air intercepters question

[Kerrang]

Just did some testing on air intercepts, the methodology and results follow:

I set up the first round of tests by starting on a duel map with one AI, and immediately placing my first city, declaring war on the AI, and entering the World Builder. In the world builder I placed an AI city 3 tiles away from my city with a Castle and Walls. I created 10 fighters for my city, and 1 SAM Infantry for the AI city. I bombeb the AI city with all of my fighters in one turn, recording how many were intercepted, then ended the turn and entered the World Builder again. I deleted both cities and their units, then started over with 10 fighters in my city and incremented the AI SAM Infantry by one after each test, until there were 10 AI SAM units, here are the results:

1 SAM -> 1 Intercept
2 SAM -> 2 Intercepts
3 SAM -> 3 Intercepts
4 SAM -> 4 Intercepts
5 SAM -> 4 Intercepts
6 SAM -> 2 Intercepts
7 SAM -> 5 Intercepts
8 SAM -> 7 Intercepts
9 SAM -> 5 Intercepts
10 SAM -> 4 Intercepts

I planned to do the same test with Fighters, then Jet fighters, then few tests with multiple unit types, but there were some factors that I did not account for. One issue was that I had trouble guaranteeing that the AI would set all of their units on intercept, and unfortunately the World Builder is not accesible in multiplayer (even hotseat).

I ended up just testing this a couple of times with 10 of my fighters against a few AI fighters. I did the same initial setup, but used fighters in the enemy city rather than SAM Infantry. The first round of attacks I was not intercepted, but I also noticed that the AI fighter did not appear to be in intercept mode, so I ended the turn, and the fighter started circling the city on the next turn. I attacked with all of my fighters again, and was intercepted once, but noticed that immediately after the intercept, the fighter no longer circled the city. I ended the turn and added another fighter to the enemy city and once again they did not immediately go into intercept mode. I ended the turn again, and then saw both fighters circling the city, so I tried another round, and was intercepted twice. After each intercept, one less fighter was circling the city.

Given the results, we can make some safe assumptions:

1) Intercepting units can only intercept one unit per turn, be they land or air units.

2) Additional units do not increase the chance for a single unit to intercept.

3) I have no life.

 

[Roland Johansen]

A ship is a mobile artillery platform with very sophisticated gun direction. One WW2 destroyer was equivelent to the artillery park accompanying an army division. A battleship was equal to that of several armies, or nations when comparing some of the smaller country's artillery parks, and it had much better gun direction. And this while the ships were moving. When stationary - as in port - they are even more accurate. There is no reason historically why a ship can not fire on land troops as accurately while it is in port as it can while out at sea. They really should be able to fire more accurately since ranging and spotting would be much better on their own territory which would be known and closely observed. Many a German got killed by the fire from Soviet ships in port while sieging Leningrad during WW2. There should be no lessening of offensive power of ships while in port and they should be able to fire on troops attacking a city they are in to better effect than if they were sailing off the coast. This goes for fighting air attacks as the aa fire would also be equally deadly. The ships at Pearl Harbor put up quite a good aa defence once they recovered from the initial shock considering how backward their aa equipment was. Almost 10% of the attacking Japanese got shot down in that attack and at least half were from aa, I believe. Obviously, ships firing torpedoes would be an exception to this and virtually useless. The only other exception I can think of is when the ships are shelled from a place they cant reply to such as the siege of Port Arthur in 1905 and that involved other artillery, not troops in a city.

This is one of the major combat failings in Civ2 and on up.

Where the ships are at a disadvantage is in defence. Cant move about, they are easier to hit. This is where the factor should be lowered. Not sure how this could be worked into Civ4 with its already retro single factor for defence and offence. Perhaps through promotions?

When a city is taken, I think ships and aircraft should have a percentage chance of being captured, destroyed or of escaping at the last moment.

