strange tank spotting behaviour

[siffo998]

I`ve just ran over something very strange i want to share.

Recently i played a german defense mission with two panthers. One of the panthers got some pounding from an sherman on the edge of the map that he could not spot (even so he was unbottened when the sherman opened up on him, and even after multiple rounds from the sherman and a cover arc in the direction).

So i assumed that maybe somthing is flawed with spotting when tanks are standing directly on the edge of the map. So i started a test scenario.

I placed a sherman m4a3 75mm on the edge of a map and 1000m opposing a panther A mid in the middle of the map (open field no cover at all between them, no height difference). Each facing the other with a cover arc. Experience and everything to normal/average/regular. Both tanks unbuttoned. I wanted to find out which of the tank would spot and first shoot at the other more often. I repeated the test 40 times each for one minute.

here are the results:

Sherman first spots and shoots at panther:

27 times (67,5%)

Panther first spots and shoots at sherman:

11 times (27,5%)

Tanks spot each other at the same time and shoot at each other:

2 times (5%)

http://www.2shared.com/file/-MPDvudf/test_sherman_edge_001.html

at this point i was very confident that my testing proves my theory that a tank at the edge of map the could be spotted far less often.

so i repeated the the test with the panther on the edge and expected similar results now in favor for the panther.

Here are the results:

Sherman first spots and shoots at panther:

25 times (62,5%)

Panther first spots and shoots at sherman:

14 times (35%)

Tanks spot each other at the same time and shoot at each other:

1 times (2,5 %)

save:

http://www.2shared.com/file/2QSHgvur/test_panther_edge_001.html

now my theory was destroyed insted it seems like a m4a3 allways has an advantage over the panther in first spotting him...

so i set up a third test with both tank in the middle of the map 1000m away (again each other facing with cover arc).

Sherman first spots and shoots at panther:

24 times (60%)

Panther first spots and shoots at sherman:

15 times (37,5)

Tanks spot each other at the same time and shoot at each other:

1 times (2,5%)

save:

http://www.2shared.com/file/Ve7pU4Fl/test_both_middle_001.html

this seems very strange to me. A sherman m4a3 could spot and first shoot at a panther (1000m distance) more often than the panther on the sherman ??? from all I`ve read a panther at the russian steppe was able to spot and kill targets like the russian t34 at ranges around 2000m. also he had magnificent sights with two magnification steps. 2,5X and 5X. the 2,5X to search targets and 5X to lock on targets. On the other hand their isnt any account i know of that a sherman m4a3 75mm ever spotted or fought targets at 2000m and his sights only had 5X magnification. (Both sight informations according to Ospreys Duell Panther vs. Sherman 76mm Battle of the Bulge).

I am aware that the silouette of the panther is slightly bigger than the one of the sherman but a height from 20cm+ also brings the advantage that the tank commander could spot more easily if on open ground. (image you climb on a tower...the higher you get...the more you can see.)

sorry but in my opinion a panther should at least have an 50/50 chance, if not better, to first spot and first shoot at an sherman m4a3...(havent testet it with other vecicles yet).

[dieseltaylor]

Interesting stuff.

I have an observation. Firstly that hopefully this is already being dealt with to a degree in V1.01 in terms of speed of target acquisition.

Secondly, you say spots and shoots first. I would be very interested to know if the spotting time is the significant feature. The delay to shooting time is something I could live with as the Sherman I would expect to fire more swiftly following sighting at the same time.

I will do a download : )

[siffo998]

Interesting stuff.

I have an observation. Firstly that hopefully this is already being dealt with to a degree in V1.01 in terms of speed of target acquisition.

I hope so too... ;-)

Secondly, you say spots and shoots first. I would be very interested to know if the spotting time is the significant feature. The delay to shooting time is something I could live with as the Sherman I would expect to fire more swiftly following sighting at the same time.

I will do a download : )

I have never encountered a situation during my testing were the panther spotted first and nonetheless fired after the sherman... the time between spotting and firing seems to be nearly equal for both... in case they spotted and fired at nearly the same time i`ve put it under the third section.

but i`ve encountered multiple situation (stopped counting at 5) were the panther was shot at from sherman side for nearly one minute and still hasnt spotted the sherman...! I can not recall that this was reproduceable with the sherman (sherman was fired at for nearly one minute and the panther was not spotted).

[dieseltaylor]

I have just run two tests and in neither did the Panther spot the Sherman until it was hit by the 75mm shell.. However the variation in spotting and firing between the two tests was remarkable.

In the second test the Panther was spotted within 2 seconds and the first shot 29.50.

