Casualty Question

[DzrtFox]

Something I've noticed over the course of a ton of battles is that when my infantry/anti-tank squads first take a hit, 95% of the time it is either the leader or the anti-tank soldier who takes the casualty. This seems very unrealistic and makes the battle frustrating, as it's almost always the most important soldier in a squad being taken out.

Now I understand that in real life leaders are very high-priority targets. But it seems unrealistic that almost without fail they are the first one to be hit when in real life measures are taken to protect the leaders.

Is this a bug or a "feature"? Has anyone else noticed this?

[ May 14, 2008, 11:08 AM: Message edited by: DzrtFox ]

[birdstrike]

Probably a psychological issue - at least with me

I notice my important units getting hit all the time, while the enemy only loses "second-rate" personnel and equipment.

[Field Marshal Blücher]

I've had pretty similar experiences. I've played some "not too difficult, but win with minimal casualties" missions as the US. I end up with under 10 casualties but of those 10, at least 5 are Leaders.

[Pandur]

i think HQ squads are easier to spott too. i just finished a QB where i had a big bunch of units in sight from the start. the first 6 spotted units where all HQ´s.

only after that other units got spotted.

[DzrtFox]

It's definitely not psychological. I've played a LOT of battles both online and off, and this happens every time. Any time I see the red cross and click on a squad, either the leader or anti-tank unit is the casualty almost without fail.

I'm going to continue to test this, but I KNOW it's not only in my head.

[DzrtFox]

Also, I should note that this happens even when a squad is hit with a blast area effect like a tank round. In other words, the unit is almost always "sniped", even when hit by a blast. I seriously doubt in real life a squad would put its leader up as "cannon fodder" when coming under fire from a tank or artillery.

[DzrtFox]

OK just tested this out again on Al Huqf (a good test for infantry). Infantry squad 1 moves up to level 1 of a house and takes position. The full squad is in place in the house in cover. All of a sudden they come under fire from a Blue infantry squad. Very first casualty the squad takes is the Leader. I have a screenshot but no way to host images. I'll be happy to email the image to anyone who can post it.

[slug88]

Sorry, but I think it's psychological. Consider that we have 1-to-1 representation, which means that individual bullets (and presumably pieces of shrapnel from explosions) are tracked as well as individual soldiers. So for leaders and specialists to be singled out, they would actually have to program the tac-ai specifically for the purpose of making those units more careless. As with any sort of ai programming, this would be a not at all straightforward task. So, unless they went through a great deal of effort to introduce this one unrealistic annoyance just to spite the player, it's psychological.

Also, for what it's worth, I can recall many instances in which leaders outlast the majority of their squads in my battles. In fact, in my current game, I had a squad get blown up by a T-72. Every man was killed, except for the squad leader, who was injured.

[DzrtFox]

With all due respect, it is not psychological. I have no idea how the game tracks shrapnel and bullets and individual soldiers, but however it's doing it it is inaccurate.

Try a game on Al Huqf. Keep your squads inside buildings. Take note the first time a squad takes a casualty. In my last test a squad of 9 men were all inside a building on one floor and the leader is the ONLY unit that took a casualty. There is no way that in an entire squad of men on one floor of a building the leader could consistently be sniped out of the pack by another squad.

[DzrtFox]

I signed up for an image hosting account just to prove my point. Look at these and tell me again that it's all in my head.

Battle 1, Al Huqf, I'm using Red. Squad enters building and is fully stationary. Blue infantry starts shooting:

Battle 2, Al Huqf, I'm using Blue this time. My infantry squad is running for cover in a building and comes under fire WITH THE UNITS ALL BUNCHED UP. Take a look at the casualties this time:

[slug88]

I've made a simple blue-on-blue scenario to test this. In a little bit I'll report back with my results.

