Do we have combat all wrong??

[Venger]

Curiously, in each round of combat, is there a per chance hit on each side, or is there only a per chance to win the round? If it's the latter, then the system is already more screwed up than I thought...

Example:

Unit 4/2 attacks unit 1/1.

In the "per side" system, each side has a chance to hit, in this case the attacked would have a 80% chance to hit, the defender a 20% chance to hit. The chance of the defender hitting without the attacker hitting back is a very low 4%.

In the "per round" system, only one side can score a hit, so the attacked has an 80% chance of winning the round, the defender a 20% chance.

What does this mean for gameplay?

Assuming regular units and the per side system, the chances of the defender winning 3-0 is an infintesimal .0064%.

Assuming regular units and the per round system, the chances of the defender winning 3-0 is .8%.

Still small, but 125 TIMES HIGHER than the per side system. It would also explain the bizarre swordsman beats infantry scenario, where winning a round is easier than "going up" a round.

Let me explain -

With even hit points, winning is hitting without being hit. In the per side system, where each side has a chance to hit, this is much more difficult than the per round system. Using the above example, in order to hit without being hit, the 1 strength defender must score a hit (20% chance), and at the same time not be hit (20% chance the attacker misses). Those probabilities combined are only 4%.

Now, if it's per round, the chances are 20%, because if there is only one percentage check for that round of combat, and the defender wins in his 20%, by DEFINITION the other unit does not score a hit. That a 5 times higher chance to succeed.

Now, does anyone KNOW, for sure, that the system in Civ3 is a per round system? Also, can anyone confirm that which I am fairly sure of, that Civ2 was per side?

If Civ2 was per side and Civ3 is per round, that will compound the already statstically screwed up Civ3 system into a unusable mess. Oh wait, I think that's what we have...

Venger

[BigNick]

Originally posted by Venger
Curiously, in each round of combat, is there a per chance hit on each side, or is there only a per chance to win the round?

I'm fairly sure it's the latter. Each 'round', one (and only one)point of damage will be done. In your scenario, this point has an 80% chance of being assigned to the defender and a 20% chance of being assigned to the attacker.

I think this was the case in Civ2 and SMAC, too.

[WhiteElephants]

I'm not sure how the "per side" and "per round" styles of combat differ. Aren't you simply resolving 2 rounds of combat at once? And would there be results were both the attack and defender died in the same round.

Wouldn't that send the realists through the roof?

[Venger]

Originally posted by BigNick


I'm fairly sure it's the latter. Each 'round', one (and only one)point of damage will be done. In your scenario, this point has an 80% chance of being assigned to the defender and a 20% chance of being assigned to the attacker.

I think this was the case in Civ2 and SMAC, too.

You know though, I swear I've seen "rounds" where nobody scored a hit though - that would seem to indicate the "per side" argument.

If they both are per round, NO WONDER we see these goofy combat results. Someone in another thread mentioned there may be a way to change the number of HP per experience level - is this true, where is it, and has anyone tried it? This is a godsend if so and could allow those of us who want a sensible combat model to make changes and allow those who wants to sing Kumbaya in Combat Fantasyland to keep the system as is...

Anyone know about this?

Venger

[Venger]

Originally posted by WhiteElephants
I'm not sure how the "per side" and "per round" styles of combat differ. Aren't you simply resolving 2 rounds of combat at once? And would there be results were both the attack and defender died in the same round.

Good question - no, it's not resolving 2 combat rounds at once. Basically, it treats rounds of combat like - rounds of combat, with a chance of success for EACH side, rather than just only one side scoring a hit per round.

Yes - it would results in situations where both died in the same round, except that could simply be in the game mechanics - the defender rolls first in every round. That way you avoid it. I WANT to say I've seen units annihilate each other in Civ2 rarely - but cannot be sure. Maybe I'm thinking of Archon on the C64...heh...

Wouldn't that send the realists through the roof?

I have no more roof left...

Venger

[WhiteElephants]

Originally posted by Venger
Good question - no, it's not resolving 2 combat rounds at once. Basically, it treats rounds of combat like - rounds of combat, with a chance of success for EACH side, rather than just only one side scoring a hit per round.

