etc. The game's just not as much of a challenge otherwise.

Hmm.. I'd have to disagree a bit with the last bit. Intelligent decision making would be to know to attack or fortify in woods, on a hill as opposed to grassland or plains, etc. That's using intelligence to pick the "right answer". Having to calculate if your 2/3 healthy unit can really beat a 3/4 healthy, but 'weaker' unit isn't using your game intelligence, it's number crunching.

Yoyr example of Arathorn forgetting the 25% bonus vs. gunpowder units is not relevant since that would (or should) already be in the unit calculations with you mouse over the potential combat to see the unit strengths. I assumed it was already calculated int he display but I could be wrong.

In any event, the "intelligent decision making" isn't number crunching any more than it was micro-managing your way to a win on the higher difficulty levels. Intelligent decision making should entail having all the facts, and then making the best choice you can based on available information. Pouring over every tiny detail and trying to tweak all the minutae to make sure everything is 'perfect' isn't decision making, it's another term for micro'ing, which most of us want less of, not more.

Players should -not- have to pull out a calculator or some spread sheet with formula's or whatever to know the reality of the situation and the real strengths of the units in the game including their health as well no matter if the units are used on attack or defense.

Say that people would still complain if they lost that 1% of the time and you'd be correct, however, would the vast majority complain if they lost a seemingly superior unit, like Infantry, to something like an axeman even if the wounded Infantry appeared as strong (was wounded down to the axeman's strength) but was actually much more likely to loose?

I think if people knew the -real- unit strength based not only on promotions and base strength but hit points as well, you would generally see less complaining. People complain when what they see doesn't match the reality, and as long as there are 'fake' or incomplete numbers involved people will continue to be misled because what they think they know, and what they've been told, isn't exactly how the reality of the situation is.

Having all the facts, and not having to 'guess' about what's going on in the background can be the difference between a highly sucesful game, and one that fails almost completely, MOO3 was a prime example of that. It wouldn't be hard, IMHO, to simply show the chances when the hitpoint/health units is also factored in. After all, we all know this calculation is what's really used to resolve the com,bat, so why not simply use it for the display chances/strengths anyway?

 

[Toni Metsänkylä]

I know this sounds horrible, but after reading this thread, it seems almost unbearable reading other, less eloquent and less experienced player threads. Not to disrespect the other SG players, (as I am not good enough to even think about SG), but it really is almost like putting down a Pulitzer winning Novel and going to comic books.

My thoughts exactly!

Been reading this thread for a long time and finally decided to register again. Sirian and Sulla are very good in telling an interesting story as well as playing an interesting game! I've also looked at the other SG's but they lack that 'something'.

I'm getting my copy of CivIV this week, going to crank it up on Win98 (heard it works!) and see if 512MB is enough...

 

[Beamup]

It wouldn't be hard, IMHO, to simply show the chances when the hitpoint/health units is also factored in.

That's the thing though. It would be much harder. What is currently displayed is a near-trivial calculation: A/(A+D). Calculating the actual odds of winning is orders of magnitude more complex. And you need a very significant amount of mathematical knowledge to work out how to do it at all. And, yes, it's far more complex than just resolving the combat, too - which is simpler, figuring out the odds of getting more than 70 on (1d6)d10+5d8 or just rolling the dice and seeing what you get?

Could it be done? Probably (though I wouldn't be terribly surprised if none of the programmers knew enough probability to work it out). Would it be a worthwhile use of man-hours, particularly considering the fact that it's not universally desirable? That's a lot harder.

And that's assuming it wasn't a deliberate design decision in the first place!

No, it's very far from clear what "should" have been done.

 

[vergoetter]

That's the thing though. It would be much harder. What is currently displayed is a near-trivial calculation: A/(A+D). Calculating the actual odds of winning is orders of magnitude more complex. And you need a very significant amount of mathematical knowledge to work out how to do it at all. And, yes, it's far more complex than just resolving the combat, too - which is simpler, figuring out the odds of getting more than 70 on (1d6)d10+5d8 or just rolling the dice and seeing what you get?

Could it be done? Probably (though I wouldn't be terribly surprised if none of the programmers knew enough probability to work it out). Would it be a worthwhile use of man-hours, particularly considering the fact that it's not universally desirable? That's a lot harder.

And that's assuming it wasn't a deliberate design decision in the first place!

No, it's very far from clear what "should" have been done.

I would guess that it was/is a design decision. There's no "right" choice between probabilities and strength values, and I'm not trying to say there is. I am saying that there seems to be an interest in a mod or something that shows probabilities instead.

My first try to accurately work out the probability would be to use a Markov chain. That's entirely possible to work up programmatically. I don't really know how the Civ 4 scripting system works, but I hear it's in Python, so it's doable.

