Is it too easy to spot AT guns?

[Scipio]

Originally posted by redwolf:

Stealth-bonus HQ control do not make firing guns harder to spot in CM.

Why?

[Sigurd]

Originally posted by Scipio:

</font><blockquote>quote:</font><hr />Originally posted by redwolf:

Stealth-bonus HQ control do not make firing guns harder to spot in CM.

Why?</font>

[Redwolf]

What can I say, you need to run tests. I couldn't notice that comparable AT guns in comparable positions in comparable distances are harder to spot when they are under extra-stealth bonus HQ control when they fire.

For a change I don't think this detail is an unrealistic aspect of CMBB. There is not much a HQ can do to make his gun be spotted harder once it start shooting.

[YankeeDog]

Originally posted by JasonC:

Now, is all of this historical? I submit that it is not. That T-34 company, without supporting artillery to suppress the 88s, or anything ahead to scout out their locations, or any folds in the ground to hide in during their approach - ought to be toast. Part of the problem is definitely the accuracy of their HE replies, and the ease with which they therefore KO dug in guns at ranges well over a kilometer.

First of all, thank for the interesting test. It certainly gives an insight as to how spotting works in CM.

I'm not sure what to think about the issue of whether or not your test results are "realistic". Frankly, I've never read any actual combat accounts that come close to the situation you depict in the test. To be sure, I've read combat accounts of a few well-placed and well-hidden ATGs holding up armored advances for quite some time, but there have usually been at least a few things that are different. Here's a few of the issues that come up for me:

In some of the accounts of a few PaKs holding up large armored advances, the ATGs succeeded by scoring several tank kills at very long range without the attacking tanks having any clue where the guns are. The rest of the tanks then get spooked and retreat, or just pull back and wait for support assets to move up for a more coordinated attack to be mounted.

These Real World™ stories don't jibe with your results because in your test, it is distinctly to the guns' disadvantage to open up at long ranges because the hit percentage is very low and they run out of ammo before they can score an appreciable number of kills.

I think this discrepancy is relatively easy to reconcile. First of all, your test was entirely without TRPs. I suspect Real World™ PaK fronts made heavy use of pre-ranged fire to improve accuracy of emplaced guns. I have even read stories of gun commanders pacing distances to nearby terrain features on foot during quiet moments in order to establish ranging information. In my opinion, in CM, this kind of preparation would be best represented by the liberal use of TRPs along likely avenues of approach. I suspect a dozen or so TRPs (and narrow covered arcs focused upon them) would have improved your Pak43's kill percentages at long range dramatically. This would mean that substantially fewer T-34s would actually penetrate to 1.5km, when they have a chance of spotting the guns. Redwolf’s comments about mine belts are also well taken and would further improve the effectiveness of the TRPs.

Furthermore, as we are all aware, CM does not model ammo replenishment. This becomes a particularly glaring omission when you have two ATGs with a limited ammo supply facing what could be used as a textbook example of a "target rich environment". IRL, I'm guessing that as soon as those PaKs see 13 T-34s charging them headlong, they wire back to the local CIC for (1) more ammo and (2) any available mobile anti-armor reserves. In the Real World™, I suspect the guns might be willing to expend more ammo at longer range in the hopes that this would slow up the advancing tanks long enough for more ammo and reinforcements to come up. The defensive use of artillery to force the tanks to button up and slow down is also a distinct possibility.

Finally, there is an inherent artificiality in your winning strategy of charging to 1.5km from the PaKs, and then hunting until you see them. I submit that such a strategy would not be used by a Real World™ tank company commander under long-range ATG fire. Why? Because, IRL that company commander doesn’t know that the couple of ATGs he has vague sound contacts to are alone. For all he knows, there is a battery of Pak38s positioned 500m in front of the Pak43s, just waiting for the Tanks to approach close enough to get good kill percentage shots. Or, there could be another pair of PaK 43s positioned 500m behind the first two (i.e., defense in depth). There are many possibilities that the Tank Commander has to worry about. Even an HMG team or two in front of the guns should have improved the guns’ chances by forcing the T-34 to button up as they got closer. It is also worth noting that these tactical concerns also apply to CM players. On the defense, I will happily let you charge my 88s set up at the rear of the map if that means I get flanking shots with my PaK38s and maybe even IAT teams as you get closer!

