Casualty Question

[slug88]

DzrtFox, the problem is that your tests are still statistically insignificant. When I ran my most recent test, leaders died the first four times. Had I not run it another 16 times, I would have been convinced that leaders do die first. The point is that four times in a row, or six times in a row, or even 8 times in a row is all well within the realm of plausible coincidence. I have counted, and you have a total of 8 men dying in all the screenshots you've posted. In all my tests so far 233 men have died. So if we averaged our data, it would still indicate that there is no intrinsic bias.

You also have previous experience of this from months of playing, but the undeniable reality is that the human mind does not treat all data equally. If you play a battalion sized battle in which three different platoons lose their AT man first, it will stick out in your mind, especially if you've got enemy armor to worry about. On the other hand, how significant would it be to you if three more, or 6 more platoons lost riflemen first in the same battle?

Now, I do not at all discount the possibility that other factors may bias leader and specialist deaths in complex situations. For example, it's possible that heavily-laden RPG soldiers run slower than their squadmates, and so are exposed to fire for longer periods. Or perhaps leaders begin moving to their waypoint before the rest of the squad, and so by getting there first, they're more likely to die first. Mainly what I am contesting is that the game biases them artificially, which is what is implied when you state that HE and artillery is more likely to kill leaders than soldiers.

[DzrtFox]

The only difference I can see between your setup and what I'm doing is that I am playing on Veteran difficulty and you are playing on Basic. On Veteran spotting is more difficult which also makes it more unrealistic that this is happening.

Maybe try your tests again on Veteran?

[DzrtFox]

I'm going to leave it at this...

I'm in a PBEM game right now with a friend. I have maybe 13 or so squads and units around the map. Of the five squads that have taken casualties (all from Bradley Area Fire), ALL FIVE of them are missing either the leader or the AT unit and nothing else. If that combined with what I've shown you earlier doesn't convince you that somewhere there is a problem, then like I said before nothing will.

[Pandur]

maybe that is specific to blue on red!? slug did you test blue on blue only!?

i played both sides much lately, and i never "really" had that problem on blue side. just on red. however i didnt ran tests myself.

[slug88]

I suppose it could be a red issue. As I specified, my tests involved US Stryker infantry. Also, I doubt that difficulty level would affect it, since it the tests were on a flat, featureless, 300m long map, so spotting is not much of an issue. I played on basic in order to track the casualties of the opposing side as well as my own.

Dzrtfox, the latest figure you've given doesn't convince me that there is a problem, but it does convince me that further testing is necessary, for red squads. My suspicion, however, is that your results, again, are a result of complex factors and that the biasing could very well be realistic in a way. Remember, I've been arguin against the presence of an artificial bias.

[ May 13, 2008, 04:32 PM: Message edited by: slug88 ]

[DzrtFox]

All I'm asking is that you try it on Hotseat mode on Veteran difficulty. Maybe the problem is somewhere in the spotting model.

[DzrtFox]

Also it is possible (although I don't see how) that it is a Red issue, since every time I've noticed the problem I think I have been playing Red.

Although I tend to play Red way more often for the challenge...

[slug88]

Ah, hotseat is a good idea, hadn't thought of it. Will do, though not now because I just got off work and I'm in the mood for some actual playing .

[DzrtFox]

Slug once we get to the bottom of this you and I need to play a PBEM game. I like the way you think.

[DzrtFox]

OK, I had time to sneak in a battle tonight. I set up a Quick Battle tiny Syrian vs. US Meeting Engagement. I took Red and received an infantry company.

As I'm advancing my squads, I come under fire from two Abrams tanks across the map. Here are my first casualties of the battle:

In fairness, the first shell that hit actually killed a soldier, but the second shell that hit miraculously killed both the Leader and Team Leader in one shot while leaving the rest of the squad unscathed.

Directly behind that squad is an HQ unit who also is hit by the shrapnel:

On their right flank, in a trench (or ditch) a unit is hit from the same indirect fire from an Abrams:

These are the only casulties I took the entire battle. Can you not see how something is wrong here?

[the Fighting Seabee]

It's funny this topic came up. Just last week I was wondering about this because I had SEVERAL instances where the Leaders were the LAST to die. And it seemed even more evident that the enemy leaders were extremely tough to kill. I had one enemy battalion commander becoming an army of one. He was under fire for several minutes before he was injured. Another couple of minutes of firing and he was finally dead.

[DzrtFox]

Interesting.

Either way, it seems like there is some sort of artificial bias going on, and it needs a little bit of tweaking.

[Kwazydog]

Originally posted by DzrtFox:

Either way, it seems like there is some sort of artificial bias going on, and it needs a little bit of tweaking.

Just for the record guys Ive run this past Charles and there isnt any sort of code in the game that would cause team leaders to be hit with more frequency than regular solider. Each soldier is treated as an individual.

