need glasses for the sharpshooters ....

[siffo998]

http://www.2shared.com/file/qsQDCUYM/test_lane.html

here is a upload from my test lane (it should be equipped with the last setup from my test... you can use it as you want to)

you can upload your map too if you want... but i will not share my e-mail sorry.

[JG11Preusse]

http://www.2shared.com/file/qsQDCUYM/test_lane.html

here is a upload from my test lane (it should be equipped with the last setup from my test... you can use it as you want to)

you can upload your map too if you want... but i will not share my e-mail sorry.

its no problem taht you wont share you emailadress.

i hit the link, but looks like i cant have access from here, where i am. some links dont work, facebook, youtoube and some others also wont work, but if you agree i can give you my emailadress and you send it there ?

you maybe also can create a 1 time emailadress like webmail or google, yahoo or others.

for repository i use the same logindata i use to buy the game.

p.s. with generalizing i meant that i talked about 3 or 4 situations and you wrote : "sorry preusse but i cannot reproduce your "my sniper hits nothing against infantry" experience"

sounds like generalizing or maybe i didnt understand enought due to language lacks

[Bastables]

Sorry typo regt number is E1002850, was seeing red and did not note it was a 6 not a 5.

[Magpie_Oz]

Sorry typo regt number is E1002850, was seeing red and did not note it was a 6 not a 5.

I had noticed that, was about to get you to run around the parade ground 5000 times shouting at the top of your lungs " E 1 0 0 2 8 5 0 SAH !"

[JG11Preusse]

I had noticed that, was about to get you to run around the parade ground 5000 times shouting at the top of your lungs " E 1 0 0 2 8 5 0 SAH !"

aha ... sounds good and im sure it has to do something with the topic itself, but i only dont understand

[Magpie_Oz]

Of course it has nothing to do with the topic CMBN threads are nothing if not tangential.

[user38]

Still running tests with this, but I am running into extreme variability in outcomes. Playing real time, I've seen 3 TCs hit with the first shot in one run and in another run all 10 TCs untouched after 15 minutes (all of sniper's ammo expended). In another run, no TCs were hit for over 5 minutes, then all 10 were hit in the following 10 minutes.

I also conducted a few runs of this test range using turn-based play (all my previous tests were real time), and I am yet to see a single TC hit at 100m. This hints there may be a divergence between real-time and turn-based outcomes, but as I note above, I have had one instance in real time play where there were also no hits at all, so I can't really draw that conclusion yet. I also did one turn-based run at 50m instead of 100m and did see hits on TCs, so there is no categorical "impossible" to hit here either.

Here are my 100m and 50m test ranges:

http://dl.dropbox.com/u/36437512/800m_10lanes_snipertest_elite_panthers_100m.btt

http://dl.dropbox.com/u/36437512/800m_10lanes_snipertest_elite_panthers_50m.btt

akd, thanks for posting the test ranges.

I did some testing on your ranges. I gave the tanks short cover arcs and turned them 90 degrees to face the snipers. I used v1.01 on a PC. Here are my results:

Real time - 100 metres

35 minutes - no tank commander casualties (reset after 15 minutes as the snipers run out of ammo)

Turn based - 100 metres

40 turns - no tank commander casualties (reset at turn 15 as the snipers run out of ammo)

Real time - 50 metres

5 minutes - tank commander casualties 9, 8, 8, 7, 8 (reset each minute as the tanks run out of tank commanders)

Turn based - 50 metres

10 turns - tank commander casualties 6, 5, 1, 4, 4, 6, 4, 5, 4, 4 (reset after 1 turn)

Analysing the data

50 metres

The mean tank commander hits after one minute of play was 4.3 in turn based and 8 in real time play. The standard deviation for the real time data was .63 so on this test one would expect 7 to 9 tank commander casualties 95% of the time.

The standard deviation in turn based play was 1.35. On my data the single hit on my third test falls 1.7 standard deviations under the mean (which gives it a 5% probability of occurring). If we recalculate the mean and standard deviation without this data point we get a mean of 4.7 and a standard deviation of .81. As such, on this test on would expect either 2 to 7 or 3 to 6 tank commander casualties 95% of the time (depending on whether we include or exclude the 1 hit data point).

The above analysis assumes a normal distribution of results (and why wouldn't we). On this basis there is a statistically significant variation between results in real time play and turn based play. It gives one pause.

100 metres

On my data elite snipers in this game (v1.01 on a PC) cannot hit a head sized target at 100 metres. (It is open to debate whether this is realistic or not.) I note akd reported hits on tank commanders at this range in real time play. If so, it further supports a hypothesis that the game engine is doing something different with snipers in real time and turn based play.

Conclusion

On the basis of the above tests the game is fundamentally flawed and broken. We should all play Achtung Panzer while we wait for the game to be fixed.

[c3k]

user38,

Thanks for posting. I have to take exception to your comment "the game is fundamentally flawed and broken." If it was meant in jest, the tone didn't get conveyed. The rest of your post is concise and serious. There is nothing to differentiate this final statement. If you are serious, we have divergent views on the seriousness of TC vulnerability to marksmen.

Recognize this SEEMS to be a specific case of exposed TC's being fired at by aimed rifle fire. Sometimes it hits. It is a specific case with a low probability of a hit. Hardly fundamental.

