Deer-in-the-Headlights: Reaction of Tanks to Strafing Run

[von Lucke]

I had six T34's line-abreast in a balka (that's a ditch, to you non-Russkis), ready to go over the top in a cavalry charge. Each had a squad of tankodesantniki holding on for dear life to the engine covers. They start Fast moving --- maybe a whole meter --- and then the 2nd one from the right get's targeted by a jabo, some 1500m up, coming in perpendicular (rather sweet target for the pilot, I suppose).

So, then what?

All six tanks stop dead and their TC's slam the hatches! All six infantry squads bail and start crawling away on the ground!

After the jabo strafes it's target (and it only targeted the single tank), with no periferal damage to any other unit, do the tanks start moving again --- of course, now without their passengers. At the end of the turn, the armor ends up some 100m from it's infantry support, who are on the other side of an open field. Bad.

So, then: Should all 6 tanks have frozen at the sight of a jabo, or just the one targeted? And should they have stopped at all? Wouldn't it have made sense to try and move out of the strafing run --- after all, once the AC lines up, moving out from under it would prolly be a good idea. And why did all six infantry squads bail? Wouldn't just the targeted one jump ship?

[Malakovski]

How could the targetted tank tell it was the one targetted? All they see is a plane coming in their general direction, probably about to drop some heavy ordinance on their heads.

That said, I would think freezing would be the last thing the tanks would want to do. If I were an infantrymen, I might well bolt for cover, but as a tanker I think I'd step on the gas, so to speak.

[Michael Emrys]

I guess the guys on the ground couldn't tell for sure who was being targetted. When a big, mean ol' enemy plane starts diving in your direction, it's always a good idea to assume he's got you in his sights, or soon will if you're the only donkey standing out after everybody else has dived into the ditch.

As to whether it's a better idea for the vehicles to disperse under attack, I'd say so. But it doesn't seem to be what got done sometimes. For one thing, if the plane is able to disrupt their formation and disorganize their attack, he's already done a pretty good job of what he was sent out to do.

Michael

[akdavis]

For the infantry, they did the best thing possible. Let them dismount and clear the fire-drawing tanks. For themselves, they did the worst possible thing. Option 1: Move into closest cover as quickly as possible. Option 2: At least move somewhere as quickly as possible. Option 3: Move anywhere even if it's just ten feet at a slow crawl, at least you won't be exactly where the jabo saw you on his last run. Option 4 (a.k.a. defect to the other side): stop in place making yourself as easy of a target as possible.

[akdavis]

Well, I just ran a test through several times. German air (cannon, mg) versus two platoons of t-34s (one crack, one green) with inf riders. Pretty consistent results. Tanks will not stop under air attack unless immobilized (which is common), but inf will usually dismount immediately. I even had a green tank lose its TC during a strafing run and keep on moving to waypoint despite being shocked.

BUT, here's an interesting little observation: the aircraft seem to have an uncanny ability to spot more experienced tank crews, always attacking my crack units before green ones. Repeated several times with same results. 4 aircraft in the air, all attacked crack tanks first, only switching to green tanks once all 3 crack tanks were KO'd or immobilized. Me thinks this could lead to some unbalancing effects of air support.

[Fly Pusher]

Originally posted by akdavis

BUT, here's an interesting little observation: the aircraft seem to have an uncanny ability to spot more experienced tank crews, always attacking my crack units before green ones. Repeated several times with same results. 4 aircraft in the air, all attacked crack tanks first, only switching to green tanks once all 3 crack tanks were KO'd or immobilized. Me thinks this could lead to some unbalancing effects of air support.

If this is true I would agree that the selective destruction of more experienced crews would greatly enhance the relative 'power' of aircraft in an artificial and non-historic manner. I would guess that more experienced crews would be LESS likely to be targeted due to better use of cover or more acccurate anti-aircraft fire.

Has anyone else made this observation ?