Furthermore, there is one “Gamey” aspect to CM battles that can probably never be eliminated: CM fights take place on maps of determinate size. Real World™ PaK fronts were not 480m wide; they could extend for many kilometers along the front. The closer those T-34s get to the PaKs in front of them, the more likely it is that they put themselves into range of deadly enfilading fire from guns slightly farther up or down the defensive line. Of course, Real World™ attacks don’t happen just along 480m frontages most of the time, either, so the guns up and down the line might have their own problems. The point is, IRL, the T-34 company commander has many more potential dangers to worry about, and I think he would likely your “Charge to 1.5km and Hunt for the guns” an unattractive strategy. If a T-34 stops 1.5 km in front of one PaK to engage it, he is also potentially only 2.1 km from another PaK 43 a full 1.0 km further up or down the front (assuming a straight line front, of course). As your test shows, at 2.1km, a PaK 43 can engage and destroy T-34s with a reasonable hit chance so long as the T-34 isn’t moving .

So, I don’t think the “Charge to Spotting Range” solution is very practicable for the Real World™, or for CM players, for that matter. Your points about the effectiveness of relatively small caliber AT fire on dug-in guns and the predictability of spotting range are well taken, though, and I think bear further examination. Part of the HE fire effectiveness thing may have to do with the fact that gun crews cannot re-man guns in CM, so abandonments are permanent in CM – I knock out many more guns by crew abandonment than I do actually KOing the gun in CM.

The predictability of the spotting algorithms is curious. It seems to only apply to firing units. Spotting hiding guns (or infantry, or whatever) is much less predictable. While you generally have to get very close to a hiding unit to spot it, sometimes you may not spot a hiding unit until you’re practically on top of it, and sometimes you get lucky and spot it from further away. Yesterday, I had an infantry squad practically walk over an enemy 76.2mm gun in scattered trees (daytime) before spotting it. Usually you have to get pretty close to see them, but that was the shortest first sighting range I’d ever seen for an infantry unit moving up on an ATG. Conversely, in the past, I have gotten lucky and spotted hiding guns from 100m out or so, though only rarely does this happen. It seems to me both spotting algorithms should have a considerable random factor built in.

Cheers,

YD

[Scipio]

Originally posted by redwolf:

What can I say, you need to run tests. I couldn't notice that comparable AT guns in comparable positions in comparable distances are harder to spot when they are under extra-stealth bonus HQ control when they fire.

For a change I don't think this detail is an unrealistic aspect of CMBB. There is not much a HQ can do to make his gun be spotted harder once it start shooting.

Not in reality. But we are talking about a computer program that (hopefully ) does what the programer wants it to do. The simple question that only the programer can answer :

Gives a HQ a stealth bonus to a gun, and if yes, does the program set it to zero when the guns has fired a shot?

[JasonC]

Thanks Yankee Dog for the extensive reply. I don't think my situation is so artificial, actually. There are indeed plenty of AARs of tank masses held up at range by unlocated PAK. But there are also any number of AARs, particularly from early to mid war, of masses of tanks trying to charge through the whole range envelope of dug in PAK, without combined arms support. And these affairs generally resulted in the destruction of entire tank brigades in a single afternoon. I will address closing tactics for CMBB as it stands more, below.

Another fellow mentioned homing in. It is much better in CMBB, that is true. And against the cluelessly non-aggressive AI, you can get good results at very long range because of it. See, homing in depends heavily on tanks stopping, or moving only slowly.

The AI will mill around aimlessly after a few shots have come its way, when it has only a sound contact and a flag to relate to. And even at 2.5km, if tanks remain stationary while the vet 88s home in with round after round, the hit chance does climb, and dramatically so.

Using the combo of saving only HE for closer range fire, and homing in, an AI commanded T-34 company can be stopped before it closes. The hit chances seen in practice rise to 10-12%. That is enough for a pair to get a tank a turn, occasionally 2. Fire off 60-70 AP, and half or more of the company is KOed, without ever really getting into gear (under AI leadership, understand).

The HE is not effective at long range, incidentally. At 1600-1750 yards - thus outside the "spotting range", just - you wind up getting only 1/4 to 1/2 of the HE hits to do anything. But that is enough, if there aren't many tanks left at that point. So the AI can indeed be stopped (by 88s, that is. With PAK 40, the ammo limit problem is harder to get around because the HE is not much of a backup).