You present it as if we disagree, but I kind of agree with most of your statements (especially the statements about the firepower of ships, people tend to think that bombers are more capable at bombarding, but the firepower of even one battleship is almost unequaled in the WWII era). If ships could bombard land units in civ4, then I also think they should be able to do it while in harbor. This would represent the ship firing at units that are sieging the city (like Leningrad). However, I view units that attack a city as representing city fights and such. That's why certain units in civ4 can get the city raider promotion. These are typically units that could excel at city fighting (although the tank shouldn't get this promotion, infantry should get it). And in the case of city fighting, I can't see what a battleship in port could do to defend the city. Yes it could obliterate a division of the attacking army, but it would take a part of the city with it. I think the most logical result in this case would be for the battleship to leave the city to an adjacent tile. The battleship would need to be taken by surprise attack to be captured and I can't see this surprise attack happening in a city siege. Airplanes and artillery should be able to sink ships in port and should do extra damage as the ships are immobile while in port. But the AA of the ships should of course be effective.

It could be argued that as civilization isn't a wargame that the representation of war in the game is limited. However, we do have units retreating in this game, so there's no reason this couldn't apply to ships in port when a city is captured. There are also instances where bombers do varying amounts of damage (bomb shelters, defence bonuses), so why not a damage increase of bombers and artillery against ships in port?

 

[Roland Johansen]

A ship could be captured simply cause the crew surrenders after the city is taken, there's no fuel or the "out" channel is blocked and they have no other choice other than die fighting. The other two options, destroyed or escaping also make sense historically and in making the game more enjoyable to play.

I didn't think of the case where ships have no fuel. I that case, surrender or blowing up the ship (self destruction) seem like the only options.

Probably, we disagree in the case of civ3 bombardment. I don't want to go back to that kind of bombardment where with enough bombardment units one can totally obliterate a stack of units without any danger to your own units. I can see that some people dislike the suicide collateral damage attacks of civ4 because of realism arguments. But other realism arguments could be used to make the bombardment of civ3 ridiculous. As if a stack full of units would sit and wait for their death while countless artillery units are shelling them. They would counterattack. But the turn based nature of civ-combat makes that impossible. And that's why I don't like ranged bombardment in a turn based game.
It works for planes because there's a counter unit that can react (the fighter). But ranged bombardment without any counter is naturally overpowered in turn based combat.

What I would like to see, is some form of stack fighting. The system should work in such a way, that if you attack with a well balanced stack, then it will perform the best. An artillery heavy stack would have a nice strong hit in the initial attack, but the meager infantry defence in this stack would be insufficient to hold back the counterattacks of the tanks in the stack that is being attacked. So the tanks break through the infantry defenses and massacre the artillery. The stacks should be limited in size so that combat is more interesting than 'the biggest stack wins'. However more stacks could be present in a single tile. The intelligent player could then balance its stacks in such a way that they perform well against the stacks of the opponent. It has been done in previous games and it can work. I don't think that it is more difficult to program an AI that can handle this kind of combat than one that can handle the civ4 combat with all its promotions and its rock-paper-scissors system.

 

[Bushface]

@Kerrang - so you had 37 interceptions in all during 100 attacks, which makes precisely 37% chance of interception. This, I think, is close enough to the 40% interception chance that a SAM has to be good corroboration. I had 17 fighter interceptions among the 56 attacks made against fighter-defended cities, a 30% rate, and 7 SAM intercepts scattered about the 15 attacked cities (I say again that one solitary SAM in a city with no possible air cover made two, successive, interceptions in the same turn). It isn't clear if the fighters always tried to intercept before the SAMs had their chance, though this seems very likely because of the greater range of the fighters, but on that basis the SAMs were left to deal with 48 attacks and got a 15% intercept rate. This is worse than the expected 20% (SAM basis 40%, halved for Stealth) whereas the fighters did better. I still think that my overall 55% success rate is a lot lower than ought to be expected of Stealth Bombers.
As an aside, this thread seems to be unravelling into two quite differen discussions - (1) air interceptions, and (2) ships in harbour. Odd.

 

[Roland Johansen]

Sorry about that.