Dan

[ May 14, 2008, 07:12 PM: Message edited by: KwazyDog ]

[DzrtFox]

So is the choice of which unit takes the casualty a completely random function? How is it possible that this is happening? Has anyone else tried this out and/or experienced the same thing?

[Clavicula_Nox]

Next time you run a test, see where the soldiers are positioned. I'm curious to see if leaders are generally out of cover, or what.

[Battlefront.com]

DzrtFox,

So is the choice of which unit takes the casualty a completely random function?

Yes, in the sense that there is no equation in the code to assess who should become a casualty based on who they are in the game. No in the sense that casualties are taken due to the proximity of the source of the enemy fire. So the guy that is the furthest away from a shell hit is the least likely to become a casualty. That could be anybody since there is no formulaic way for shots to impact nor for soldiers to be assigned to a specific spot. So it is random in the sense that casualties are caused based on being at the wrong place at the wrong time.

How is it possible that this is happening? Has anyone else tried this out and/or experienced the same thing?

To see if there is anything wrong (though we can't imagine what that might be) one would have to run a rather huge number of tests in a variety of different circumstances, the do a statistical comparison which neutralizes perception bias.

For example, a 9 man squad has three leaders, which means that statistically there's about a 1 in 3 chance of a leader becoming a casualty. That's rather high odds in any one particular circumstance. You could easily run a test three times and see nothing but leader casualties each time. So instead, run the test 100 times and then count up how many times each of the 9 guys became a casualty. If 100 samples is enough (and I don't know that it is. Some statistics guy would have to chime in here) you would expect each of the 9 positions in that squad to have been hit roughly the same number of times.

Steve

[Battlefront.com]

Whoops... I missed Slug88's post at the top of this thread. Looks like someone has already done what I suggested and shown that there isn't a sign of a problem.

He is correct, though, that there are circumstances that can make a leader more prone to be hit than another specific soldier type. But it isn't because he's a leader per se, but rather because he's the first one to move (generally) or one of the lighter soldiers (therefore first on the scene) or any number of other factors. That's why I said in my previous post that different tests need to be done before anything more-or-less definitive can be concluded. Personally, I think there are more productive things to do with one's time

Steve

[Battlefront.com]

I know the discussion has been about HE type fire so far, but to remind people... small arms fire is vector based between shooter and where the rounds strike. If the vector intersects a soldier, then it hits, otherwise it misses. So there is no randomness with small arms fire either. Soldiers who fall victim to small arms fire were indeed hit by it, not randomly selected to become a casualty like CMx1 games.

Steve

[Caesar]

Could the HE be hitting the leaders more because the HE is aimed at the centre of the target and that is where the leader typically is. I guess the same could be said for small arms fire, that it would be typically be aimed at the highest density of troops. When troops are on the move the leader moves first and is in the front and is therefore more likely to be hit. I guess if this is the cause (assuming the problem exists of course) then the AI could conceivably be altered to get the leader to hold back some and not located at the centre so much.

[Clavicula_Nox]

No way, the infantry motto is "Follow me!" not "Follow that other guy I put in front of me!"

[Field Marshal Blücher]

Well, according to one book I have, in standard formations the squad leader is in fact in the middle, between his two fire teams . . . has this changed at all?

-FMB

[Battlefront.com]

Caesar,

Could the HE be hitting the leaders more because the HE is aimed at the centre of the target and that is where the leader typically is

Outside factors like this, which are themselves realistic, may make a tiny bit of a difference. However, I doubt it would be noticeable. Remember that just because an HE gunner aims "center mass" doesn't mean that's where he hits, either because of blockage or an inaccurate shot. In fact, the whole point of HE is to get "close enough".

The relative position of soldiers within an Action Spot is also variable, not fixed. There is, however, a default order. Real life has that too, though I'm sure ours doesn't always reflect that.

Steve

[Secondbrooks]

Field Marshal Blücher: Acording statics i've seen from ww2 it seems that NCOs and junior officers are prone to have increased casuality rate. This is because of the fact that they had to (and were trained to) lead by showing exmaple for their men.

And many times there is no time to give orders. In our soldier's hand book there are many referings to: "if you don't know what to do in situation X, look at your leader and do as he does" (this rule is golden)

I think this is atleast partially what Clavicula_Nox was after.

[ May 15, 2008, 01:02 PM: Message edited by: Secondbrooks ]

[Clavicula_Nox]

I think this is atleast partially what Clavicula_Nox was after.

Absolutely.

[Field Marshal Blücher]

Originally posted by Secondbrooks:

Field Marshal Blücher: Acording statics i've seen from ww2 it seems that NCOs and junior officers are prone to have increased casuality rate. This is because of the fact that they had to (and were trained to) lead by showing exmaple for their men.

And many times there is no time to give orders. In our soldier's hand book there are many referings to: "if you don't know what to do in situation X, look at your leader and do as he does" (this rule is golden)

I think this is atleast partially what Clavicula_Nox was after.

Makes sense.