Yes - it would results in situations where both died in the same round, except that could simply be in the game mechanics - the defender rolls first in every round. That way you avoid it.

Hmmm... not sure I follow the differnce.

[reds4ever]

i agree

[Excelsior]

Originally posted by Venger
Now, does anyone KNOW, for sure, that the system in Civ3 is a per round system? Also, can anyone confirm that which I am fairly sure of, that Civ2 was per side?
Venger

I'm pretty sure that Civ2 used a per round system. The reason that a 4 attack unit versus a 1 defense unit is more likely to lose in Civ3 is because units only have 3-5 hitpoints versus a minimum of 10 in Civ2. Assuming 10 hitpoints each, the odds of a 1 defense unit defeating a 4 attack unit in Civ2 was roughly 1.5%. With 3 hitpoints in Civ3 the probability grows to 5.7%.

I just tested this in Civ2 with legions vs. warriors and I found that not once was the attacking legion defeated. Never did a legion lose more than 50% of its hitpoints. I then edited the rules such that both units had firepower of 3, essentially reducing the hitpoints to 3.33. Thereafter, the legions had a much more difficult go of it. Out of ten subsequent combats, 3 were unharmed, 3 lost 3 hitpoints, 2 lost 6 hitpoints, 1 lost 9 hitpoints, and 1 was destroyed.

[Evil_Eric_4]

Hi Venger.

Ive been wanting to tell you ever since i started reading these forums 2 weeks ago that you are one funny mofo.
always amused by the things you say

(sounds like ive been playing civ3 too much as Im obviously trolling for allies

)

[BigNick]

Originally posted by Venger


You know though, I swear I've seen "rounds" where nobody scored a hit though - that would seem to indicate the "per side" argument.

It would, I've just never seen that myself.

If they both are per round, NO WONDER we see these goofy combat results.

I think it's just a matter of simplicity. Not from a programming perspective, but I don't mind letting it be fairly easy to calculate odds. I'm not sure the per-round system is neccessarily inferior to the RPG-style system you'd originally assumed - it's just a matter of game balance. Under your system units are far less likely to get out of a battle uninjured than in the per-round system, which would probably mean that a powerful (more modern) attacking force might find it a bit tougher to sweep through a primitive empire in a limited period.

[Excelsior]

Originally posted by Venger
Someone in another thread mentioned there may be a way to change the number of HP per experience level - is this true, where is it, and has anyone tried it? This is a godsend if so and could allow those of us who want a sensible combat model to make changes and allow those who wants to sing Kumbaya in Combat Fantasyland to keep the system as is...

Anyone know about this?
Venger

Well, I mentioned it to you in "Is Civ3 even worth it?" a couple of hours ago. It is in the editor under the tab "Combat Experience."

[Green Giant]

Gee, wouldn't it be nice if someone at Firaxis could clear this up for us?

[smellymummy]

Originally posted by Venger
You know though, I swear I've seen "rounds" where nobody scored a hit though - that would seem to indicate the "per side" argument.

not so. every round someone gets hit. it might seem like no one scores a hit, but i'm pretty sure that's just our eyes lying to us.

Originally posted by Venger
Someone in another thread mentioned there may be a way to change the number of HP per experience level - is this true, where is it, and has anyone tried it?

in the civ3 editor - the menu --> Tools --> uncheck "use default rules". click OK on the pop up.

Then, again the menu, go to Rules --> Edit rules (or CTRL R).

Then in the rules editor, go to the "Combat Experience" tab.

You can then rename each experience level and adjust the amount of HP accordingly. In essence you can give "conscripts" the name of "draftees" or maybe "suckers" however you can assign them an HP of 10. Maybe even 50, who knows..... I haven't tried it

[Excelsior]

Originally posted by smellymummy
You can then rename each experience level and adjust the amount of HP accordingly. In essence you can give "conscripts" the name of "draftees" or maybe "suckers" however you can assign them an HP of 10. Maybe even 50, who knows..... I haven't tried it

It only goes up to 20.

[Altuar]

Originally posted by Venger
You know though, I swear I've seen "rounds" where nobody scored a hit though - that would seem to indicate the "per side" argument.
Venger

Venger, I cant prove this but statistically your per side argument does not hold true...