Ken

[user38]

I am emoticon challenged.

[siffo998]

nice testing user...! allways good to have some people around who are good at probability calculations...

[RockinHarry]

Not directly related to accuracy of shooting, but rather of general behavior. What makes a sharpshooter swapping his scoped rifle for a handgun, while engaging the same target?

check video http://www.youtube.com/watch?v=H7RznDaGAW0

[MengJiao]

hope you are right, my testes today and yesterday are all with 1.01 patch

I'm always a little cranky when I patch my testes too.

[Blackcat]

I'm always a little cranky when I patch my testes too.

Nobody is allowed to touch my testes except my wife. Mind you at my age it doesn't do her much good.

[argie]

Is it the thread in which everybody is playing with everyone testes?

[Boche]

ahhhhhhhhh good old fun!

[sburke]

well at least we haven't degenerated to talking about peanut butter and farm animals...oh wait..oops

[c3k]

I am emoticon challenged.

They can seem silly, but they do provide the useful service of ensuring a certain outlook gets expressed.

Ken

[Magpie_Oz]

I'm always a little cranky when I patch my testes too.

Nobody is allowed to touch my testes except my wife. Mind you at my age it doesn't do her much good.

Is it the thread in which everybody is playing with everyone testes?

I warned him !

[BloodyBucket]

I am emoticon challenged.

Thank goodness. For a minute I thought the game was broken and flawed, but the simple addition of an imaginary smiley saved my bacon.

[user38]

Is the subject matter of this thread now making jokes about JG11Preusse's bad English? JG11Preusse brought up a valid point about snipers firing against tank commanders. It has been confirmed by a few posters to this tread that there is a problem in this area. It is, of course, enjoyable to make fun of foreigners, but I am sure JG11Preusse is doing the best he can to communicate in a difficult language. I take no issue with posters who wish to debate the ideas raised by JG11Preusse. A silly idea is a silly idea and we should all be free to say so. But attacking someone's ideas by attacking the person does nothing to advance the debate.

[Magpie_Oz]

Is the subject matter of this thread now making jokes about JG11Preusse's bad English? .......... But attacking someone's ideas by attacking the person does nothing to advance the debate.

I think it is more about the tone rather than the grammar and the last sentence of you post does apply to both sides.

[JG11Preusse]

maybe it can help, if i tell how i make my tests.

i start the game and choose quickbattle.

here i choose small battle and setup "moving engagement" (not attacking or defending or else).

i also choose troops for both sides myself and i use tiny open map from the 11 qb map-pack or other small maps, but most time the tiny open one.

for german side i choose 10 elite (sometimes not only elite) snipertroops and for us side i choose infantry only, but a huge ammount more then these 10 snipertroops or i choose tanks only or a mix of both.

then i place the snipers in a line near the spawnpoint behint a row of bushes. can not directly place them there, have to run some dozent meters after game starts.

after starting the battle, i let everything almost himself go its way and observe what happens.

this way i think i will have an enviroment almost like in a normal battle from this game and not an artifical build scenario.

sometimes i use small maps with buildings on it, to get one or two snipers get moving in such buildings while the game runs.

i dont use a mapeditor or something else, just let get happen, what happens in this game and observe what each of these snipertroops will do.

[Battlefront.com]

Is the subject matter of this thread now making jokes about JG11Preusse's bad English?

If that were the case I'd be giving people a warning. As it was, people were making fun FROM a typo, not at JG11Preusse. I would expect someone to start that "ball" (pun intended!) rolling even if you made that typo. It's just too tempting

But attacking someone's ideas by attacking the person does nothing to advance the debate.

Absolutely. However, attitude often manages to be translated just fine even if the grammar is not perfect. People are responding to that in a reserved way that, so far, I don't find inappropriate.

Steve

[Battlefront.com]

maybe it can help, if i tell how i make my tests.

That is helpful. What you are doing is not a "test", only an environment for producing "anecdotal" results.

this way i think i will have an enviroment almost like in a normal battle from this game and not an artifical build scenario.

But it isn't normal to have that many snipers doing that sort of stuff. So it's also artificial.

The important reason to conduct "tests" in the Editor is to isolate variables and to be able to replicate identical situations many, many, many times. This is the standard, and most basic, approach to scientific results.

i dont use a mapeditor or something else, just let get happen, what happens in this game and observe what each of these snipertroops will do.

This might show that there is an issue worth looking into, but that is all it does. It is not helpful beyond that because there's no way to test specific circumstances. There's no way to see how common the results are for a given number of tests. There's no way to track improvements if we make changes.

In short, "tests" that are done in the Editor are hard enough to do in a way that is helpful. "Tests" done in Quick Battles, after there is already suspicion of a problem, have no practical value.

Steve

[IMHO]

That is helpful. What you are doing is not a "test", only an environment for producing "anecdotal" results.

Steve,

1. May be you can tell what test setup you would consider helpful? I mean this way you would get more help from the community.

Like when I did CMSF tests I ran it 10 times for each variable change - distance, elevation, building type etc. Variance was manageable and I learned what I wanted to learn. I tested in hot seat mode as it provides better control. Though it seems a post here gives pretty convincing evidence that RT and WeGo gives different results for some reason.

2. May be you could consider telling which variables go into the model? Apart from the obvious stuff. I understand it might be tricky.