combat without artillery

[MSGT John Drew]

At ease!

Gentlemen, I need feedback on how many men are needed to overwhelm a city's defenses without the help of my men. Personally, if I were the Field Infantry Commander, I would not smash my men to their deaths at the city gates but that is not my call. I would rather blast anything and everything that stood with artillery before ground forces occupy the rubble that remains but again I am in no position to do that. All I hope are insights on offensive versus defensive army ratios. All information hereby received will be forwarded to Army Headquarters. Thank you.

Carry on.

Attention!
*salutes then leaves*

 

[Zouave]

After I modded Firaxis' silly unit values, I was left with Infantry at 10.13.1. (I concede there may be argument for increasing defense to 14). Regardless, if attacking fortified enemy Infantry on a hill or in forests, I figure even three Infantry will be hard-pressed indeed to take the position without Artillery support. And that is how it should be.

Too bad the AI can't figure out how to use bombardment units offensively.

Moderator Action: The poster was not asking for your opinons on artillery, or how silly firaxis' unit values were. Seeing as this has been an ongoing problem. You are the recepient of a 1 month ban from posting at CFC.

- Civ3 Mods

Gonzo
Cunobelin of Hippo
Please read the forum rules: http://forums.civfanatics.com/showthread.php?t=422889

 

[Sirp]

Drew: it largely depends on how good your attackers are compared to their defenders.

Generally attackers range from having a 2:1 A ratio, to having a 1:1 A ratio. Anything above or below this is likely going to be a slaughter.

If you have a 2:1 ratio, you probably need slightly more than 1 soldier for every soldier in the city. If you have a 1:1 ratio, I would want 2-3 times as many troops as troops in the city.

Of course, if they are on hills etc, it makes a big difference, as do city sizes.

 

[Vorlin]

IMO, if you are going to attack in the modern age without any preparatory bombardment then it's best (if possible) to have 3-4 armies to lead off with. A lot of people don't like putting modern units into armies (and they have a lot of good reasons for not doing so), but if you can't soften up enemy cities first then have a couple of armies to take down those first few defenders is a -huge- help.

However, if you do use this method and only have 3-4 armies then your advance will be slow as you'll have to wait 2-3 turns after taking each city so the armies can heal up. Or you can split the difference and divide your invasion force into multiple attack groups (2 groups with 2 armies each, or 4 groups with 1 army each), you'll take more casualties but your offensive will progress faster.

Obviously the Military Academy and Pentagon small wonders are important if you use this tactic heavily (the Academy is pretty much mandatory).

 

[Sanaz]

Just an additional point, which may be obvious but wasn't to me... When you attack with the regular units (no army, no artillery), you can end up doing one point of damage, lose your unit, then their unit promotes and regains that point. If you then don't get the city first attempt, the defense is likely in barracks and will heal 100% by the next turn. So, you may waste all of your units and still not reach your goal. The worst is if you also give the enemy a great leader. So, unless you're ahead in technology (i.e. Tanks against Riflemen), no percentage of attacking units may be enough (especially if the enemy is over a river, on a hill, and is larger than size 7) without softening them up first. You can break your military, and then when you rebuild you're up against a bunch of Elites.

 

[Lt. 'Killer' M.]

very good point, Sanaz.

I usually bring a stack of at least 8 Horsemen against a city with two Spearmen and 1 archer......
same thing for hugher levels, just that Musketmen may require a 4,5:1 ratio of Knights. This pretty much guarantees you a win.

 

[cracker]

MSGT,

You have got to get down to understanding some more specifics or you absolutely will get garbage answers. Generic garbage questions get you garbage advice and Zouave rants as your prizes.

The key to understanding your question is to look at the defenders that may be in the town or city and learn the skill to figure out what the defensive strength will be.

The strongest defender will always be the one that is displayed and then they will stack down from there. If a 3/3 spearman is showing then you know that there cannor be a 4/4 swordsman or a 3/3 pikeman in the town. Get good at this assessment or die.

Then learn to assess the terrain. You can click on enemy toens to display their terrain info and see if they are on hills or just on flat ground. Hills give a 50% defender bonus while ALL flat ground gives a 10% defender bonus. Also look closely at the edges around the town to see if there is river between the two squares. Rivers give a 25% bonus to the defender when the attacker is crossing the river while attacking.

