Arioch's Analyst Thread

[PieceOfMind]

Spearmen practically never beat tanks in civ4 anyway. Usually complaints are more about longbowmen which, when beefed up with a fair number of defense bonuses (often the case!), could defeat tanks every now and then. A spear needs about a 265% defense bonus before it has even a 1% chance of beating a tank.

If a technological edge in combat becomes even more important, won't it make optimising research so as to gain military techs faster even more important than in civ4?

 

[Aramel]

Spearmen practically never beat tanks in civ4 anyway. Usually complaints are more about longbowmen which, when beefed up with a fair number of defense bonuses (often the case!), could defeat tanks every now and then. A spear needs about a 265% defense bonus before it has even a 1% chance of beating a tank.

I think a lot of the spearman vs. tank complaints were from people who blitzed their tanks in after taking heavy damage in the first attack, and then were amazed their 5 strength tank got taken out by a 4 strength spearman.

 

[Souron]

Spearmen vs tank seems to be a problem in every other version of civ. It was a problem in civ 1, so in civ 2 they added firepower and hitpoints. In civ 3, they changed it back, and again spearmen vs tank was a bit of a problem. They fixed it again in civ 4.

If the didn't take the spearmen vs tank problem seriously again, it has a real danger of coming back.

 

[12agnar0k]

Civ4 used a % chance system, so a tank had a 99.9% chance of beating a spear, so it wasn't likely a spear could win, but its possible. Also a large number of spears could plough into the tank damaging it each time and eventually get a good % chance to kill it.

In civ5, we can't be sure how the combat results are determined, it may still use some kind of "chance" factor, but then again it might not, it certainly doesn't show it. I think civ5 may very well deal in certainties as can be shown below.

Spoiler :

The end result is shown before the combat, the battle result is determined, the health loss is calculated, and its all shown, so one can summarise the random % chance of loss is gone, which In my opinion is a good thing, it only ever frustrated me.

But what is clear is that units that lose still deal out damage according to thier strength figure, so while the "1 in 1000 spearmen's victory" is gone, the other issue from civ4 was that a spearman could greatly injure the tank even though it couldnt really win. Now instead of this being a random chance too and the amount of damage random as it appeared to be in civ4, it is actually calculated and shown, and is relative to strength. So depending on the strength of a spearman will depend how many would need to be sacraficed in bringing the tanks health down. While we don't have the strength of a spearman, it should be about half that of the pikeman, and we know its strength.

Spoiler :

and here are some caprissons in strengths as they increase through tech's.

So as we can clearly see (assuming a spear is half the strength of the pike, it could be a little higher like 1.5 time's less) then the tank will be 9-12 ish times stronger in strength, which means technically if a tank does nothing, doesnt heal, doesnt run off, doesnt upgrade with all the kills 10-15 spears could kill the tank. But in civ5 that would be a sizeable army of spears

If a tank was all alone in the middle of your terriotry you could surround it with 18 spears, (2 hex wide circle) and attack the next turn but generally the tank will be in a line with other units and then depending if its on a poking out hex or poking in hex or a striaght hex, it can only be attacked by 4 units in the safer "in position", 7 spears able to attack in the straight position and still only 11 spears in the "out" position. (Which may not be enough to kill it) So technically, unless your an idiot, you probably wont ever have spears killing tanks without other units helping out.

What I meant by "in" or "out" or "straight" is simple, a line in a hex grid can either be vertical, (which has in and out points), "sort of diagonal" which is a striaght line, or horizontal, which is a straight line, see Click

 

[Grey2ham]

@Source for Air,

Okay well that can be interpreted two ways, either as the Air zone can have both 1 plane and 1 missile
Or perhaps Air will work differently than that of military and civilians in that thier is no "upper limit" other than that imposed by carrier space / airfield space. so we might be allowed a few units in the cities or the carriers as apposed to just 1 or 1 of each, thansk for the quote.

In terms of gameplay my opinion is that it can only be interpreted in one way. Anything in the air layer can stack at a base, and that the base's capacity is the stack limit as you say. The base's capacity could be infinite. As the attacks are a 'mission' to the target and back to base (mimicking 'flight range' , if movement is 12 then flight range is 5 unless air attacks dont cost movement then its six) there must be stacking at base.