If that were the case, early battles between warriors would have a 50% chance of missing a hit every round. This is not the case. I have seen the "misses" as they are, but they occur very occasionally, and I think this occurs due to insychronization between unit animation/sound effects and the actual timing of the rounds.

[reds4ever]

it's all down to probality math, i believe!

[Your.Master]

Originally posted by WhiteElephants


Hmmm... not sure I follow the differnce.

Per Round:

Round 1:

80% chance 4/2 guy (hereafter referred to as "George") hits mr. 1/1 (hereafter referred to as "Crazy Old Bob"...too long, make that "Bob").

if George doesn't hit Bob, Bob hits George.

Result: 1 point of Damage inflicted on one target.

Inevitably this battle, assuming regular units, will go on for less than 5 rounds (3-5, to be exact).

----------------------------------------------------------------------------------

Per side:

80% chance George hits Bob.

an INDEPENDENT 20% chance Bob hits George.

Result: 0-2 points of Damage inflicted on neither, one, or both targets.

Theoretically, this battle might not ever resolve, but the chances of that are 1/infinity, and thus 0%, and therefore that will not happen (Theoretically...calculus is SCREWY!). However, this battle could end in three rounds with 5 points of damage inflicted (assuming the game disallows "ties", otherwise, 6 damage points of mutual obliteration could be achieved).

[Venger]

YM has reiterated the difference quite well...

This does change my numbers a bit - not for the better mind you. I had assumed it was per side, not per round. That makes the chances for goofy results even higher than I had predicted...

I am modifying those HP's like a mofo in just a minute!!

Venger

[Venger]

One question -

If we change the HP values, will that screw up the armies?

Reason is, the graphics seem to be such that 20HP is the maximum you can get (an army of four elites would be 20 HP), represented by the graduated HP bar.

If we change the HP, that means that the army could have up to, say, 80 HP. Would that bomb out the game?

Venger

[WhiteElephants]

Originally posted by Your.Master


Per Round:

Round 1:

80% chance 4/2 guy (hereafter referred to as "George") hits mr. 1/1 (hereafter referred to as "Crazy Old Bob"...too long, make that "Bob").

if George doesn't hit Bob, Bob hits George.

Result: 1 point of Damage inflicted on one target.

Inevitably this battle, assuming regular units, will go on for less than 5 rounds (3-5, to be exact).

----------------------------------------------------------------------------------

Per side:

80% chance George hits Bob.

an INDEPENDENT 20% chance Bob hits George.

Result: 0-2 points of Damage inflicted on neither, one, or both targets.

Theoretically, this battle might not ever resolve, but the chances of that are 1/infinity, and thus 0%, and therefore that will not happen (Theoretically...calculus is SCREWY!). However, this battle could end in three rounds with 5 points of damage inflicted (assuming the game disallows "ties", otherwise, 6 damage points of mutual obliteration could be achieved).

And that's different because...

[Venger]

Just think about it. In order for a unit to "win" it has to hit without being hit.

That's easy for the 4 unit, and hard for the 1 unit.

The 4 unit has a 64% (80% chance to hit * 80% change of othe runit missing) chance of hitting without being hit.

The 1 unit has a 4% (20% chance to hit * 20% chance of other unit missing) chance of hitting wihtout being hit.

The chance of the 1 unit hitting without being hit in a per round instance is 20%.

See?

Venger

[WhiteElephants]

Originally posted by Venger
Just think about it. In order for a unit to "win" it has to hit without being hit.

That's easy for the 4 unit, and hard for the 1 unit.

The 4 unit has a 64% (80% chance to hit * 80% change of othe runit missing) chance of hitting without being hit.

The 1 unit has a 4% (20% chance to hit * 20% chance of other unit missing) chance of hitting wihtout being hit.

The chance of the 1 unit hitting without being hit in a per round instance is 20%.

See?

Venger

Right, but your math is exactly the same if you carry it out over two rounds of a "rounds" system, no?

[Excelsior]

Originally posted by Venger
One question -

If we change the HP values, will that screw up the armies?

Reason is, the graphics seem to be such that 20HP is the maximum you can get (an army of four elites would be 20 HP), represented by the graduated HP bar.