Most defenders in towns will be fortified in place. This will make them look meaner and give then the little white halo around their health bar. A fortified defender gets a 25% defender bonus.

Towns of less than 7 population do not get an extra urban area defens bonus unless they have walls. Cities of 7 and above give a 50% defender bonus and metropolii of 13 or more pop points give a 100% bonus.

What all this means is that the same units can be stronger and may take more units to kill him depending on how calculating or how incompetent of a general you might be.

A fortified pikeman defender in a town of 3 people on the plains will defend at 4.125 {3.0 * 1.25 (fortified) * 1.1 (plains) }. If you can catch that same defend in the middle of moving across open terrain he will defend at 3.3 {3.0 * 1.1 (plains) }.

A fortified pikeman on a mountain will defend at 6.00 {3.0 * 2.00 (mtns) }.

A fortified pikeman defender in a city of 7 people located on a hill behind a river will defend at 10.55 (3.0 * 1.25 (fortified) * 1.50 (hills) * 1.25 (river) * 1.50 (city bonus) }. If you attack this eact same defender from a different direction where the river is not in play, then the defenders strength drops to 8.44.

Your success in the attacks is driven by a sequnce of random numbers that get compared to the attack and defense values that YOU choose by how you process the attack.

Most people that rant and scream about the combat values and the combat system are absolutely clueless about how these statistics work and are just looking for something to scream about to hide their lack of understanding.

When you attack a unit, then the combat outcomes are generated by using the ratio of the strength of your attacker versus the total found by adding the strength of your attacker to the effective strength of the defender.

When you attack a pikeman with a swordsman the chances of winning are never as simple or as favorable as just dividing 3 bu (3 + 3) to get 50%. You never have a 50/50 chance of winning. Using the pikemen examples shown above, you can attack with a swordsman and have anywhere from a 47.6% chance of winning each stroke all the way down to a 22.1% chance of winning each stroke.

If you are stupid about attacking under the wrong circumstances, it may actually take more units to kill a regular or conscript unit because of the fact that defenders will almost always get promoted each time they win two battles in the same turn.

Using our pikeman/swordsman example again, if you attacked a regular 3/3 pikeman (fortified+city+hills+river) with a stack of regular swordsman, then you would need 4 hit points of swordsmen to plan on killing each one hit point worth of pikeman. If you are not going to kill the pikeman right away (which you won't under these conditions) then plan on it getting promoted and gaining one hit point for every two times it kills one of your brave but stupid swordsmen.

So, to kill the 3/3 pikeman fortified in a city, on hills, behind a river you would usually need 4 swordsmen if no promotions occured. Because this is a very weak attack against a strong defense, the pikeman will get promoted two times, so you will need 20 hit points worth of 3/3 swordsmen to win the battle 50% of the time. This would translate into at least 7 swordsmen of which 6 would die due to stupid generalship.

You can apply this same logic to situations with mixes unit types and stronger/weaker combinations of units.

There is no fixed exact number ofunits required to win a battle. It all really depends on what units are involved on what terrain. Thats what makes the game exciting. Some lesser players would have you modify the game to to take away some of these variables but most of these comments are based on a basic inability to play the game as it is designed or to understand and apply the basic rules of combat strengths as modified for varioes terrain types and defensive conditions.

Using an army or swordsman will only delay the chance for the defender to win an individual comabt event and thus delay any chance of promotion. Armies do not make your units stronger or increase their chances of winning ANY of the individual strokes in the battle.

Ask more specific questions if you want more specifc answers and fewer Zouavesque rants or just plain crying and drivel.

Good luck,

 

[Vorlin]

Cracker, aren't all bonuses added up first and then applied, instead of being applied in succession? For instance, in your example of 'fortified pikeman defender in a city of 7 people located on a hill behind a river ' you came up with a defense strength of 10.55, but if the bonuses are all added first and then applied the defensive strength is 7.5 I believe all bonuses are added first, at least that's been the pattern of all Firaxis/Microprose games I've ever played. I do know that factory and powerplant bonuses are added first and then applied, and not applied in succession, but I haven't checked the other areas of Civ3 where this comes up (like marketplaces and banks).

It would be nice if Civ3 had the same info box that SMAC did, where you could see the actual odds the computer uses when two units fight. I really liked that feature, as often I miss certain modifiers when trying to figure it out on my own (as you pointed out in your post, there are a lot of factors to consider).

