Small rant + Question on Air support

[Gen-x87]

Ok my small rant goes like this. Yesterday playing the AI. I moved a King Tiger to the top of a hill. It did its work killing a Sherman and 3 Half Tracks. Then US Infantry rushed the hill since it was a capture point.

I had infantry supporting the tank killing the US infantry in the open. The tank started exchanging shots with a Sherman from about 500 meters. The Sherman dropped smoke and my King lost its LOS. Anywho when the smoked cleared my King turned to fight the infantry and open its side up completly. Not a big problem thinking that since the Sherman was targeting the King the King would eventually target back and turn to face the Sherman. It never happened. My King kept on fighting the Infantry and 55 seconds later and 3 shots that deflected off the turret the King was knocked out on the 4th.

Why would the King not think the Sherman was more of a threat than the infantry? Especially since I had a HMG 42 and 2 squads of SS rifle troops whacking the Americans out. I found that rather annoying.

My question is about Air support. The AI bought fighter\bomber support. Now how many turns does this last? I counted 17 mins of strafing and bombing. They killed 2 75s, a 150 inf gun, 1 puma, 2 haltracks, strafed the hell out of one company causing almost 2 full platoons to go into panic. Really whacked a different platoon on the other side of the map. Overall really disrupted my whole force.

Gen

[Dunnee]

I've found that air support strafes the enemy a random number of times in different games. An aircraft will usually expend its heavy ordinance(bombs, rockets) on the first two turns of its appearance, and I think it will continue strafing with its guns if there are targets exposed, but for a random number of turns. I played one game and was using a mechanized german company, that damned plane practically took out all my transportation, stranding my troops a good 10 min from the action.

[Juju]

Hmmm, As far as I know any plane will not show up more than 5 times in a given scenario, so it may have been 4 fighter/bombers you had against you. Seems rather a lot really...

Can anyone back me up on this?

[Kanonier Reichmann]

<BLOCKQUOTE>quote:</font><HR>Originally posted by Juju:

Hmmm, As far as I know any plane will not show up more than 5 times in a given scenario, so it may have been 4 fighter/bombers you had against you. Seems rather a lot really...

Can anyone back me up on this?<HR></BLOCKQUOTE>

Nup... you're on your own.

Regards

Jim R.

[Panther131]

Unfortunatly CM is not completely up to snuff in some area's of AI. For instance, in your case with the King Tiger, you would think that the TC would know that there is that sherman there on his flank, after all he was just firing at it. Hopefully these annoying things will be fixed up for the next series of CM, CM2 eastern front.

[ 04-22-2001: Message edited by: Panther131 ]

[Schugger]

Gen, if your KT was buttoned he might have never noticed that a Sherman was busy spanking him.

[I/O Error]

But the essential problem is that the AI has no "recall", no short-term memory of threats around it, even 5 seconds after losing sight.

"Out of sight, out of mind"

What utter chowderheads!

(Of course, would YOU want to be the one to code an AI that could "remember" all that stuff?)

[Mr. Hankey]

I'm not sure of the ins an outs of programming, but from what I understand, creating AI with a memory recall is extremely difficult and a dodgy proposition at best. There have (obviously) been numerous discussions on this topic. You may want to do a search if you're curious. Try not to get too excited about the AI not being omniscient. It knows too much as it is.

[Radar]

I'm sorry, but I don't get why coding the AI to 'remember' prior threat areas (especially from immenent destruction!)for a limited time after a blocked LOS to react accordingly, is overlooked. IMHO, its one of the more important things that should be improved upon. I've seen it already move and plan with a visible threat (and not too bad in most cases) on its own.

[CptSwampy]

Actually, if you think about it, it isn't a question of the AI having a "memory," it is a question of constantly (i.e., often) scanning the battlefield and assessing-prioritizing-reacting to threats. Memory is not necessary. When that tank went behind smoke, it is no longer "there" as a threat. The Kt's threat scan therefore doesn't include it. It doesn't exist.

But... when it moved out of the smoke, it should have been picked up and given a higher priority on the KT's next scan of the battlefield.

Speaking of "memory" just complicates the issue. The question is: is the KT scanning the battlefield often enough and assessing-prioritizing-reacting? Is it SMART in the way it assesses-prioritizes-reacts?

Just my 2cps.

