Where is the realism?

[Michael Emrys]

Originally posted by Leutnant Hortlund:

But for the umpteenth time, the gunner in the T-34 can see the StuG all the time. According to the game engine the T-34 and the StuG see eachother. You then have a veteran gunner with line of sight to a tank (or tank sized shadow if you will)...

But he may not. As has been pointed out to you already, he may only have sighted the muzzle flashes of of the StuG (or whatever it is). It may be difficult to impossible for him to estimate where the center of mass is of the AFV is from only that clue.

Michael

[Andreas]

Or which direction it is driving and how fast. Or if is moving.

[Pak_43]

It seems quite clear to me that BFC have modelled the maximum % hit chance at 80% - regardless of conditions, missing 4 times is therefore well within reasonable probability....

Old BFC forum

Short of posting our equations (which I don't want to do) I'm not sure how to answer this question.

quote:

--------------------------------------------------

Heinz: at 300 meters would you say the odds of a hit on the third round if the first to miss, should be very close to %100?

--------------------------------------------------

You didn't direct this to me, but I'll answer anyway: No. (80%? Yes. 100%? In the thick of combat? No.)

[YankeeDog]

OK, let's step back and take a look at the way CM's hit model for AT shots work.

AS we all know, CM's anti-armor hit and penetration algorithms are incredibly complicated, and highly detailed, blah, blah, blah.

Fundamentally, though, for a given specific situation, the "to hit" calculation boils down to a percentage chance. In this way, the game is no different from the old pen and paper strategy games like ASL. The calculations it uses to determine the "to hit" chance may be much more complicated (and hopefully realistic), but in the end the program still "rolls the dice" to figure out whether or not the shot hits.

Now CM rightly models the fact that following shots generally have a higher chance of hitting than the first shot because the gunner can use fall of shot to adjust range and bearing. However, in CM as in life, the "to hit%" NEVER reaches 100%. I have had units show hit% very close to 100% - a zook team 15m from a tank with a 98% hit chance on it's second shot, etc.

This means that there is ALWAYS a chance that the next shot will miss, no matter how many shots have been taken previously. Unless you are prepared to argue that CM should make the hit chance 100% after a certain number of shots, then there will always be the chance that you will "win the lottery" and watch a tank or gun miss a 'sure shot' multiple times.

So we can have some hard data to deal with, I ran a quick test. I set up 5 Vet T-34/85s, each exactly 58m from a Stug III at night in snow. The T-34s have perfect flank shots on the Stugs.

The LOS tool lists the 'to hit' chance for the T-34s at between 55 and 58% for the first shot. (Evidently, there was enough minor variation in my setup to create slight difference in hit chance). For reasons that will become very clear when you look at the test results, I was not able to get a hit% for following shots.

In 4 runs of the scenario (i.e., 20 individual 'tests'), Not once did a T-34 fail to hit at least once in four shots. Once, a T-34 took until the fourth shot to hit (that T-34 KOed the Stug on the first hit). One T-34 took 5 shots to actually KO the Stug, but had already hit on shots 2 and 4 (second shot caused one casualty, and the fourth shot ricocheted off the front hull armor).

Two StuGs survived turn 1, but all of the matchups were 'resolved' by the end of turn 1.

In one StuG survival, the T-34 missed its first shot but hit on the second shot. Unfortunately for the T-34, the shell ricocheted off the front hull armor (the StuG was rotated far enough at this point to give a chance of a frontal hit). At this point, the T-34 got cold feet and decided to back out of LOS as the Stug brought it's gun to bear. This was the only T-34 that decided to break off engagement.

The other surviving Stug was also missed on the first shot, and then took one side penetration (no damage), and one front partial penetration (no damage) before finally getting a shot off and KOing the T-34 in one shot. This was the one really lucky (or unlucky) result - I would say the chance of a Stug taking an 85mm side penetration and an 85mm partial front penetration and not suffering at least one crew casualty should be very small. Again, though, it might happen once in a blue moon.

Make of these results what you will. In any event, I think it is clear that 4 misses at 58m, even at night in snow, is a very rare occurance, as it should be. I certainly don't agree that is should never happen, though. If someone is willing to run a test with a larger sample and show that the T-34 will miss 4 shots, say, 7% of the time (statistically quite possible given the small sample size of my test), I would then agree that this percentage is too high, and that hit algorithms should be adjusted. Otherwise, I'm just going to chalk it up to colossally bad luck.

IMHO, if you want a game where luck doesn't play a role, go play chess. I rather enjoy the sweet frustration of the winds of war.

Final disclamer: I have no first-hand knowledge about how the hit calculations in CM actually work. The above assertions are simply made on the basis of what I have read here on the forum, and what I have experienced playing the game.