Note: Probably should add that the helicopter gunship we've seen I'd personally confine to the ground layer as it has no base requirements. It may very well be classed as Air for counter purposes but will not be present in the air layer.

 

[PieceOfMind]

Civ4 used a % chance system, so a tank had a 99.9% chance of beating a spear, so it wasn't likely a spear could win, but its possible.

With all due respect agnar0k, I know you made that figure up, or perhaps you misunderstood the > when the odds display says >99.9%.

When a tank in civ4 attacks a spear, assuming no other complications, the odds of it winning when rounded correctly are:
99.9999999%.

About 1 chance in a billion that the spear could win.

For all intents and purposes it was impossible. As I said earlier, it was usually longbows that caused headaches for people. Or as Aramel said, it may have been people using severely injured tanks.

EDIT... By the way, I find it very unlikely the random chance of combat is removed entirely. CivRev didn't explicitly show probabilities either but it still used them.

 

[The Leviathan]

12agnar0k:

The combat system for ALL the different editions of Civ, and virtually every single game ever made is based on chance. The only difference is that Civ4 calculates those odds for you and displays them in an easy to read format.

If a combat system wasn't based on luck, then it meets that when two units meet in the same conditions the combat will always end in exactly the same way. Ie, if a combat 2 knight attacked a combat 3 crossbowman in a plain with no flanking units or other bonuses then the knight will always (for example) take 1.3 points of damage and the crossbowman 4.6 points of damage. And this would be exactly the same every single time a battle like that happens.

This would be ridiculously boring, remove most of the strategy and all of the risk taking from combat. Civ (and 99.999999% of all other strategy games) has never worked like that, and never will.

They could always make the damage a unit does deal a range of damage instead of a static number. So instead of a unpromoted knight taking 2 damage and a unpromoted crossbowmen takes 5. They could make it so the crossbow men could take 4-6 damage while the knight can take 1-3 damage. For example. And that's not counting defensive bonuses on tiles and promotions.

The Rock Paper Scissors formula, Ranged combat, Promotions, Generals, and Defensive tile bonuses will make 1UpT combat infinity more fun than IV's current system if done well IMO. Having a lower chance of RNG effecting the outcome of combat than in IV doesn't suddenly make combat less fun. However I agree that there has to be some chance/RNG involved in the combat.

 

[KMRblue1027]

They better not use CivRev's RNG. Loosing a 27 defense archer army to a 5 attack knight was terrible. It was even worse in multiplayer because of the desync where combat happens and you didn't even see it. I swear that game conspires against the human player so bad it's not funny at all.

Also looking at the Civ info I think Arabia is going to be very over powered. Double resources from oil is going to murder in the modern era. They can build so many tanks.

 

[Silvestris]

Maybe Tanks will need more than one ressource.

 

[12agnar0k]

OK I understand what you guys are saying, in most games yes chance is left it, but the thing is, when the results of the combat are displayed before combat (not probable results but the actual results), this is removing the "chance" factor. As normally you do the action, the chance factor comes in, and then you get your results, if you know the outcome definitively before the action is even attempted, this is removing chance.

Also I think civ4 rounds too 2 decimal places. I can't remember if it was 99.99% when a tank is facing off against a spear, or if it was 99.98 or 99.97 %, the later suggests it to be 2-3 in 10,000 not 1 in a billion, though again it depends a little bit on if they use the rounded odds in working out the victory, or if they use actual odds which could be 99.9849999999999999 for example and then only round them for us to see.

 

[Gre_Magus]

12agnar0k: Thanks for posting those odds, they're pretty interesting. The calculated strength seems pretty straightforward: it's 1000xStrengthx(1+SUM(Modifiers%)) There does not seem to be a penalty for starting a combat with low health EXCEPT in the one example where one tank is attacking another (whose starting health is very low and has a 3350 strength instead of 5000, with no visible modifiers). The total strength for the cannon being attacked by the cavalry also seems off - I don't know if that's from an earlier build where modifiers aren't displayed.
You can see the breakdown for current health and health after the battle (Green, Yellow, and different shades of Red). There's no chance of victory or range of damage indicated here, which makes me wonder whether: there is no random element to combat or the random element has already been figured in this prediction. I'm not sure what I think about that...of course, are we sure that these are combat predictions and not combat results?