[DzrtFox]

Battle 3, Cavalry Charge at Apache Pass. My infantry squad is HIDING IN A TRENCH and comes under AREA FIRE from a Bradley. Look at the casualty:

Same Battle. Anti-tank squad in a trench further back. One casualty:

[slug88]

Alright, I've run a test in the following way. I've created a flat map, roughly 300m x 300m. Along two opposing ends I have evenly spaced out two identical, opposing Stryker rifle platoons (without their weapons teams or vehicles). Playing in turn based, basic difficulty, I let them shoot at eachother for two minutes. At the end, I count up all casualties from both sides, categorizing them in the following way:

Leaders: squad leaders, team leaders, platoon commander (total 40 men)

Specialists: marksmen, AT, forward observer (total 28 men)

Regular: soldiers, radiomen (total 52 men))

I repeated the test 10 times, which is equivalent to 20 minutes of continous firing. Following are the casualties, where the first number is total men, and the second the percentage from that category:

_____L_____________S___________R

1. 7(17.5%)_____3(10.7%)_____4(7.7%)

2. 8(20.0%)_____2(7.1%)______12(23.1%)

3. 7(17.5%)_____2(7.1%)______12(23.1%)

4. 9(22.5%)_____3(10.7%)_____11(21.2%)

5. 6(15.0%)_____5(17.9%)_____10(19.2%)

6. 7(17.5%)_____4(14.3%)_____9(17.3%)

7. 7(17.5%)_____7(25.0%)_____13(25.0%)

8. 3(7.5&)______1(3.6%)______12(23.1%)

9. 9(22.5%)_____5(17.9%)_____10(19.2%)

10. 9(22.5%)____8(28.6%)_____8(15.4%)

Average:

L - 7.2 +/- 1.7 (18% +/- 4%)

S - 4.0 +/- 2.1 (14.2% +/- 7.7%)

R - 10.1 +/- 2.5 (19.4% +/- 4.8%)

So, from 20 minutes of battle between 120 men with effectively unlimited ammo, we see that regular soldiers were actually *more* likely to die than either leaders or specialists (but of course they are all easily within a standard deviation of eachother). Of course it's not conclusive unless you run it 100 or 1000 times, but I think the burden of proof lies on you in this instance. Anecdotes aren't proof, and neither are screenshots when you cherry pick the ones that support your hypothesis.

For referance, here is the map I used for the test:

http://hep-www.colorado.edu/~oleinik/casualtyTest.btt

[ May 13, 2008, 01:35 PM: Message edited by: slug88 ]

[DzrtFox]

I haven't cherry-picked anything. The screenshots I posted are 4 instances that occurred TODAY. It happened 4 out of 4 times. I didn't go back and replay these scenarios a bunch of times to get the results I wanted.

These are results I've noticed over the course of a lot of games, and when today I went and tested them they happened WITHOUT FAIL.

I have no agenda here other than to have a game that works properly after almost a year of waiting. I'm going to continue to test this but I find it very hard to believe that I am the only one experiencing this and that it is just coincidence.

[slug88]

I'm sure you don't have an agenda here, and neither do I. In fact when I ran the first test I noticed that leader casualties were almost double soldier casualties, and I was thinking that you may be completely right. But my results speak for themselves, from 20 minutes of combat soldiers are actually more likely to die. And for each screenshot you post like the above, I could easily post a counter screen. And I bet I could get all the ones I need from a single scenario, too. And I can also recount various instances in which AT specialists and leaders outlasted the rest of the squad. I have seen it happen many times.

My point is, if your experience leads you to believe this is really an issue, the only way to convince anyone else is with statistics. Hard data. 4 out of 4 times is not statistically significant, and doesn't even warrant mentioning.

[handihoc]

Interesting discussion.

I haven't fully tested this like you guys (and I appreciate the time you've put in to do this) but I HAVE noticed that generally it's the team leader and/or AT guy who gets it first.

I'd found this quite acceptable, if annoying, putting it down to the pure logic of the enemy sniping for the most logical/important target in a team.

However, IF, as Dzrt Fox says, this also applies to tank and artillery blasts, then we DEFINITELY have a problem that needs to be remedied, as it's obviously thoroughly unrealistic.

It would be great to have a comment by BFC on this, and possibly more testing for verifiable results.

Anyone else tested this?