Me, too. I think that discussion has ended though, since meisen and I agree too much.

 

[Snerk]

Given the results, we can make some safe assumptions:

1) Intercepting units can only intercept one unit per turn, be they land or air units.

2) Additional units do not increase the chance for a single unit to intercept.

Yes that would be a safe assumption indeed, and answers some of my questions.
Thank you!
..however, im still much wondering about the result of multiple/different unit intercepting air-attacks.
Like a fighter a SAM and a Mech. Inf. all intercepting.

 

[Bushface]

@Meisen - Whilst there were, as you say, 55 SAMs who managed 37 interceptions, an apparent 67% interception rate, it appears to me that if a SAM fires and misses (a thing never reported) it can cheerfully have another go at the next attacker. This would indicate 100 attempted interceptions, of which 37% caused hits. For example, I made three successive attacks on a city which had one SAM, the first two being successful while my third plane was intercepted, and it is only reasonable to expect that the SAM did indeed try to hit the preceding two aircraft, achieving a 33% intercept rate - versus Stealth Bombers, may I remind you. And I have had another case of two hits from a single SAM in the same turn, and the plane I lost to a hit from a MechInf was the first pass at a city which also had a Fighter and a SAM; did the latter two miss, or did the MI get first go and strike it lucky ? The Combat Log is no help, as it ignores air attacks, and the Event Log only reports successful raids and successful interceptions.
Somebody's earlier question as to whether it is better to spread SAMs over a few adjacent tiles rather than stacking them would also need elaborate investigation, These are verily murky waters in which we flounder.
I confidently expect that my current war will soon be over. My allied enemies have six cities left, scattered across tiny islands, and their defences will soon be zeroed by warship bombardment while their garrisons get ground down to 50% strength by waves of air strikes so that my lovely City Raider 3 Modern Armour units can wipe them out. Why didn't I get some Amphibious promotions ? Good question.

 

[Kerrang]

@Kerrang - so you had 37 interceptions in all during 100 attacks, which makes precisely 37% chance of interception. This, I think, is close enough to the 40% interception chance that a SAM has to be good corroboration. I had 17 fighter interceptions among the 56 attacks made against fighter-defended cities, a 30% rate, and 7 SAM intercepts scattered about the 15 attacked cities (I say again that one solitary SAM in a city with no possible air cover made two, successive, interceptions in the same turn).

Is it possible that you were mistaken? Maybe a SAM infantry was moved in or out just before or after the attack? The reason I ask this is in all the testing I did with SAMs, I was never intercepted more times than there were SAM Infantry in the defending city. It is not proof positive, but it is a strong indication, that individual SAM Infantry can only intercept once per turn. Given what I observed with the defending fighters, I am certain that they can only intercept once per turn, and it would make sense that SAM Infantry would follow the same rules.

It isn't clear if the fighters always tried to intercept before the SAMs had their chance, though this seems very likely because of the greater range of the fighters, but on that basis the SAMs were left to deal with 48 attacks and got a 15% intercept rate.

This is something I wanted to test as well, but given the inability to control what the AI would do when it came to using air units, I could not come up with a reliable method for testing this. I think the best bet would be to set something up in multiplayer, but that would take more time to accomplish, and I would have to bribe someone in my family to assist me with the testing.

This is worse than the expected 20% (SAM basis 40%, halved for Stealth) whereas the fighters did better. I still think that my overall 55% success rate is a lot lower than ought to be expected of Stealth Bombers.

I have read in the Strategy forum, that the chance of interception is calculated first, and after it is determined that an intercepting unit would have made a hit, the SBs 50% chance of avoiding detection is factored in. This seems counterintuitive to me, and I also do not know how this would effect the intercepting unit as far as whether or not it would be able to perform another intercept that turn. From a programming standpoint (I am a programmer), it seems like would be easier to calculate the SBs evasion chance first on each intercepting unit and then, if the SB fails to evade, you would calculate whether the intercepting unit hits it.