If we change the HP, that means that the army could have up to, say, 80 HP. Would that bomb out the game?
Venger

Unfortunately, I think it does. I haven't tested it directly (IMO leaders are too rare to waste on armies) but I have seen AI armies composed of one unit.

[Venger]

Originally posted by WhiteElephants


Right, but your math is exactly the same if you carry it out over two rounds of a "rounds" system, no?

No.

Again, in order to "win" you have to cause more damage than you get. That's doesn't happen very often for the 1 defender. 4%. Chances of getting that twice are REAL low. Even if it lucked out and got it once, the chances are that the exact opposite would happen next round.

Venger

[Marquis de Sodaq]

CivII uses a "per round" method, according to how you define it. The odds of a combat round are calculated based on the attacker's modified attack value vs. the defender's modified defense value. Somebody scores a hit each round - if not the attacker, then the defender by default.

The result is smoothed out by the use of hit points being 10x the number listed in rules.txt. That is, a legion has 10hp, a musketeer 20hp. The warrior in your scenario would have to win 10 times in a row against a legion, not just 1 or 2.

[GePap]

Can the 'rate fo fire' or arty and bombers and so forth be changed? If we get to units with 10 hp but arty only does 2 damage still, then arty or bombers become much weaker. Also, if the 'rate of fire' is increased, does this mean that the chances of hitting improvements would rise?

[GePap]

the explination over changing Hp seems to rely on experience levels, not tech levels. Let say that you say regulars get 4, veteran 6, elites 10 . Well, if warrior becomes elite, does he still not reach 10? Is there a way to limit the ability of units to reach certain experience levels, which according to how the rules edditing was described here, is what determines hitpoints? Also, is there any way of insuring that upgraded units don't lose HP?

[Rusty Nail]

Marqis de S is correct. Both civ2 and civ3 use a per round model where one side has to win a hit. See page 100 of the civ2 manual. Defensive strength represents the damage inflicted on the attacker i.e. with the defender's probabily of winning the round and with the defenders damage defined by his firepower . The major difference betwen civ2 and civ3 is the NUMBER of rounds (greater in civ2) and the concept of firepower, which is always 1 in civ3. This is the model used in the civ3 "civulator". Having more rounds skews the result in favour of the one having a better chance of winning a single round and reduces the standard deviation of the result, i.e fewer outliers. It is a metter of subjective taste as to what level of certaintly you feel is best, i.e chance for the underdog to win. He has a better chance in civ3 than in civ2. But, in a mod, by imodifying the number of hitpoints on both sides we can tailormake this to any standard. This multiplier could even be a parameter that the user could choose.

[Arkatreides]

Originally posted by GePap
Can the 'rate fo fire' or arty and bombers and so forth be changed? If we get to units with 10 hp but arty only does 2 damage still, then arty or bombers become much weaker. Also, if the 'rate of fire' is increased, does this mean that the chances of hitting improvements would rise?

Yes it can. I multiplied all the hitpoints by 2 and then realised that bombards were useless. So I multiplied bomard rate of fire by 2 as well and now it evens out. Basically the bomabrd rate of fire is exactly what the firepower in Civ2 was. Un fortunately it only works for bombard units, not regular units.

Armies are not screwed up if you go over the 20HP limit. In fact for a while I had elite units with 20HP and I could easily make a 60HP army. It just doesn't show all the HP dots on the screen, but who cares. The problem with 20HP units (a factor of 4 vs the original setup) is that you cannot multiply bomabrd by 4 as the maximum it allows is 10 and some units like bombers have an original value of 3.

[J10]

Personally, I do not like underdogs to win, I do NOT.

Reason: Battle tactics are made less useful than it was in Civ2 if you screw it up so that underdogs can win.

I mean com'on, yesterday I lost two 5hp Elite Calvaries attacking one 1hp Calvary on grassland in a row, and finally had to kill that 1hp Calvary with a 4hp Calvary which wounded up having 1 hp left. Someone who's still in college doing probability can tell us the chance of that in Civ2 and in Civ3 so we can all be amazed at how sh|t out of luck I was. I'm pretty sure it's a lot more difficult to happen in Civ2 then in Civ3.

And once again, if underdogs are made easlier to win, then battle tactics are not as important anymore, bad news for us warlike leaders