[von Lucke]

In terms of Real Life, I agree, if I was sitting on the back of a big metal box with no cover, and an NME AC came swooping in, I'd bail. That's because, as was pointed out, you wouldn't be able to tell who he was going after.

Unfortunately(?), in CMBB you see a big yellow line connecting the target to the AC. I'm just wondering why the AI, which knows who is targeting whom much better than I, would cause all vehicles / troops to behave like the target, when they obviously were not.

If you target one tank with direct fire and cause it to cease it's action, should every tank nearby do the same?

As for the tanks freezing, I ran the turn a couple of times, and in every instance they paused while under attack --- in brush, wheat, and clear terrain.

[Foxbat]

Originally posted by von Lucke:

In terms of Real Life, I agree, if I was sitting on the back of a big metal box with no cover, and an NME AC came swooping in, I'd bail. That's because, as was pointed out, you wouldn't be able to tell who he was going after.

Unfortunately(?), in CMBB you see a big yellow line connecting the target to the AC. I'm just wondering why the AI, which knows who is targeting whom much better than I, would cause all vehicles / troops to behave like the target, when they obviously were not.

If you target one tank with direct fire and cause it to cease it's action, should every tank nearby do the same?

So you want your virtual tankers to look up at the sky spot the yellow line and do something that in real life would have been absolutely impossible?

The direct fire comparison doesn't hold, when engaged by direct fire the tanks have a pretty good idea who's being shot at, while that incoming plane could drop it's ordanance on each and every of these tanks (the tankers can't tell where the ordanance is going until it starts coming down).

As for the tanks freezing, I ran the turn a couple of times, and in every instance they paused while under attack --- in brush, wheat, and clear terrain.

Where they freezing because they stopped, or because they had to let the animation play out? This can cause situation to appear visually different from how they actually occur (think Arty still falling after the 60 second mark has been reached).

[Foxbat]

Originally posted by Fly Pusher:

I would guess that more experienced crews would be LESS likely to be targeted due to better use of cover or more acccurate anti-aircraft fire.

While I agree that planes shouldn't be able to pick of experienced troops first I don't think the opposite is true either. There isn't much cover to use if you're in a tank (besides you only see the a/c coming when it's to late to start using cover), and anti-aircraft fire would apply only to units that have aamg's (Shermans) and that are actually willing to use it (I know that among Panzer crews in France at least this was a big- no-no).

[Bruno Weiss]

To my thinking, the only part of what was described that might not portray an expected reaction to the aircraft attack, would be the tanks just sitting there. More than likely, in a real aircraft attack, the tanks would scatter out of line/formation. Otherwise, I would have expected the infantry reaction. Not sure however, what sort of AI programming it would take (probably more than technology allows at the moment), to get the individual tanks to recognize an aircraft attack on one unit in their line/formation as a threat to them all and then proceed to disperse their units. I would think that would call for some pretty fancy AI calculations to say the least. Than again you say the infantry did, so I dunno.

[ October 02, 2002, 09:05 AM: Message edited by: Bruno Weiss ]

[von Lucke]

Originally posted by Foxbat:

Actually, yr reiterating my point. I don't want them to do that --- but that's what's happening: 1 nanosecond after the primary tank is targeted, all six tanks button up and freeze at the same time. All six squads bailed at the same time. Borg spotting in full effect. Considering the AC was some 1500m to the east, and all tanks were facing north, that was one hell of a whip-lash those TC's have...

As for direct fire, try playing EFOW. How often do you see where the first few shots come from --- especially with ATG's? You always spot an aircraft, now don't you?

I'm sorry, I don't quite know what you mean here. You talking about a hardware issue? There is no "animation to play out" with a strafing run. It doesn't continue after the fact, like arty fire. Besides, this all occured in the first 10 seconds of the turn.

[Schoerner]

Does anyone know about special doctrines dealing with air-attacks against tanks?

Something like:

if Inf transported on tanks and no cover available, tank stops and Inf. takes cover behind the tank?

Or no matter if cover or not, inf. running away from tanks as far as possible?