 

[Snerk]

I would agree in the fact that it sound quite unlikely for a SAM unit (or any) to intercept more than once per turn.
Nevertheless, im finding it rather difficult to get any sort of full understanding regarding the whole matter of interception.
Even after reading about all the results posted here.
I was just about to do some testing myself when my laptop decided late-game on a huge map wasn't to it likings. I did knew there would be problems but i didn't think it would be that bad. I have now starded on a standard map wich i know will work fine, even late game will be playable.
However, regarding interception i wish i could talk to the head guy down at firaxis and find out how its all connected.
But after reading the posts here and my own game experience, i think that the biggest remaining mysteri to me is about intercepting when different kinds of intercepting units are involved in an interception.

 

[Bushface]

@Kerrang - I agree that it would be easier to factor in the evasion chance before calculating hit damage (I was a programmer once). Otherwise what happens would be a factor in the damage calculation, not the interception chance.

Any way, my war is now over and here are the final scores :-
Missions flown, 120; total interceptions 36 (which is 30%), successful 84.
Fighter interceptions, 19 (incl. 2 bombers shot down)
SAM interceptions, 14 (incl.2 bombers shot down)
Mech Inf, 2 interceps (incl. 1 bomber shot down)
Destroyer, 1 intercep (bomber shot down)
Fighters made the first interception in all but 3 cases, 1 each by SAM, MI and Destroyer.
There were 2 occasions when a single SAM made 2 interceptions in the same turn, and 1 when a lone Fighter did the same.
One bomber was lost to a fighter which apparently materialised out of nowhere. It certainly was not based in the remote island city which was under attack, beside which I had a destroyer.
Make of these figures what you will. Maybe after another thousand missions a clearer pattern would emerge, but I'm not proposing to check.

 

[Arkturus]

Hi,
I'm new in this forum and a litte bit concerned that I'm too late in this thread, as the last post is over a week old already.
I read almost all of the posts in this thread, more and more amazed that nobody complains about the issue I want to address:
Why isn't it possible in Civ 4 to gain air superiority and why does nobody care? I started a major invasion of a sophisticated Civ that has jet fighters and stealth bombers.
What I want to do is disable the air defense (defending jet fighters, not ground units) in order to allow my bombers to bomb without being intercepted. To achieve that, I send my own jet fighters on bombing runs (an interception mission is not available) to let them be engaged by enemy fighters. This works very well - except one thing: In over 90% of the cases my own fighter is heavily damaged or shot down, while the enemy fighter is returning to base, ready to intercept in the next turn.
Just to make that clear: I'm talking of an enemy that has about 5 fighters defending each city, and as we already know, the fighters of the surrounding cities intercept as well, not just the ones of the target).
This results in about 20-30 successful interceptions each turn (not counting the SAMs and Mech.Inf.) and I don't know how to build so many bombers... in the first turn they get damaged, in the second they get shot down... This way there is no chance of hurting the enemy effectively by air

Why do the attacking fighters have such a tiny chance to damage the defending fighters? And is there a way to influence these chances?

 

[Roland Johansen]

Hi,
I'm new in this forum and a litte bit concerned that I'm too late in this thread, as the last post is over a week old already.
I read almost all of the posts in this thread, more and more amazed that nobody complains about the issue I want to address:
Why isn't it possible in Civ 4 to gain air superiority and why does nobody care? I started a major invasion of a sophisticated Civ that has jet fighters and stealth bombers.
What I want to do is disable the air defense (defending jet fighters, not ground units) in order to allow my bombers to bomb without being intercepted. To achieve that, I send my own jet fighters on bombing runs (an interception mission is not available) to let them be engaged by enemy fighters. This works very well - except one thing: In over 90% of the cases my own fighter is heavily damaged or shot down, while the enemy fighter is returning to base, ready to intercept in the next turn.
Just to make that clear: I'm talking of an enemy that has about 5 fighters defending each city, and as we already know, the fighters of the surrounding cities intercept as well, not just the ones of the target).
This results in about 20-30 successful interceptions each turn (not counting the SAMs and Mech.Inf.) and I don't know how to build so many bombers... in the first turn they get damaged, in the second they get shot down... This way there is no chance of hurting the enemy effectively by air

Why do the attacking fighters have such a tiny chance to damage the defending fighters? And is there a way to influence these chances?