Cheers,

YD

[Hortlund]

Originally posted by Michael emrys:

</font><blockquote>quote:</font><hr />Originally posted by Leutnant Hortlund:

But for the umpteenth time, the gunner in the T-34 can see the StuG all the time. According to the game engine the T-34 and the StuG see eachother. You then have a veteran gunner with line of sight to a tank (or tank sized shadow if you will)...

But he may not. As has been pointed out to you already, he may only have sighted the muzzle flashes of of the StuG (or whatever it is). It may be difficult to impossible for him to estimate where the center of mass is of the AFV is from only that clue.

Michael</font>

[YankeeDog]

Originally posted by Pak_43:

[QB]It seems quite clear to me that BFC have modelled the maximum % hit chance at 80% - regardless of conditions, missing 4 times is therefore well within reasonable probability....

Maybe for a certain weapon under certain conditions, but 80% is definitely not the upper limit of hit percentage for all weapons at all times. I have seen hit chances higher than this many times.

Cheers,

YD

[Hat Trick]

From Brightblade

The hue of the LOS tool was quite dark, the hit chance for the next shot is 79%.

For what its worth, if all four shots had a 79 percent chance of hitting (admittedly, a big "if"), the odds of missing on all four shots are about 0.2 percent, or one in 500. Sounds like a run of bad luck to me.

[YankeeDog]

Oh, I forgot to mention the one thing that I did find quite unrealistic about my tests: Not once did the StuG try to break off the engagement.

The StuG is at a clear disadvantage. In 20 runs, it survived twice. All it has to do to break off the engagement is hit the gas and drive forward at top speed - since the contact is on the very edge of LOS conditions, it doesn't even need to turn. If it does this right away (i.e., as soon as it sees the T-34), the T-34 doesn't even get a second shot off (I tested this by actually giving the StuG a FAST order straight ahead).

Other AFVs in similarly low-survival chance engagements (see any of the recent threads on IS-2s and ISU series) will often get of of LOS before taking a single shot even if they are facing the enemy AFV and could get at least the round in the chamber off before retreating.

For some reason, the StuG crews seem to be quite courageous - they usually have to weather 2 flank and 2 frontal shots (all with a hit chance of over 50%, and all with at least a decent chance of KOing the StuG) before even getting a shot off.

Perhaps the TacAI can't figure out that it can get out of LOS by moving tangent to the contact and so stands and fights as it's only possibility of survival?

[ January 16, 2003, 02:11 PM: Message edited by: YankeeDog ]

[xerxes]

I was going to post on this earlier, but I thought it would be a waste.

Just because you (player) "see" the tank doesn't mean it is as clear as day. You can "see" tanks as something other than what they are. Sighting is not a binary thing EVEN though the user interface shows it that way. There are degrees of how clearly you see the target. The only way to determine this is to use the LOS tool. The darker the line, the "fuzzier" your tankers view of the target is. Imagine if you will looking through 25m of trees at a tank beyond the trees. You will NEVER see the tank clearly. This is why your to-hit chance will be lower than if there was a straight clear LOS. The same effect occurs at night, snow, or fog to varying degrees. Your tank was probably aiming at a "blobby" tank that was just at the limit of your tanks ability to see.

This is my understanding of the CM to-hit model. I believe it's quite accurate based on playing 5 gagzillion games.

On a side note, it would be really nice in the new engine for the "clarity" of the target to be graphically depicted. Of course, in CMBB if you turn on the full weather, fog, trees, etc. and get down to the point of view of your tank it would probably be pretty clear how "clear" your tankers view is.

[gibsonm]

Originally posted by Sardaukar:

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

BTW, why didn´t the Russians have Bazookas? The USA sent virtually anything to them, food, boots, telephone wire, HTs, tanks, aircrafts, why not Bazookas?

<snip>. Russians made some use for captured Pzfausts, though. They could probably be modelled in game...who knows, they may be in next patch

Cheers,

M.S.</font>

[Brightblade]

Originally posted by Xerxes:

Just because you (player) "see" the tank doesn't mean it is as clear as day. You can "see" tanks as something other than what they are. Sighting is not a binary thing EVEN though the user interface shows it that way. There are degrees of how clearly you see the target. The only way to determine this is to use the LOS tool. The darker the line, the "fuzzier" your tankers view of the target is. Imagine if you will looking through 25m of trees at a tank beyond the trees. You will NEVER see the tank clearly. This is why your to-hit chance will be lower than if there was a straight clear LOS. The same effect occurs at night, snow, or fog to varying degrees. Your tank was probably aiming at a "blobby" tank that was just at the limit of your tanks ability to see.