If the right closing behavior is used instead, the 88s still don't have a chance, even opening up instantly and relying on homing for the AP at long range, and the reserve of HE for when the T-34s get closer. The reason is that the hit chance never really rises beyond 7-9% (at best), even with considerable homing, against "fast" moving targets. And it only take about 3 minutes for the T-34s to cross the outer part of the range envelope, to within sighting range. In that period, only 1-2 T-34s will be taken out. The difference compared to the AI is (1) less exposure time and (2) less homing bonus, both due to "fast" movement.

Now, Yankee Dog raises a legitimate issue when he asks if this is something a commander - in CMBB, mind - can afford to actually do. And I think the answer is yes, despite supposed fear of other ATGs "waiting for him to get close", etc.

Very soon after coming under fire, the tanks have a sound contact. They are not ignorant of the direction of fire, or even its likely sources by glancing over the nearby terrain. They don't have to artificially pick the direction due to map orientation, or the range to close to due to map size, or anything of the sort.

You can argue this is unrealistic on the part of CMBB as it stands, but it is the spotting behavior as it stands. You hear the incoming, yes, you also get a good fix on the source and so know how to engage. You don't just get ambushed by completely unknown PAK, with no idea where to close toward.

Would intervening ATGs make the closing tactic suicidal? No. Because intervening ATGs have just as low a life expectancy as the far ones, against the oncoming company. You can afford to fast move closer. If anything opens up at shorter range, you just pause or change your waypoints to "hunt", target the new gun(s) with all available tanks, and resume the advance a minute later.

This is important info for CMBB players, because that is the system we have at the moment. You do not want to mill around at range, or try to silence a mere sound contact with area fire. Nor do you want to charge to point blank, or fire on the move, or spread your formation. The combined firepower of the massed tanks will defeat anything within full spot distance. Fast move will defeat longer range fire, at least if you have numbers over the enemy weapons (if you don't, you've no business attacking, obviously).

To do it right, you only need to add an empirical knowledge of the practical spotting distance to various enemy gun calibers. Which you can judge by (1) sound (2) whether you've spotted them yet already (3) shell effect when they hit, combined with range to sound contacts.

Thus e.g. if a Pz III takes a front hit and "shell broke up" is the result, and you have a sound contact at 600 yards, chances are it is 45mm and you need to "fast move" to 400m or so, then hunt, and your problem will be solved.

I hope this is interesting.

[Redwolf]

I think there are a few aspects to put Jason's example into context (ome of them already mentioned). Even assuming they know they will run into a front of some 88s gun.

1) In real life the tanks couldn't be sure whether there are minefields in front of the PaKs, or infantry with short-range AT weapons.

2) A single Panzer IV/70 or Tiger would throughroughly ruin their day, a real-life tank company commander probably doesn't have enough into to race right up to 1500 meters of that ridge. Same thing for a pillbox with an AT gun. The moment the defending AT shooter is too hard the charging tactics doesn't really work anymore.

3) 13 T-34 is a vast superiority over two 88 FlaK guns, as expensive as they are in CM.

Still, the predictablity of spotting shooting AT guns in CM is too easily exploited. It is worse in CM than in reality because you have more ideas about locations of minefields or other obstacles (remember trenches are always immedeately spotted at 200m nomatter what spotter and spotter status and what terrain the trench is in). A randomization of these hard-threshold spotting aspects in CMBB would be most welcome.

Again: did anyone ever observe an accuracy benefit for defending vehicles that did not move?

[GRUMLIN]

Interesting idea bout the rush approach in CMBB - but it is, I think, based on a too simplistic assumption. Any opponent worth their salt will have armour / snipers / second AT guns hidden ready to engage oncxe you rush in. Your return fire will be relatively inaccurate, and I have seen plenty of instances of my tanks picking off fast moving enemy AFVs from flanks......

defence in depth will, as always, ruin your day.

G

[JasonC]

Laying minefields 2 km from your front isn't very effective. Infantry walk up and clear the mines.

It is 10 T-34s, not 13. Also, there is no "ridge" involved, it is very flat.

You have a "gun?" sound contact in the first minute.

If the target proves too tough (Tigers, say) after you reach 1500m, reverse. You won't lose much more getting out than getting in.