 

[Wyatan]

Umm, no. The whole reason to base the calculations on the real strength of the unit (including hit points, although it's hidden now) vs. it's "best strength" when fully healed is to balance game play and have players choose to either keep attacking or stop.

I'm not suggesting to base everything upon the full strength of a unit : only the damage part.

Thus a wounded unit would still die quicker and hit less, but it would hit as hard... when it hits.

In the example that started the discussion, the only thing that would be changed would be the damage of the defending infantry : 29 instead of 21, meaning 4 hits needed to kill the attacking rifles instead of 3. The change wouldn't be dramatic, but it would even the odds somewhat.

Currently, a 10/20 vs a 10/10 :
- hits as often
- hits as hard
- dies twice quicker
--> twice weaker
That seems kinda broken to me...

With the change I suggest, a 10/20 vs a 10/10 would:
- hit as often
- hit nearly twice harder
- die twice quicker
--> roughly even
More intuitive, more balanced, but maintains a different behavior...

 

[KrikkitTwo]

I think the big problem is the fact that hps DON'T stay at 100 like was previously thought, meaning that the Strength Values can be HIGHLY misleading, ie the unit with a higher strength displayed Does NOT have the higher chance if it is more damaged. (as pointed out by Wyatan)

While that was a useful gameplay decision, the display of relative strengths is misleading, and simply modifying it so that ithat display of strengths also included some line that mentioned relative amounts of damage (say an * by a unit that had less than 2/3 as many hp as the other one, or ** by a unit that had less than 1/3 as many hp as the other one) This would give SOME indication that the Strength values need to be taken with a grain of salt

actually when displayed, there should probably some type of 'disclaimer' list (all factors that are not displayed in the strength but affect the results of combat)
* X First Strike
* X % Collateral damage
* X % Retreat probability
* low hps (=<2/3 of opponent)
* Very low hps (=<1/3 of opponent)

Where the * indicates that this factor affects combat beyond the strength calculations

 

[ChrTh]

I think I agree with the idea that either strength or hit points should be reduced, but not both. A unit that has taken 50% damage is effectively 75% weaker versus a similar opponent with the same base strength and modifiers; as Arathorn noted, they are redlined at that point.

I'll use a concrete example:

I have an infantry squad of 4 members. Now, if that squad takes 50% damage, what should the result be? Either
--2 of the members are dead, leaving 2 to fight, or
--all 4 suffer 50% injury, and are each half as effective

But that's not what's happening in the game. 50% damage translates into
--2 of the members are dead, leaving 2 to fight, AND
--The 2 surviving members suffer 50% injury, and are each half as effective.

That seems squirrelly to me (of course, I may be misunderstanding things).

 

[weakciv]

...
I have an infantry squad of 4 members. Now, if that squad takes 50% damage, what should the result be? Either
--2 of the members are dead, leaving 2 to fight, or
--all 4 suffer 50% injury, and are each half as effective

But that's not what's happening in the game. 50% damage translates into
--2 of the members are dead, leaving 2 to fight, AND
--The 2 surviving members suffer 50% injury, and are each half as effective.
...

I thought the exact same thing after reading that HP factored into combat effectiveness in Civ4.

in a simplistic example:
A str4 unit has 4 members, each do 1 unit of damage.
This str4 unit now takes 50% damage
This results in the unit only having 2 members, each doing 1 unit of damage. A str2 unit, right?
Wrong. It results in 2 members each doing .5 units of damage. A str1 unit.

same thing you wrote ChrTh (just in a way I understand. I hate %)

 

[Beamup]

No, it really does make relatively realistic sense. 2 of the members are dead, so it's both easier to kill them and their firepower is reduced. i.e. HP and strength both go down.

But it's really more than 75% weaker (87.5%, really) - they're less likely to hit, do less damage when they hit, and take less damage to kill. Essentially combat power goes as (HP)^3, not just squared.

Now, what would really make sense is to recalculate strength ratios for each round of combat - not just for each individual combat. So if you lose the first round, you're weaker for the second.

 

[Spatzimaus]

Now, what would really make sense is to recalculate strength ratios for each round of combat - not just for each individual combat. So if you lose the first round, you're weaker for the second.

Of course, if you did that, then First Strike would take over the game. If I could get a few extra hits in on you, before you have a chance to start fighting, I'd win easily.

 

[knupp715]

I personally think that the game should have showed the "right" odds. But the fact is it doesn't. Most likely just a programming decision. Either way, this thread used to be about the Cuban Isolationists IIRC. So let's get back on topic. I want to see some Cuban armors kill the Greek infidels!

 

[weakciv]

I personally think that the game should have showed the "right" odds. But the fact is it doesn't. Most likely just a programming decesion. Either way, this thread used to be about the Cuban Isolationists IIRC. So let's get back on topic. I want to see some Cuban armors kill the Greek infidels!

I agree whole-heartedly

 

[nevs69]

Cuba! CUba! CUBa! CUBA!!!!!!