What about movement routes of tanks under air-attack? Towards the attacking-line or vertical to it?

[Foxbat]

Originally posted by von Lucke:

Actually, yr reiterating my point. I don't want them to do that --- but that's what's happening: 1 nanosecond after the primary tank is targeted, all six tanks button up and freeze at the same time. All six squads bailed at the same time. Borg spotting in full effect. Considering the AC was some 1500m to the east, and all tanks were facing north, that was one hell of a whip-lash those TC's have...

Ok I see what you mean. It's all a bit fast*/coincidental, but by and large planes coming in for a strafing run are pretty easy to spot and if one TC sees the guy next to him button up and the infantry bail out he's going to have a pretty good idea what is going on (and if he doesn't there's a good chance he'll button up first and ask questions later ).

* Buttoning up tanks litteraly no time in the game, we'll have to live with that.

As for direct fire, try playing EFOW. How often do you see where the first few shots come from --- especially with ATG's? You always spot an aircraft, now don't you?

I've seen massed tanks stop look around and start hammering away in EFOW when ATG's started firing, especially in Zitadelle Schwerpunkt where you have that enormous mass of tanks and a few unspotted ATG's taking potshots at them.

I'm sorry, I don't quite know what you mean here. You talking about a hardware issue? There is no "animation to play out" with a strafing run. It doesn't continue after the fact, like arty fire. Besides, this all occured in the first 10 seconds of the turn.

When CMBO came out and I still had my ancient (and even at the time outdated) compu I had things like this happen.

But I guess this is just the time it takes for the inf to dismount, or did you also do it with "unencumbered" tanks?

[Fly Pusher]

I agree that there may be an issue with the speed that tanks react to air attack.

However the thing that really worries me is the apparent ability of an air boy approaching at over 300mph to distinguish the experience level of people on the inside of buttoned tanks and take out the most experienced tanks first.

Über Borg spotting if you ask me...

A comment from on high would be appreciated !

[xerxes]

The AI has a better ability then a player to determine enemy troop quality/HQs etc. That's why a sniper will target a HQ when you (the player) haven't ID'ed the HQ.

Doesn't seem right.

But, then again, having the AI have more information that isn't shared with the player is a reasonable method of reducing borg spotting effects.

[Lt. Kije]

This [AI targeting with supernatural knowledge of unit quality/value] is, I believe, an ongoing problem with CM. Even way back in CMBO days, it seemed to me you could put five units in five halftracks, button them up, then drive them into danger and the enemy would target the one halftrack holding the multiply augmented company commander, or whichever unit was of highest value.

I could be wrong about this, and if so I look forward to someone correcting me based on empirical test. I hope I am wrong. But I have a fuzzy memory, moth-eaten and indistinct, of trying the experiment described above and finding these results.

It is also possible, memory being heavily biased by the availability heuristic, that the easily recalled outrages of losing only the most valued unit in a group of five are the only memories that come to mind. Memory is weird and I'm thus not asserting the point about supernatural targeting as boldly as I feel it within my tortured soul.

And I'm not sufficiently exercised about this to actually carry out a multi trial experiment.

The point: this clairvoyant targeting is [may not be] specific to air attacks.

-- Lt. Kije

Fuzzy. Moth-eaten. Indistinct.

[Frunze]

Originally posted by xerxes:

The AI has a better ability then a player to determine enemy troop quality/HQs etc. That's why a sniper will target a HQ when you (the player) haven't ID'ed the HQ.

Doesn't seem right.

But, then again, having the AI have more information that isn't shared with the player is a reasonable method of reducing borg spotting effects.

Also, this works for the TacAI of the human player as well as for the computer opponent. So it's not such an advantage for the computer opponent. But maybe that's not what you meant.

[Renaud]

I haven't experienced tanks vs aircraft yet in CMBB, but I'd like to chime in on general policy in real life for vehicles under air attack.