The first one the Sherman saw a shape at 29.53 but commenced aiming at 29.33 and fired 29.10

PS. I wondered if having a wall a hundred metres behind the Panther actually makes it easier to see. There may be a presumption that an empty edge is a foliage covered back-drop and that helps the Sherman blend in.

[slysniper]

I do not think it has anything to do with the map. I think it is in the mechanics of the tanks, for some reason, shermans are spotting better than panthers in the game and it might be some type of input error.

I also have noticed how hard it is to get my panthers to spot enemy in the game, where as shermans seem to see things that they should not very quickly. My gut feeling is that there is a possible data error on the sherman tanks. but how to show that in a test is a good question.

[siffo998]

actually i placed the wall to keep the infantry i had to place for the scenario out of the spotting process... but maybe it really influences the testing...

however i`ve repeated the testing with the third scenario (both in the middle) now with an eye only on the spotting process. I`ve counted the seconds until the panther and the sherman first spotted each other. In the first two rows the seconds are shown until the sherman spots the panther and vice verca. in the third row i`ve declared a winner according to which one spotted faster (sherman = S, panther = P). I`ve repeated the test 50 times one minute each. If someone did not spot the enemy in one minute i`ve counted it as 01:00.

Sherman spotted Panther after / Panther spotted Sherman after / Faster

00:01 00:33 S

01:00 00:27 P

00:13 00:59 S

00:08 00:41 S

00:10 00:14 S

01:00 00:02 P

01:00 00:22 P

00:06 00:24 S

00:09 00:27 S

01:00 00:49 P

00:34 01:00 S

01:00 00:33 P

00:02 01:00 S

00:04 00:28 S

00:02 00:25 S

00:11 00:12 S

00:38 00:40 S

00:42 00:10 P

00:16 00:23 S

01:00 00:03 P

01:00 00:07 P

00:10 00:14 S

01:00 00:07 P

00:13 00:18 S

00:09 00:35 S

01:00 00:59 P

01:00 00:22 P

01:00 00:16 P

00:05 00:02 P

00:23 00:53 S

00:01 00:16 S

01:00 00:08 P

01:00 01:00

00:51 00:34 P

00:13 00:41 S

01:00 00:01 P

00:02 01:00 S

01:00 00:18 P

00:21 00:12 P

00:30 00:42 S

00:49 00:15 P

01:00 00:52 P

00:18 00:25 S

00:27 00:02 P

00:32 00:23 P

00:22 00:07 P

00:26 00:51 S

01:00 00:23 P

00:49 00:10 P

00:25 00:24 P

results:

Panther spotted sherman faster: 26 times (52%)

Sherman spotted Panther faster: 23 times (46%)

1 draw (no ones can see the other for one minute)

hm looks like my 3 tests before this one are now revised... with about 50 testing runs the panther even overcome the sherman... I wonder why my 3 tests before this one all favored the sherman ?

[siffo998]

sorry something is wrong with the formatation of the numbers...they are now a bit pressed... ;-) but it should be readable if not write me and i will send you a excel tablet.

[siffo998]

I do not think it has anything to do with the map. I think it is in the mechanics of the tanks, for some reason, shermans are spotting better than panthers in the game and it might be some type of input error.

I also have noticed how hard it is to get my panthers to spot enemy in the game, where as shermans seem to see things that they should not very quickly. My gut feeling is that there is a possible data error on the sherman tanks. but how to show that in a test is a good question.

even when my actual test shows something opposite i`ve had the same feeling until now...!

[Blackcat]

"I wonder why my 3 tests before this one all favored the sherman ?"

The mathematics of probability. If a roulette wheel comes up red nine times in succession, what are the chances of the ball landing on red on the tenth spin? Casino operators mke a lot of money out of people who think it isn't 1 in 2.

[slysniper]

Oh well, maybe it is just one of them percheived things until you test it out and find everything is OK.

There could be terrain factors that help one side spot the other that are in effect, never can tell with these guys, like maybe siluette against a sky background compared to undergrowth behind units, who knows where the logic starts or stops in the game.

Also there might be a advantage to one tank to the other if the set up requires the turrets to move, that could easily give one tank the advantage of getting off the first shot.

[PanzerMiller]

Blackcat is exactly right.

Siffo998, you need to run at least 250 trials -- and more like 500 trials -- before you can start drawing conclusions from your test results. Fifty trials isn't gonna cut it...

[dieseltaylor]

There is acomplication in that tanks sometimes report an enemy icon but until it becomes a tank will not fire. I think this mucks the figures slightly. Out of 6 tests the Shermans are 5 to 1 ahead.