[DzrtFox]

Also slug, your results are not really refuting my point, which is that Leaders and Specialty units are almost ALWAYS the first to die. Over the course of 20 minutes of continuous firing it is entirely possible that casualties will level out. But I'm telling you that almost every time a squad takes a hit the first unit to die is the leader or anti-tank soldier.

I would suggest going back and re-testing with your scenario, but this time check each squad the first time it takes a casualty. Those are the numbers I'm interested in.

[slug88]

Read how I performed the test again. It wasn't 20 minutes of a single battle, it was 10 seperate engagements lasting two minutes each. In each engagement, there were a couple squads that had not taken casualties at all, and most had taken between one and three. So I believe it does refute your argument. However, again, I encourage you to perform your own test, for there may very well have been errors in my methodology. I am open to the possibility of you being correct; for instance, perhaps leaders and specialists as more likely to be standing or crouching when the rest are prone, and I could see this as being a programmed behavior that would lead to what you describe. However, I simply haven't seen any evidence to support it.

[Clavicula_Nox]

I just did Al Huqf as Blue. 3 WIA. One was a rifleman who was wounded by an RPG blast, the other 2 were from the Bradley crew who were shot while evacuating their vehicle.

I've seen what you're talking about happen, and it's quite frustrating. Maybe the TacAI has some coding to give priority target to the guy who is leading the formation, shouting orders, and carrying the biggest weapon?

*edit*

I will say this, I have an M203 gunner who just earned himself a Bronze Star with V device. He took out that BMP with a single grenade, I even saved it to video. I will upload to Youtube, barring any problems.

[slug88]

Originally posted by DzrtFox:

I would suggest going back and re-testing with your scenario, but this time check each squad the first time it takes a casualty. Those are the numbers I'm interested in.

I suppose it wouldn't hurt just to check, so I will run a few tests as you describe above. I will report shortly.

[slug88]

Alright, I've run a new test. Two opposing Stryker rifle squads face eachother at 300m. I play in realtime, basic difficulty, until the first casualty (red base) sustained by either side. Repeating the test 20 times, I found:

Leaders died first: 8 times

Specialists died first: 1 time

Soldiers died first: 11 times

The squads consisted of 3 leaders, 4 soldiers, and 2 specialists each. When weighing each type by by their frequency in the squad, a soldier is 3% more likely to die *before* a leader, and 450% more likely to die *before* a specialist. Ideally, if there's no bias, then these weighted percentages should each be 0%. Clearly 20 times is not enough for a conclusive analysis. But, again, the burden of proof is on your shoulders, since you claim there is a problem. Here is the scenario I created to run the tests:

http://hep-www.colorado.edu/~oleinik/casualtyTest2.btt

[DzrtFox]

Here's my latest test. I fired up a Hotseat battle on Al Huqf, positioned two opposing squads in buildings on opposite ends of a street and ordered each to fire at the other.

First casualty for red:

Next casualty for red:

Third and final casualty for red:

Blue took no casualties. So are you telling me that it is realistic for an infantry squad to be able to snipe two leaders and an anti-tank unit INSIDE A BUILDING while taking return fire?

[slug88]

DzrtFox, you have 3 data points supporting your argument, while I have 20 refuting it. That's if you discount the first test, which contains far more, and which also contradicted your claim.

Also, note where those casualties took place. Each man was right next to a window. That's the discriminating factor, not the soldier type.

[DzrtFox]

OK I'm not going to waste any more time on this. This has happened every time I've tested it today in an actual battle. If this isn't enough evidence to prove that there is a problem somewhere then there is nothing I can do that is going to convince you.

I do appreciate you taking the time to perform these tests, but this has happened EVERY TIME I've tested it today. It's happened enough over time that I've noticed it on my own and it's happened without fail today.

[DzrtFox]

Also slug, I see where the casualties took place (near the window), but we don't have control over individual soldier placement. I would assume in a real-life situation the riflemen would be in the windows with the leader hanging back.

No matter how you look at it, this is a problem. Maybe not for you, but it is a huge problem for me.