 

[cracker]

Vorlin,

I cannot say for certainm but I have been tracking some actual combat events in stroke-by-stoke detail and so far the results are matching with the cummulate interpretation instead of the "add-em-up-first" intepretation.

I you say that previous you have proof that prior Firaxis coding efforts have used the "add-em-up-first" method that is a fairly strong argument in favor of that method being repeated again for CIV3.

The key point in the discussion is that asking any questions about combat success rates without understanding the basics of the combat defense bonuses first, will lead you to the wrong answer.

At least when using one of the interpretations of the combat defense bonuses, a player can begin to make some reasonably intelligent generalship type decisions when conducting combat operations.

If we conducted a scientific poll to see how many players regularly conduct attacks by horsemen against spearmen or swordsmen on hills, when these spearmen and swordsmen are not endangering a road, city or resource, then I think you would be shocked to find that about 80% of the target market for the civ3 game makes mistakes of this type on a regular basis. Once you understand the terrain bonus issues, the right move tactically is just to make sure you have units that can strike the enemy units when they move off their defensive positions. You can often bypass the units and then just pick them off one-by-one with mobile units when the terrain favors you.

 

[Lt. 'Killer' M.]

Vorlin: cracker is dead on - I started using defence bonus extensively, also started to assume a doubled defence for any terrain, trippled for mountains, then decided on whether to attack or not - and to my utter surprise I keep steamrolling the AI since then.

 

[Lt. 'Killer' M.]

ah, one thing I should add: I see a surprising number of combats (even with HP doubled) where a 2 attacker defeats a 2 + fortified + town + terrain defender. Usually, an AI attacker defeats my defender.

But after i started taking notes on these results, I found them to be even - it is just that the AI tends to bring defenders along in huge numbers on higher levels, and thus doesn't loose so many units to counterattacks - and thus often these surprise results lead to the falling of towns, while the other way round nothing exceptional seems to happen...

 

[Vorlin]

At this point the best solution would be for someone who can get a Firaxis coder's attention to have them get this info. That would be the only way to be definitive about it.

As far as needing so many extra troops than appears necessary at first, I think a big part of this is the promotion effect that Cracker referenced. Nothing quite like having your 'army of conquest' turn into a promotion tool for the AI's defensive units. Once you see that happen you don't tend to forget it, pain is a good teacher.

 

[ptorek]

After I modded Firaxis' silly unit values, I was left with Infantry at 10.13.1. (I concede there may be argument for increasing defense to 14). Regardless, if attacking fortified enemy Infantry on a hill or in forests, I figure even three Infantry will be hard-pressed indeed to take the position without Artillery support. And that is how it should be.

Moderator Action: The poster was not asking for your opinons on artillery, or how silly firaxis' unit values were. Seeing as this has been an ongoing problem. You are the recepient of a 1 month ban from posting at CFC.

Gack! In a list of off-topic posts, from most off-topic to least, Zouave's would be number 45,379. The only possible REAL reason for banning the guy is an antipathy for his viewpoint.

Go ahead, ban me too. I dare you.

Moderator Action: Go ahead and taunt me again, then see if you can post. And please read my post below, mod whining isn't allowed outside of site feedback.

Gonzo
Please read the forum rules: http://forums.civfanatics.com/showthread.php?t=422889

 

[GI Josh]

http://www.columbia.edu/~sdc2002/civulator.html

This will answer most of your combat questions - kudos to whomever designed the site. The only thing missing is UUs.

Cracker, you're advice is spot-on, but I think you might be off on how defensive bonuses are calculated. It is my understanding that they are *not* cumulative, but are all calculated off of the base. For example, a unit with defense 4 is fortified on a hill. Fortification is +25% and hill is +50%. On a cumulative basis, the base of 4 is +25%, which is 5, to which is then added the hill bonus to reach def. value of 7.5. I believe the correct calculation is always off the base number and not cumulative, e.g. base 4, add 25% of 4 for fortified (+1) and add hill bonus to base value of 4 (+2), to get a defense value of 7, instead of your 7.5. This result is not great in this example, but add in city bonuses and the results can be significant.

The underlying lesson - on the modern battlefield, artillery is king.