You are right, in reality a target will be obscured by whatever in almost any cases.

However that will in reality influence how easy or how difficult it is to spot the target, not how easy it is to hit. Once spotted, the target may hide behind brushes or scattered trees as much as it wants, as long as it isn´t moving, it will be hit just as if it stands on open ground. It´s a bit different with shell blocking obstacles like houses, hills, thick walls etc. Then the target is actually smaller and therefore more difficult to hit.

Once you spotted some part of a tank, you can deduce from this where the rest has to be, so you will aim in the middle of some brushes or right through a hedge where you expect a good point to aim and although you don´t see the specific part of the target where you aim at, you will hit it very likely.

If the target or shooter is moving however, keeping track of the target even through only visual obstacles can be hard.

[ January 16, 2003, 03:19 PM: Message edited by: Brightblade ]

[Seahawk-vfa201]

as xerxes pointed out LOS is not a binary variable: "either you have it or you have it not". In between there are all possible variation hence you might have a LOS on the tank - and see it perfectly on the map with your eyes naked - but the quality of the LOS could be damned close to the extreme "have-it-not" threshold.

BTW: yoy would see it perfectly on the map with your eyes even if dressed: no need to test it while naked!

When CM will introduce bmp for different level of visibility I bet with a very DARK LOS even the player on the map (either naked or dressed) will see the sh!t of before and ask himself: who the heck - or what - my unit is targerting: I can only see sh!t there..

Second: a veteran unit (or a crack) are not there in order to do miracles: they *react* to situation faster. That is, in a situation where you see sh!t they get to sh!t much faster than a regular unit. They still get sh!t though: only faster (maybe slightly better sh!t but not that much better)

Well, have been talking about ex-edible stuff for too long. Have to run now: the sh!t hit the fan!

[Seahawk-vfa201]

Originally posted by Brightblade:

Once you spotted some part of a tank, you can deduce from this where the rest has to be, so you will aim in the middle of some brushes or right through a hedge where you expect a good point to aim and although you don´t see the specific part of the target where you aim at, you will hit it very likely.

If the target or shooter is moving however, keeping track of the target even through only visual obstacles can be hard.[/QB]

This is correct if you also could identify which part of the tak you have spotted: the front or the back. So you aim to the left (ideally getting into the middle of the tank) or to the right? - and risk to hit sh!t again (sorry guys, must be some of those moments)

Imagine the situation: "JEez, I think I spotted a tank there - WHERE, WHERE!!?!??! There sh!t (recurring item it seems). Fire , Fire, Fire there."

Ops missed. Try a bit left: FIRE FIRE. definitely somethign is there.

OOOpps missed again. SH!T aim to the right then FIRE FIRE FIRE.

ARRGGGGHH: we hit SH!T

FIIIIREEEEEE.....

In those conditions I agree though that missing 4 in a row is truly bad luck but I would not be surprised at all if it was just 3!

And if in your games (I played thousands) it happens once in a while again I would not be that surprised (in extremely bad visibility).

Last comment on LOS and having a tank silhouette on the map. No one ever noticed cases where a still tank fires and WOW it pops into view, few secs later turns into a target. then Fires again and *POP*, visible again, lasts few secs then marker again.

Now how do you explain this *rather common* behavior? LOS might change during the second and even if your tank has a RED aiming line to a target that does not means it has a LOS: it migh have just lost it (it happens and I verified and saw this many times).

From what I see LOS calculation has many factor into play. (including the famous sh!t)

[Moon]

This isn't necessarily correct, Brightblade, because under such circumstances it will also be more difficult to observe fall of shot. So you fire at the brush... then what? Did you hit? Did you totally miss? How do you adjust?

Night and snow is also different from LOS obscured through brush and other obstacles, because night and snow make it more difficult to estimate ranges. Coupled with the difficulties to adjust aim properly, and one can see that even an experienced crew can have difficulties hitting. This doesn't mean that they will miss 4 shots in a row everytime, but it isn't a "first shot, first kill" situation either, despite the close range.

And then one last point - in darkness, with snowfall, you might be able to just barely make out a target with the nake eye, but "lose" it through your gun sight. This could be a real life explanation for *just* having LOS to a target near the max visibility range, but having a low "to hit chance".

Martin

[Seahawk-vfa201]

Originally posted by Moon:

This isn't necessarily correct, Brightblade, because under such circumstances it will also be more difficult to observe fall of shot. So you fire at the brush... then what? Did you hit? Did you totally miss? How do you adjust?

[...]

Martin

Exactly: you might know you missed it (no rewarding BANG and metal screaming sound) but where? left, right, high, low? The second shot does not have higher chances than the first as the amount of information is practically the same. The SA (situation Awareness) is still below your ankles.