First things first: welcome to civfanatics!

A thread is not closed very fast in this forum. I've seen people respond to threads over a year old. Not that that was a very useful thing to do, but it's not forbidden or something like that.

I agree with you that the defender has an advantage in air vs. air combat and that that advantage is a bit too big. However, I do think that there should be an advantage in air combat for the defender, just because the defender would have the advantage of ground based radar installations and such in real life. It's just that this advantage is too big in the game.

You shouldn't lose that many bombers in bombing runs. Never, never bomb with damaged bombers. They do less bombing damage proportionate to the amount of damage they have taken and they are far easier to shoot down. I don't think the AI uses damaged aircraft in bombing runs or for air interception missions. If you only use healthy aircraft in bombing missions, then you will almost never be shot down. Actually, I think that the chances of destroying a healthy aircraft by an interception mission are too small. It's something like 1 in 10 when the aircraft technology on both sides is equal.

Another thing that you can use to your advantage is the range of interception. Aircraft can only intercept within their operational range and the AI will not use damaged aircraft for interception. (Note that diagonal tiles count as 1.5 distance. Add all horizontal, vertical and diagonal distance in tiles and round down to an integer. You can see the operational distance of your own aircraft by picking a fighter and selecting strike or bombing mission.) If you know where the healthy enemy fighters and jet fighters are stationed (spy, religious spying), then you can use this knowledge to your advantage by picking a city that is protected by as few fighters as possible.

Also, only start bombing just before an attack by ground forces. This way, you will not lose aircraft in bombing mission that have no real purpose or effect.

 

[Older than Dirt]

So, Roland, do fighters in patch 1.52 have a diagonal operational distance of 4squares? I'm at work now and can't check my game, but I seem to remember having at least a 5 range diagonally and one instance where a fighter wouldn't go the full 6, which mystified me, but which must be because it was fully diagonal.

 

[Bushface]

A diagonal move is taken as being about 1.5 times as long as an orthogonal one. So a fighter's coverage area is shaped much like the cultural radius of an expanded city: it covers 6 tiles in a column or row, from base, but only 4 tiles on a diagonal, or in combination, for example, 3 diagonal plus 2 orthogonal. Similarly for other aircraft - 7 diagonal for a jet with nominal 10 range and 8 diagonal for a stealth with range 12. Just look at the marked tiles which show up when you've selected an aircraft and you'll see the pattern.

 

[Arkturus]

I agree with you that the defender has an advantage in air vs. air combat and that that advantage is a bit too big. However, I do think that there should be an advantage in air combat for the defender, just because the defender would have the advantage of ground based radar installations and such in real life. It's just that this advantage is too big in the game.

Thanks for the warm welcome
Well, OK. Let's say the defender has the advantage of a ground based radar (and perhaps better morale because he is defending his home). But still, if I attack 20 times with my fighters, 19 times mine is damaged and only in 1 case the enemy is hit, that's a rate of about 95:5 and strikes me as too extreme - I would say 70:30 would be a good rate. BTW, AWACS planes can do the same job for the attacker as ground based radar for the defender...

Anybody here who knows how to change that? I think it's not part of the xml files, otherwise I would have already done it...

 

[Roland Johansen]

Thanks for the warm welcome
Well, OK. Let's say the defender has the advantage of a ground based radar (and perhaps better morale because he is defending his home). But still, if I attack 20 times with my fighters, 19 times mine is damaged and only in 1 case the enemy is hit, that's a rate of about 95:5 and strikes me as too extreme - I would say 70:30 would be a good rate. BTW, AWACS planes can do the same job for the attacker as ground based radar for the defender...

Anybody here who knows how to change that? I think it's not part of the xml files, otherwise I would have already done it...