 

[Celevin]

The only change compared to Civ4's system for unit on unit combat is it should only go for a finite number of rounds.

This would mean, keeping in all damages that a longbowman and tank can do, even if the tank rolls really crappily it will probably not die. It lets units "hold their ground" for a couple of turns without expecting to win, as long as they've got enough health.

The fact that Civ4 goes until one side wins is what causes sweeping changes to the landscape in the span of just one turn.

 

[craig123]

Does anyone have a link to the gamestar video? Would love to see it!

 

[Schuesseled]

Maybe Tanks will need more than one ressource.

I wouldn't be suprised if they used up one iron as well as one oil (per tank)

 

[PieceOfMind]

Also I think civ4 rounds too 2 decimal places. I can't remember if it was 99.99% when a tank is facing off against a spear, or if it was 99.98 or 99.97 %, the later suggests it to be 2-3 in 10,000 not 1 in a billion, though again it depends a little bit on if they use the rounded odds in working out the victory, or if they use actual odds which could be 99.9849999999999999 for example and then only round them for us to see.

Dude, trust me. You're talking to the guy who made Advanced Combat Odds :hint hint: and the one who spotted a bug in the BtS odds calculator (that still exists to this day FYI, but not in ACO where it's fixed).

Since you're interested, the combat odds are only rounded in the display of the calculated odds. The function of the RNG during combat is completely separate from the odds display. For more information on this, please refer to the Combat Odds and RNG thread in my signature. For each combat round a 1000-sided "die" is rolled. So technically yes there is some rounding but it is on a per-round level - not per battle. Lord Olleus had a good way of describing it in his reply.

I will say it again, but yes the odds of the spear beating the tank are about 1 in a billion. Nowhere near 1 in a thousand as you suggested.

EDIT...
And on the matter of the "Major Victory" and "Major Defeat" etc. that gets shown, I'd be more inclined to say they are indicating what your more likely outcome is. Predetermined battles in a civ game would make warfare incredibly lame and also more exploitable. That's why I really hope your suggestion is wrong.

 

[alienXorcist]

Some screenshots from the GameStar.de video:

Moderator Action: *snip*

 

[Jaybe]

EDIT...
And on the matter of the "Major Victory" and "Major Defeat" etc. that gets shown, I'd be more inclined to say they are indicating what your more likely outcome is. Predetermined battles in a civ game would make warfare incredibly lame and also more exploitable. That's why I really hope your suggestion is wrong.

Now, I do not totally recall the vids, but is it possible the combat results are from AFTER the combat was committed, but before the combat animations were displayed?

 

[Aramel]

I was wondering what they'd be doing with great people in Civ 5. It seems like from that city screen that they are tracking GPP for each type of specialist separately, but that it's still split between cities like in Civ 4. Interesting. Should get rid of those situations when I'm working a bunch of scientists in a city but it still randomly pops an artist.

 

[bjbrains]

I was wondering what they'd be doing with great people in Civ 5. It seems like from that city screen that they are tracking GPP for each type of specialist separately, but that it's still split between cities like in Civ 4. Interesting. Should get rid of those situations when I'm working a bunch of scientists in a city but it still randomly pops an artist.

Yea, it looks like GP will be on independent counters that appear to be based (partially) on how many cities you have. Perhaps the required points no longer increase with each GP? (or to a lesser extent?)

 

[Souron]

The Berlin city screen is interesting. First of all "resource demanded: sugar" reminds one of civ 2. Are civ 2 trade routes like back?

Second, it looks like there's a road leading to a furs resource, but the position of the resources is such that it cannot be in any city. This means that it must be possible to get resources from outside your cities, as was enable by colonies in civ 3.

Third, the great people match up with the three specialist buildings. This suggests that specialists generate GPP as in civ 4, except now GPP are so that each type of great person accrues points independently.