 

[KrikkitTwo]

No, it really does make relatively realistic sense. 2 of the members are dead, so it's both easier to kill them and their firepower is reduced. i.e. HP and strength both go down.

But it's really more than 75% weaker (87.5%, really) - they're less likely to hit, do less damage when they hit, and take less damage to kill. Essentially combat power goes as (HP)^3, not just squared.

Now, what would really make sense is to recalculate strength ratios for each round of combat - not just for each individual combat. So if you lose the first round, you're weaker for the second.

Actually its to the 4th (since they not only do less damage, they also take more)

 

[Sirian]

Actually its to the 4th (since they not only do less damage, they also take more)

In the real world, the amount of casualties inflicted on a unit before its military effectiveness is deemed to be "destroyed" is nowhere near 100%.

Not that this means anything -- I'm regularly pointing out where gameplay has trumped realism -- but at least the realism factor is there in this case!


How many of you have ever played a 4v4 teamer in Warcraft III? Ever had one of your teammates drop in the early going, reducing you to a 4v3? What are your odds in that situation? Four to three? *snort* Yeah, right.


Anyway, Civ4 combat involves a LOT less luck than Civ3 combat, but still enough that you don't have to worry about crunching all the numbers. If you are in a war in Civ4, you are in the war. It's not so much about theoretical ideals, at that point, but about deciding whether or not to attack. The idea that you will be able to control things to where you only attack when the odds on the FIRST battle are greater than x% is amusing me. The fact is, you will often have to fight long odds, but then if you do SOME damage, the odds go way up for subsequent units.

I find that this does a very nice job of "simulating" the results of simultaneous warfare, within a sequential framework. That is exactly what Civ and other games like it have been MISSING all this time. Instead of results out of some kung fu fantasy movie where all the Ensign Slods of the world line up single file to take on the hero in 1v1 combat, we get a much more interesting "gang up" effect, where it is as if six of them come at him simultaneously and he's going to go down!


If someone wants to mod a real combat odds calculator, have at it! That's why the modding is so open. However, I do recommend that more than one knowledgable set of eyes examine any such mod, to make sure that it is NOT a Diablo2-style "Lying Character Screen".


As for updates... I'll try to trickle some in here, some there. A holiday weekend is coming up, though, and Sulla can't take his next turn until afterward anyway, so no hurry, right?

BUT THAT'S OK! I'VE GOT GOOD NEWS!

(I just saved a bunch of money--) NO! GET THE DEMON OUT OF MY HEAD! !@#$%! "Madison Avenue" marketing slogans. !@#$%!


I'VE GOT GOOD NEWS!

While the turn logs will be a bit slower, I will continue to entertain and amaze you all with my fabulously sharp wit. The "Laugh to Groan Ratio(TM)" will be at least one to three OR BETTER! Get your tickets now!


- Sirian

 

[perky_goth]

since i've registered in order to follow the thread better, i might as well post.
good going it saddens me that the victory types are kind of broken... the only way to win a conquest is to have a huge lead everywhere else, as per this game.

as to the odds, i can't see a reason why the correct ones aren't seen. it isn't computacionaly intensive (a modern pc could calcute a few million cases per second).
it is important because unless you know the numbers well (or play a lot), you might not be able to realize if a war is worthwhile, or if you should skip targets, or wait a turn... so the combat isn't as random, but you can't tell unless you know the math.
i'm rambling. cute game, guys.

 

[Mazarin]

the main point of this post is to subscribe to this truly amazing thread. Great game, guys and one of the most entertainig write-ups I've ever read. I once had the chance of playing in a civ3-SG with Sirian and would definetly love to repeat that with civ4

One comment on the odds-display discussion

That's the thing though. It would be much harder. What is currently displayed is a near-trivial calculation: A/(A+D). Calculating the actual odds of winning is orders of magnitude more complex. And you need a very significant amount of mathematical knowledge to work out how to do it at all.

It is harder but not by that much: a very basic statistics course should be sufficient to do that. I do think that only displaying the strength ratios as opposed to the actual odds of combat was a wise move by Firaxis considering many of the casual players: We all know how many complaints there were because of combat results in civ3: an attack 10 unit losing to a defense 3 unit was reason enough for some people to endlessly scream around how the AI was cheating in combat. Although the odds were more than 10:3 in favour of the stronger unit people were still mainly noticing the unfavourable results and felt cheated. We now have a system where the gap between strength ratios and probabilities of success has been widened for healthy units which might reduce the amount of perceived injustice. Just imagine the display showing 99% probability of winning a battle and some of the aformentioned users losing two such battles in a row...this and other fora might go down in flames that way However, for those of us who can handle unlikely outcomes better, a modification of actual displays would be great, and I'm sure that the great modding community will provide us with these tools if they are technically possible.