I agree with you completely and said so in my previous post. I don't know how to change it. On the other hand, when you change this, then you should also change the damage that is done when a plane is intercepted since too many planes survive an interception (when they're fully healthy). I don't think that it is very easy to change it. What would be useful is if we knew how the interception chances are calculated and how the damage from an interception is calculated. Both are unknown as far as I know.



Bushface is completely correct about the diagonal movement. This is also the way the distance it calculated when determining base production from forests. At range 3 or less, chopped forests produce the full 30 base shields (modified for production bonuses from the city and game speed). But every tile further, the value of chopping forests decreases by 5. And again diagonal tiles count for 1.5 tile, rounded down.

If someone would like to see a screenshot of how distances are calculated, then I could make one. Is someone interested or is it completely clear?

 

[Arkturus]

I agree with you completely and said so in my previous post. I don't know how to change it. On the other hand, when you change this, then you should also change the damage that is done when a plane is intercepted since too many planes survive an interception (when they're fully healthy). I don't think that it is very easy to change it. What would be useful is if we knew how the interception chances are calculated and how the damage from an interception is calculated. Both are unknown as far as I know.

Now we got to the point of the problem. Well, if it's neither possible to alter the combat advantage for the defending fighter nor the chance to shoot down a 100% Bomber, the only possibility for me would be to add a special unit for my civ that is somehow modded in a way, that it has a better chance of damaging the defender resp. attacking planes (the only way I can think of is increasing the <combat></combat> value.
I never tried intercepting a fighter with a jet fighter: Does anybody know if in this case the attacker (jet) has a better chance of damaging the defender (WWII-Fighter)? That would prove that it makes sense to create a special unit that has a greater <combat> - value.

This would be of course a bad solution, because then it would be not possible for other civs to gain air superiority over my terrain, and it would be no solution for multiplayer either. But at least my current problem would be solved...

 

[Roland Johansen]

I've successfully fought a war of fighters/bombers against jet fighters/stealth bombers in an emperor level game. So the enemy had plenty of units. I just exhausted his air defences with the first 10 fighters or bombers (maybe 1 or 2 got shot down, the rest damaged). The next 15 removed the city defence bonus and did lots of collateral damage. This allowed my ground forces to successfully attack the enemy ground forces defending the city.

In another game, I had the superior air forces and they were also sometimes shot down. But I didn't do a scientific study at the rates of interception and the amount of planes that really got shot down. It didn't seem to make a significant difference. Maybe the strength value is only used for bombing damage (I know that it is used for that at least) and the rate of interception is only used to see if an interception occurs and has nothing to do with the damage of the interception. I really don't know, this is all speculation.

If you give yourself superior units, then you will do more bombing damage per strike. The rest, I don't know.

Your best chances (in my opinion) would be to write a bug report in the bug report forum. You'll have to explain in detail what is wrong about the system and why it is wrong and maybe a patch will fix it. That would be nice.

 

[Roland Johansen]

About the limited damage. The units are supposed to represent a number of aircraft (at least a squadron, probably several squadrons) in the game. A 10% loss ratio is considered excessive and more than that is very unusual. At least outside of propaganda or suicide missions. Having the bombers units damaged most of the time rather than getting shot down is really is not that unreasonable from a historical pov since whole units of bombers rarely got completely obliterated, some usually made it back. The same should also apply to defending intercepters.

The same could be said about most other units in the game. Most armies are not destroyed completely in a battle. But in real life, armies also don't regenerate free of cost like in civilization.

The reason why ground units and ships are destroyed in battle in civ (except for some units that can retreat when they are attacking and losing) is for a balance reason. If most units would survive and retreat then this could be exploited very easily. They didn't limit the chances that a unit succesfully retreats to some very low values (for instance horse archer 30%) for no reason.

 

[Bushface]

If the Jet Fighters were Harriers (AV8B in the U.S., I think) then they too could get the 'withdraw' facility and fly away backwards. Like the fabled oozlum bird . . .