Intervening lesser arms will get overrun at least as easily as the 88s do. This is open steppe, remember.

The terrain is ideal for the defending 88s. In real terrain, folds in the ground or small patches of trees could easily obscure half of the advance.

Doctrinally, artillery was supposed to suppress such targets, infantry to scout ahead for them, tanks to remain in defilade as long as possible and coordinate attacks with other arms, etc. In CMBB, bum's rushing with armor works better than it should.

It is not as bad as it might be - as e.g. the first posts in this thread suggested - but the guns are definitely at a disadvantage. If the ATG-armor match up is less forgiving than with 88s, the problem becomes more acute. Because the lethal distance vs. medium tank armor and the practically deterministic "spotted if fire" range will be much closer together for lesser guns (except against very light armor, I suppose).

[YankeeDog]

Jason,

Would you mind emailing me the Test Scenario you used when you get a chance? I'd like to take a look at what you did, both to replicate what you did and also to try adding some different variables, like putting a couple of HMGs in front of the guns and seeing what happens. It will be easier to compare results if I'm using exactly the same map you were.

To save you the trouble of checking my profile, my email is:

Jugnickle@aol.com

Points about the predictibility are well taken. You definitly shouldn't be able to figure out EXACTLY how close you have to be to spot a gun.

Redwolf - I just ran a quick test found very little evidence of an accuracy bonus for vehicles on the defense that don't move from their setup positions.

I put 2 Marders in a patch of scattered trees. One was just deep enough to see out of the trees from it's setup, the other one I put farther back. On turn 1, I moved the second Marder up next to the first. On turn 2, I move a T-34 out from behind a pactch of woods at the very end of the turn (so that the Marders see it, but don't actually have a chance to get a shot off).

During the next orders phase, I checked the hit% for both Marders. For the non-moving one, it was 22%. For the moving one, it was 20%. The range and aspect to the target were as near identical as I could get them (Approx. 550m range, T-34 moving away from the Marders at about a 45 degree angle).

So the unmoved Marder as a 10% greater chance of hitting on the first shot. *Shrug*

It is possible, though, that the accuracy improves more quickly for unmoved vehicles after the first shot or something like that.

Cheers,

YD

[ January 15, 2003, 08:36 PM: Message edited by: YankeeDog ]

[Scipio]

Originally posted by redwolf:

I think there are a few aspects to put Jason's example into context (ome of them already mentioned). Even assuming they know they will run into a front of some 88s gun.

1) In real life the tanks couldn't be sure whether there are minefields in front of the PaKs, or infantry with short-range AT weapons.

2) A single Panzer IV/70 or Tiger would throughroughly ruin their day, a real-life tank company commander probably doesn't have enough into to race right up to 1500 meters of that ridge. Same thing for a pillbox with an AT gun. The moment the defending AT shooter is too hard the charging tactics doesn't really work anymore.

3) 13 T-34 is a vast superiority over two 88 FlaK guns, as expensive as they are in CM.

Still, the predictablity of spotting shooting AT guns in CM is too easily exploited. It is worse in CM than in reality because you have more ideas about locations of minefields or other obstacles (remember trenches are always immedeately spotted at 200m nomatter what spotter and spotter status and what terrain the trench is in). A randomization of these hard-threshold spotting aspects in CMBB would be most welcome.

Again: did anyone ever observe an accuracy benefit for defending vehicles that did not move?

Again: the BTS crew, especially the programer can give you all the informations. Ask him directly, don't waste your time with 100s of testmatches.

P.S.: Sorry, but I must confess that I'm a little bit annoyed, cause BTS said they would give us a much better and detailed manual for CMBB than for CMBO. It seems they have send me another manual.

[ January 16, 2003, 12:07 AM: Message edited by: Scipio ]

[Olle Petersson]

Originally posted by JasonC:

Laying minefields 2 km from your front isn't very effective.

They can be.

Just check the minefields used in the desert. They could be up to 5km deep. It took engineer units several nights to clear a path, not even attempting to clear the entire field.

Then it's a matter of how you define "front".

- If you equal front to MLR, then there should be MGs and other stuff covering the minefields well ahead of the front.

- If you equal front to the line of forward observation posts, then the heavy ATGs should be quite some distance behind the front.

Cheers

Olle