[SgtMuhammed]

This was a conversation between one of our T-80 gunners (here in OPFOR land) and his TC.

TC: Mike 1 (M1) 1 o'clock in the far woodline.

G: I can't see him.

TC: He's right there.

G: Where?

TC: Climb up here.

G (from loaders hatch): Oh there he is.

G (back at sight): I still can't see him.

TC then kicks gunner in back of head.

The point. Yes it is possible to see a tank with your eyes but not be able to see it with you optics. The above example was during the day, imagine at night.

Stuff happens and occasionally luck rules the battlefield.

[Brightblade]

Martin,

you are right of course, what I said is not necessarily true. I only thought of targets close behind or in a visual obstacle. Then it´s not too hard to adjust if you didn´t hit, after all you got the tracer in the round, an explosive filler in many cases which makes it easier to spot the impact etc. I admit it´s probably still not exactly as easy to hit as if the target stood on open ground, but it´s not much more difficult.

If the target is farther away from the obstacle, it might be harder to adjust the next shot as the impact might be obscured by the obstacle.

Night and snow are a different thing. I know that it´s quite hard to see anything at all under such conditions. As I said before, when I started this thread I assumed that atmosheric conditions had only a minor influence on the hit chances, if any at all. I learned that this assumption was wrong and it´s fine that way, being more realistic.

[ January 16, 2003, 04:41 PM: Message edited by: Brightblade ]

[Bruno Weiss]

Sardaukar wrote:

Russians made some use for captured Pzfausts, though. They could probably be modelled in game...who knows, they may be in next patch

Panzerfauskies?

[Hortlund]

Originally posted by Sgtgoody:

This was a conversation between one of our T-80 gunners (here in OPFOR land) and his TC.

That story reminds me of one I heard from an Israeli TC during the Yom Kippur war. At the end of the first day in the Golan heights, everything was chaos with no real frontline to speak of, rather lots and lots of small unit actions, ambushes, hedgehog defences etc.

Anyway, this lone Israeli tank (cant remember what type) was the sole survivor of its company, and it was parked in a small depression behind a ridgeline. The TC had left the tank to go over to the ridgeline and see what was on the other side. Just as he was halfway to the ridgeline, three Syrian T-62 crested the ridgeline, maybe 30 meters ahead.

The TC ran back to his tank screaming for the gunner to shoot. Meanwhile in the tank, the crew had been relaxing, and was caught completely off guard. The gunner scrambled to his sights as the TC jumped into position. "Shoot" "Shoot" the TC screamed as the T-62s were 30 meters away and they had spotted the Israeli tank. "I cant see anything, all I see is brown" the gunner replied (he had the sights on maximum magnification). The TC kicked him in the head, and the gunner took the hint and just pressed the trigger. The TC aimed the gun by looking at the gun barrel and kicking his gunner on the left or right shoulder to traverse the gun.

Amazingly enough, the Israelis managed to take out those three Syrian T-62s and the crew survived the war.

Anyway, I think it's a good story.

[ January 16, 2003, 07:00 PM: Message edited by: Leutnant Hortlund ]

[Brightblade]

Just if anyone cares: The T-34 fired its fifth shot and actually hit! This one hit was enough to eliminate the assault gun, probably a StuH because it didn´t fire again and a StuG would have reloaded much faster.

[Rother]

Originally posted by Sgtgoody:

This was a conversation between one of our T-80 gunners (here in OPFOR land) and his TC.

It's been over 10 years since my last Hohenfels. What are you using for T80s and BMPs? At NTC we used M551s for both. Their sights became totally useless well before it got fully dark. Any night gunnery involved blindly spraying coax until you flashed the target's strobe and then flipping the switch to main gun.

[SgtMuhammed]

We use M60s as T80s and M113s with a hand made gun for BMPs. The 60 thermal sites are really nice but all we have in the BMPs are little Bushnel 3x scopes and nods for night vision. We have to use to coax methode as well although we have started using PAQ4 laser aiming devices. We still can't engage more than about 150m at night though. Still once we do close the range its all over for BLUEFOR.

[gunnergoz]

Originally posted by Leutnant Hortlund:

Heck at 58 meters you can look through the barrel and aim the gun that way and hit.

OK, try this: Go out on a cold and snowy night with MAXIMUM visibility of 60 meters and see if, by looking down a 12-foot long shipping tube, you can find the ass end of a greyhound bus. Somehow, I doubt it.

I think the results speak well of the simulation of a nighttime winter engagement.

But then, I've been reading war memoirs and military history books for 40 some years...so can I help it if my imagination is a bit broader than most?