 

[Tassadar]

Originally posted by ptorek




Gack! In a list of off-topic posts, from most off-topic to least, Zouave's would be number 45,379. The only possible REAL reason for banning the guy is an antipathy for his viewpoint.

Go ahead, ban me too. I dare you.

I agree

 

[gonzo_for_civ]

Moderator Action: Tassadar/ptorek: Mod complaints go in the site feedback forum. And in a list of off-topic posts, from most off-topic to least, yours ranks pretty damn high. Stay on topic.
Please read the forum rules: http://forums.civfanatics.com/showthread.php?t=422889

 

[alpha wolf 64]

personally i rarely use artillery because it slows me down. i like massed stacks of the best mobile unit i have. i may take some losses but its worth it. i follow up with defensive units to protect my new slave workers and to protect any wounded units. the workers keep the road/RR complete to the front lines so i can rush more units to the front line.

 

[ukrneal]

A couple of extra omments:

1. Doesn't the hp level of your attacking units play a significant role? In other words, if some of my units are injured, and I want to use them for attack, won't I need even more units as my odds are not as good as when they are fully healed?

2. I think one of the combat problems is the deception in absolute numbers, which is especially true with the early units. When I have an archer agaisnt a warrier, I always think the odds should be 2:1, but in reality, it will always be less. So you really DO need to think in %s as people here have so well explained.

3. The level of the unit (elite, veteran, etc.) also palys a big role. That calculator will give you a good sense to how it works.

4. To answer your question (a little bit), it is hard to attack with less than 3-4 units with any surety. Best of all is to have a mix of units, especially a defender to protect your weak defending attack units. If the computer is hunkered down in a big city with 4-5 defeneders, you'll probably need upwards of 8 units or more if the bonuses are significant (and you are attacking with anything less than veteran units). A 2:1 ratio (attacking to defending) has worked well with me (assuming relative level of unit). Better to retreat and regroup if you see you don't have enough attack units.

 

[TheNiceOne]

As a general statement I would say that if you're not using artillery at all, then you will either lose most of the time, or you're playing on a too easy setting.

Taking out a city against a defender that is on-par in techs is just too costly after riflemen has arrived. As cracker excellently explained, you will just lose too many units without doing much harm. This is true for cavalry against riflemen or (gasp) infantry and mostly true for tanks vs. infantry.

As long as you can attack with tanks against riflemen and cavalry, then you can do well without artillery, but then you're not fighting real opposition.

Btw, I was surprised that Zoave was banned a full month. IMHO, Zouave has been the most regular spammer on this forum. Whatever topic he joined, he would usually do as above: not discuss the actual topic, but use the topic as a reason to rant about how silly and stupid CIV3 (and Firaxis) is. I think a month is very long, but think he diserved a day or three (as first warning).

Edit: Although the zouave comments aren't on-topic, the rest of your post was and it wasn't really a rant so I'll address your comments on zouave just to clear something up. This was not a first warning, this wasn't even a first banning. It was discussed to great lengths in the staff forum before the actual 1 month was instated. I talked to both Hippo(before the ban) and thunderfall(after the ban) on the same day as the ban. I would prefer it if comments about the ban could either be sent to me, thunderfall, hippo, or posted in the staff forum from now on. Thanks, Gonzo

 

[Cerryl]

Hip Hip, Hurray. Someone finally did at least a little something to shut that really cool guy up, if but for amonth.

Moderator Action: Now come on, no flaming, especially if the person you are flaming is unable to defend themselves. Gonzo
Please read the forum rules: http://forums.civfanatics.com/showthread.php?t=422889

Originally posted by Zouave
After I modded Firaxis' silly unit values, I was left with Infantry at 10.13.1. (I concede there may be argument for increasing defense to 14). Regardless, if attacking fortified enemy Infantry on a hill or in forests, I figure even three Infantry will be hard-pressed indeed to take the position without Artillery support. And that is how it should be.

Too bad the AI can't figure out how to use bombardment units offensively.

Moderator Action: The poster was not asking for your opinons on artillery, or how silly firaxis' unit values were. Seeing as this has been an ongoing problem. You are the recepient of a 1 month ban from posting at CFC.

- Civ3 Mods

Gonzo
Cunobelin of Hippo
Please read the forum rules: http://forums.civfanatics.com/showthread.php?t=422889