[Sirocco]

<BLOCKQUOTE>quote:</font><HR>Originally posted by CptSwampy:

Actually, if you think about it, it isn't a question of the AI having a "memory," it is a question of constantly (i.e., often) scanning the battlefield and assessing-prioritizing-reacting to threats. Memory is not necessary. When that tank went behind smoke, it is no longer "there" as a threat. The Kt's threat scan therefore doesn't include it. It doesn't exist.<HR></BLOCKQUOTE>

To show how erroneous that concept is I'll outline a scenario. You're the commander of a Sherman, and you're facing a King Tiger. You order your gunner to fire a smoke round in front of the KT, which generates a smoke cloud which blocks LOS between you. You then notice some grenadiers around 150 metres away. What do you do? Do you order your gunner to load HE and rotate your hull and turret to face the grenadiers, or do you back into cover as quickly as possible before the smoke cloud clears? If we follow the "it doesn't exist" logic you're already issuing orders to your gunner, and he's loading that HE round. How many rounds do you think you can get off before the KT "exists" again..?

That's one example why CM, or CM2, needs "memory".

[Sirocco]

<BLOCKQUOTE>quote:</font><HR>Originally posted by Mr. Hankey:

I'm not sure of the ins an outs of programming, but from what I understand, creating AI with a memory recall is extremely difficult and a dodgy proposition at best.<HR></BLOCKQUOTE>

The collective spotting system used in CM would make it extremely difficult, but when the replacement is coded, "memory" should be a part of the process. These things are always much harder to add later than they are to include from the beginning.

[CptSwampy]

Hmmm... good point, Sirocco... but lets say that Sherman backs down over a hill top all on its own... how long do you want to keep your turrent pointed at the empty hilltop whilst the infantry are approaching your flanks?

Works both ways, I would say.

Perhaps the KT needs to keep that hilltop "in the back of its mind..."

[Sirocco]

<BLOCKQUOTE>quote:</font><HR>Originally posted by CptSwampy:

Hmmm... good point, Sirocco... but lets say that Sherman backs down over a hill top all on its own... how long do you want to keep your turrent pointed at the empty hilltop whilst the infantry are approaching your flanks?<HR></BLOCKQUOTE>

How does a human memory work? As the TC you ordered the driver to reverse the Sherman into cover. You know the KT is still there, but you have infantry approaching. If you forget about the KT you might go forward again to improve your LOS on the grenadiers, and then be surprised by the lurking KT, with the only hope that it, too, has forgotten about you and is looking somewhere else. With "memory" you wouldn't stop reversing.

Just because you remember the KT, it doesn't have to mean that you discount other threats.

A good threat calculation algorithm allows you to prioritise them in terms of personal threat and threat to others, taking into account LOS and other factors, and likelihood of achieving a kill etc, which CM does quite well now, but an excellent one doesn't discount a threat just because, for the moment, it's no longer visible.

[Gen-x87]

Gen, if your KT was buttoned he might have never noticed that a Sherman was busy spanking him."

He was buttoned but

1. He was just fighting with him 1 minute earlier

2. The sherman deflected 3 shots off his turret.

Some of you bring up some good points. I would say that possibly this could have been alleviated if the whole tank did not turn to face the troops and only the turret?

Maybe when the whole tank turned it could not make a LOS on the sherman eventhough the Sherman could on the KT?

But does anybody have a concrete answer on the air support? I was astounished at the level of destruction it caused. I figured it may stick around for 5 mins but 17 mins? I had to go to class and quit on the 17th min. Who knows it could have gone on even longer.

Gen

[Sirocco]

<BLOCKQUOTE>quote:</font><HR>Originally posted by Gen-x87:

But does anybody have a concrete answer on the air support? I was astounished at the level of destruction it caused.<HR></BLOCKQUOTE>

I think CAS isn't modelled accurately, but BTS would disagree, and have in the past. I personally believe CAS should "shock" more than KO. And in the example you gave, you were particularly unlucky, but that happens sometimes.

[Mr. Hankey]

<BLOCKQUOTE>quote:</font><HR>Originally posted by Sirocco:

...I personally believe CAS should "shock" more than KO. And in the example you gave, you were particularly unlucky, but that happens sometimes. <HR></BLOCKQUOTE>

I don't know who told you that, but a flight (four planes) of P-47's can completely destroy a tank column of about ten tanks. One P-47 carries three 500 pound bombs, plus they have eight .50 cal machine guns and about 2500 to 3000 rounds of ammo. One 500 pounder will completely anilalate a tiger tank especially if it hits the engine compartment area, not to mention the effect that 8 .50 cal's, set to converge on the area about the size of a silver dollar will have.

[ 04-23-2001: Message edited by: Mr. Hankey ]

[FriendlyFire]

CAS is, and always has been, a crap-shoot. Sometimes it doesn't show up, sometimes it shows up and shoots the wrong tanks, sometimes it drops one bomb and then flees, and sometimes - well, sometimes it rains on your whole parade.

Look at it from the pilot's point of view: he's found a target-rich environment with no flak. Nothing's going to shoot him down, he's having a great time strafing everything in sight, and he's (apparently) got infinite fuel and ammo...