When under air attack (enemy planes spotted in the air) it is advisable to not move the vehicle. The reason for this is that the human eye (brain really) percieves moving objects far easier than stationary objects. It's very difficult to pick out camoed vehicles from the air until they move. Also, moving fast, zig-zagging or other nonsense does not actually make you harder to target. This is because your max speed relative to the plane's speed is miniscule (20kph vs 300kph) and only servers to stir dust and cause motion which attracts attention. In other words, motion makes you stand out from the surrounding ground, but isn't going to significantly impact the course adjustment of a plane approaching at 200-300+kph.

This is what I was taught in armor school and what I think most modern armor forces practice. (This applies to fixed wing not rotary wing. If an ATGM coming at you, you should zig-zag.) That being said, in WWII reality there's no telling what behavior you might see. I think only disciplined and trained TC's are going to refrain from hauling butt in a futile effort to evade attack.

Ren

[PeterX]

However the thing that really worries me is the apparent ability of an air boy approaching at over 300mph to distinguish the experience level of people on the inside of buttoned tanks and take out the most experienced tanks first.

Yeah, ouch. They're suprsingly adept at picking out command tanks as well. I also noticed in my ongoing PBEM that an enemy AT pillbox showed the same tendency to single out the most valuable units; letting pass four tanks before zeroing in on the platoon leader. Of course it might have been my disordered magination.

Apparently, the game assigns a weight to each target which incorporates the exp level. I really doubt a FB swooping down on a smoky battlefield at 300mph could distinguish a Tiger from a Mark III let alone an elite crew from a veteran crew.

This is a bug which needs to be squashed in 1.01. Too bad, because they did such a bang up job with the new airpower code.

[ October 02, 2002, 03:08 PM: Message edited by: PeterX ]

[Chad Harrison]

To answer the original question, I have had this happen to me on numerours occasions also. And to me, it makes sense. If you are in a tank, and see a plane gunning for something around you, how can you know if he has a bomb or not? Or a 37mm cannon? Especially if you dont have any close AA support.

I would of gone inside! And if I was on the back of the tank, I would of ran too! A tank makes a big fat target!

Chad

[Silvio Manuel]

It seems like the incoming Aircraft causes an Area Morale effect, much like one big 300mm shell hit them or something...which makes sense b/c the plane's ordance has an area effect if it's bombs or rockets.

[wbs]

I was interested to see that in CMBB you can now get airpower assigned to your forces in a QB. I have only played one game with airpower. I was kicking my opponent's butt, and the Sturmovik (or whatever) made three passes and knocked out three tanks, which almost cost me the game.

Of course, since he had airpower his ground forces were that much weaker, so maybe it all evened out.

I still thought that the success rate for the strafing attacks (100%) was a bit much, but I probably have too small of a sample test population to accurately generalize.

[Madmatt]

I set up a test scenario with two platoons of T-34's. One platoon was Elite and one Conscript.

Both platoons were next to each other in a open area.

Against them I had a single Ju-87G Tank Buster.

In multiple replays the Stuka seemed to randomly pick it's target from both platoons equally. It also did not seem to ever fixate on the HQ tanks anymore than the other units in the platoon.

I did notice that if the plane picks a enemy unit and does NOT kill it in the first pass it will be more likely to retarget that same unit in later pssses. I know that when I play IL-2 I often find myself doing the same thing though.

Madmatt

[ October 02, 2002, 03:21 PM: Message edited by: Madmatt ]

[PeterX]

It also did not seem to ever fixate on the HQ tanks anymore than the other units in the platoon.

Strange, Madmatt, I got the opposite result. I made a platoon of 5 tanks, elite thru conscript. The FB attacked either the Elite or the Crack crew each time, ecept once when it picked on the Veterans. Admittedly, I ran the test only 6 times.

Anybody else tried it?

[ October 02, 2002, 03:27 PM: Message edited by: PeterX ]

[Madmatt]

I usually do tests in batchs of 50 or more runthroughs as anything less is too inconclusive. I will run some other tests and see what I find though.

Madmatt