[Soddball]

Air support rain on my parade GOOD. Was playing axis assault against allied AI, AI had bought 3 planes(!). Lost 8 tanks plus a dozen supporting vehicles, and the game, as a result.

Still, I call that realistic - and next time, I will buy some AA units.

[tss]

Mr. Hankey wrote:

I don't know who told you that, but a flight (four planes) of P-47's can completely destroy a tank column of about ten tanks.

"Can", yes, but in general, not "will".

One 500 pounder will completely anilalate a tiger tank especially if it hits the engine compartment area

The key word is "if". It was pretty hard to drop a bomb close enough to knock a tank out.

I'm not certain what is the lethal radius for a 500 lbr bomb with respect to a heavy tank, but I'd guess that it is not much more than 15 meters. So with 15 meters there's about 30 m error margin when dropping a bomb. A plane zooming in at 350 km/h goes 30 meters in 0.3 seconds or so. You don't have to make a big mistake in releasing a bomb and it will miss altogether.

- Tommi

[Priest]

Uhm......hmmmmmm. Guys if you look and search for simliar tank hunting threads from the past you would know that Steve and Charles have answered this question about 3 times. Personally I find it funny that you complain about such a trivial thing. Could it happen on the battlefield? Yes it could. You guys are arguing about a point that may have actually been the TacAI resolving a random chance issue. The TC of the KT made a mistake and it cost him. Also before you post little rants you should probably remember that in CMBO your rants consist of "my guys made realistic but bad decision: rants while other games have people asking such things as playability. Personally I have not had a tank make a "bad" decision in a long time. Also what the heck are you doing letting a sherman get 500m to a KT. If it were a 1000 meters you would be posting that the KT is UberTank and can ignore shermans even when they flank to a side. That is my 2 cents.

[Mr. Hankey]

<BLOCKQUOTE>quote:</font><HR>Originally posted by tss:

The key word is "if". It was pretty hard to drop a bomb close enough to knock a tank out.

I'm not certain what is the lethal radius for a 500 lbr bomb with respect to a heavy tank, but I'd guess that it is not much more than 15 meters. So with 15 meters there's about 30 m error margin when dropping a bomb. A plane zooming in at 350 km/h goes 30 meters in 0.3 seconds or so. You don't have to make a big mistake in releasing a bomb and it will miss altogether.

- Tommi<HR></BLOCKQUOTE>

Yeah, and your point being? Saying "if" means absolutly nothing in this discussion. You make it sound as though a fighterbomber has about as much chance of hitting a tank as you do hitting the lottery. Well, as a matter of fact, tanks were destroyed by fighters on a regular and frequent basis, on the eastern and western front. The Il-2 was designed pretty specifically for tank busting. If you want to talk about the Stuka, we are looking at about a 70%+ hit rating. I remember reading about one Stuka pilot (the name escapes me now) who stated that he could, as a matter of routine, put a bomb in a three foot radius at any given time.

Sorry if I sound a bit harsh, except your statement is only sheer speculation on your part with no basis in fact. Simply because you can do algebra does not make you an authority on dropping bombs. Hell, I can do enough differental equations in my head to intercept a projectile traveling at about 40 mph for a distance of over 80 yards, doing continuous updates during it's travel and know almost exactly where it will be at it's end point. It's called catching a football. With some practice, I'm sure those guys would have no problem, given the right circumstances, of hitting pretty much anything they wanted.

Anyway, I only know what I've read, and the numbers say it wasn't all that hard for a fighter to kill a tank. 'nuff said.

[ 04-23-2001: Message edited by: Mr. Hankey ]

[Sirocco]

<BLOCKQUOTE>quote:</font><HR>Originally posted by Mr. Hankey:

Anyway, I only know what I've read, and the numbers say it wasn't all that hard for a fighter to kill a tank.<HR></BLOCKQUOTE>

This has been discussed often. Put a stationary tank in the middle of an open field, with no AA cover, and, yes, it was probably "easy" to kill a tank. Move the tank at speed and try to hit it. The chances drop considerably. Put some cover into the field and give the tank some AA protection. What are the chances of a kill then?

The biggest weakness of any tank is it's logistical tail, and that's what aircraft were best at killing, not tanks.

[Olle Petersson]

<BLOCKQUOTE>quote:</font><HR>Originally posted by Mr. Hankey:

... tanks were destroyed by fighters on a regular and frequent basis, on the eastern and western front. The Il-2 was designed pretty specifically for tank busting. If you want to talk about the Stuka, we are looking at about a 70%+ hit rating. ...

Anyway, I only know what I've read, and the numbers say it wasn't all that hard for a fighter to kill a tank. 'nuff said.<HR></BLOCKQUOTE>Tanks were destroyed by attack aircraft, allright.

The interesting point is under what circumstances it happened.

German philosophy was to score direct hits at specific targets. This was done using dive bombers (STUrtzKAmpf Flugzeug) piloted by highly skilled pilots.

They could pinpoint a single tank and hit it.

USAF was on the other end of the scale. Rarely (if ever?) using dive bombing, instead relying on level bombing with much lesser accuracy.

Their tactic was to use many-on-many. Sending several aircraft to drop their payload at a concentrated target of several tanks. This is more a matter of area fire, and surely there will be a couple of direct hits on tanks in the area. The effect being roughly the same as using concentrated heavy artillery fire.

RAF was roughly like USAF, but with a little more precicion and larger calibre guns for strafing.

Soviet air force seem closer to the German style.

Cheers

Olle

[tss]

Mr. Hankey wrote:

Yeah, and your point being?

My point being that the percentage of tank losses that were directly knocked out by aircraft was not too big. I've seen various figures, and not a single one has been larger than 7-8%. (The proportion of indirect kills caused via supply interdiction and subsequent abandoment was quite large, but difficult to quantify).

You make it sound as though a fighterbomber has about as much chance of hitting a tank as you do hitting the lottery.

Depends on the lottery. The Finnish national lottery has one change in 15 million for the main prize. Most fighter pilots were much better than that.

Well, as a matter of fact, tanks were destroyed by fighters on a regular and frequent basis, on the eastern and western front.

Yes, they were. But other losses (mines, AT guns, tanks, infantry AT weapons) were heavier. Also, there were a lot of sorties per each kill.

The Il-2 was designed pretty specifically for tank busting.

And it was quite good in it. However, having a complete air superiority and dozens of Il-2s doesn't automatically mean that all enemy tanks will go up in flames. For example, Finns lost 30-40 tanks in summer '44 (I can't remember the exact figure and my sources are at home). If I remember correctly, not a single one was destroyed by air attacks. Even anti-tank rifles caused more casualties (one T-26 platoon).

If you want to talk about the Stuka, we are looking at about a 70%+ hit rating.

There are no Stukas in CMBO. Dive bombers were inherently more accurate than fighter-bombers.

I remember reading about one Stuka pilot (the name escapes me now)

Hans-Ulrich Rudel, most probably.

ho stated that he could, as a matter of routine, put a bomb in a three foot radius at any given time.

Rudel is not a particularly good example of the capabilities of an average combat pilot. The man flew over 2500 combat missions, far more than any other pilot in any country.

At Vuosalmi the Germans lost (to light AA fire) one Stuka pilot (can't remember the name, he had a Knight's Cross) that was considered to be very experienced and one of the best. He had a little over 600 combat flights at the time.

Sorry if I sound a bit harsh, except your statement is only sheer speculation on your part with no basis in fact.

Sheer speculation true, but has the basis on studies that were conducted on tank wrecks (that showed that only 1 wreck in 20 could be identified as aerial kill). Oh, and I'm pretty thick-skinned.

Simply because you can do algebra does not make you an authority on dropping bombs.

Agreed.

Hell, I can do enough differental equations in my head to intercept a projectile traveling at about 40 mph for a distance of over 80 yards, doing continuous updates during it's travel and know almost exactly where it will be at it's end point. It's called catching a football.

The speed of the football: 40 mph = 64 km/h.

The speed of attacking aircraft: 350 km/h.

A five-fold difference.

With some practice, I'm sure those guys would have no problem, given the right circumstances, of hitting pretty much anything they wanted.

Given the right circumstances. However, the enemy wouldn't usually give the right circumstances. Actually, they would usually do as much as they could to make the circumstances as wrong as possible.

The main point of my calculation was to stress the fact that the pilot had to have a sub-second timing in releasing the bombs, if he aimed at an individual tank. (Bombing along a tank column would be easier because there would be more than one tank on his aiming line). The exact figures are irrelevant.

Note that I don't claim that "sub-second timing" makes it impossible to aim bombs and hit. (After all, baseball players need also pretty good timing to hit the ball). Just that even a small mistake from the pilots part will make the bombs fall far enough from the target to leave armored targets intact. That mistake may be simply flinching when an enemy AA mg starts firing tracers.

Assessing bombing accuracy is actually a pretty difficult thing to do. Most pilots will report that they either got a direct hit or at least a near miss. They probably even believe that themselves. It's the same thing that happens in fighter kills (all sides regularly overclaimed by a factor of 2-3) or sniper kills (whether a sniper hits or misses the target will immediately drop down and it may be difficult to find out what happened to him).

- Tommi