The last one was an icon job where despite spotting a US icon with 40 seconds to go it was never resolved as atnak and despite fire form the 0.5" and two hits from a 75mm it remained that way with the Panther backing up.

I did check for sun direction but it is laid out north south and is 9 am.

Seventh test. No spotting until last ten seconds when Panther spots the Sherman. Sherman oblivious.

[undead reindeer cavalry]

results:

Panther spotted sherman faster: 26 times (52%)

Sherman spotted Panther faster: 23 times (46%)

1 draw (no ones can see the other for one minute)

if you ignore the duels and just look at spotting times, you get something like this with your data (this is of course not accurate):

the chance of spotting the opposing tank per second:

Sherman: 1.6%

Panther: 3.3%

hm looks like my 3 tests before this one are now revised... with about 50 testing runs the panther even overcome the sherman... I wonder why my 3 tests before this one all favored the sherman ?

i suspect CMBN randomness does not have memory (or methodology simulating one), which leads to the casino effect Blackcat talks about. it's not how many things related to tactical combat work in real world, so you get the strange randomness and variation you are not intuitively expecting.

[siffo998]

Blackcat is exactly right.

Siffo998, you need to run at least 250 trials -- and more like 500 trials -- before you can start drawing conclusions from your test results. Fifty trials isn't gonna cut it...

actually in a thread some time ago steve himself posted that you should at least run 40 trials. at best 100. so i think with running 50 test rounds its quite a good compromise. if you do not agree you could download the savegame and run 250 - 500 trials yourself... ;-)

[siffo998]

There is acomplication in that tanks sometimes report an enemy icon but until it becomes a tank will not fire. I think this mucks the figures slightly. Out of 6 tests the Shermans are 5 to 1 ahead.

I`ve only counted the enemy tank as spotted when it really turned out to be a tank and not when it was a "?". But maybe i`ve done some number salad and there lies the problem. to be really sure i will rerun the test. maybe this time without the target arcs.

if the game has at least a little bit of an statistical logic than the numbers should be similar to my last test today.

[siffo998]

i suspect CMBN randomness does not have memory (or methodology simulating one), which leads to the casino effect Blackcat talks about. it's not how many things related to tactical combat work in real world, so you get the strange randomness and variation you are not intuitively expecting.

actually i havent understand what blackcat wanted to say at first so just for my understanding... do you want to say that cmbn tank combat behaviour is only completely random (not related to any real world behaviour) and that you could not get any reproduceable logic even if you run 1000 test... ? If that is true than it would be really gamebusting... ?!

[womble]

The mathematics of probability. If a roulette wheel comes up red nine times in succession, what are the chances of the ball landing on red on the tenth spin? Casino operators mke a lot of money out of people who think it isn't 1 in 2.

And the rest of their money out of people who think that it is.

[undead reindeer cavalry]

actually i havent understand what blackcat wanted to say at first so just for my understanding... do you want to say that cmbn tank combat behaviour is only completely random (not related to any real world behaviour) and that you could not get any reproduceable logic even if you run 1000 test... ? If that is true than it would be really gamebusting... ?!

the idea is that each time you roll a dice, the chances that you roll 1 are the same. it doesn't matter if you have rolled 1 a hundred times in a row previously. the chances are the same for each roll.

spotting a tank (for example) is different. crews are trained specific scanning techniques for spotting targets. the chance that a crew spots an enemy tank in any given second increases by every second they do scanning. it's a bit like if you had four boxes and one of them contained a cake -- each time you open one box the probability that you find the cake from the next box increases.

what i mean regarding the game is that i suspect it works in some areas by the dice method, not the cake-in-a-box method. this is why there's a huge variation in results for example in your spotting test, as with bad luck the Sherman may never spot the Panther during the 30 minute battle (while it may spot it on the first second if you are lucky) -- each time the game "rolls the dice" to see if Sherman spots the Panther the chance of spotting remains the same.

i don't mean that there wouldn't be logic in the game or that it would all be just randomness. i just mean that there's so much randomness that you need to do those 2000 tests to have good statistics. and because of the same reason it's almost irrelevant when you play the game (you would need to play 2000 battles to see the difference, just like when running tests).

i don't mean that the game would be broken. i'm just pointing out that there's so much variation that it's practically impossible to draw conclusions even after running 50 tests.

[Blackcat]

And the rest of their money out of people who think that it is.

Womble, stay away from gambling on games of chance! If you believe that a roulette wheel has a memory you will end up broke very quickly.

[dieseltaylor]

Roulette wheels have no memory but then the chances of them being perfectly balanced is another matter entirely : ) And as for the skill of putting the ball into play .....

For some reason casinos keep a very good record of the numbers turning up on each wheel - perhaps they don't believe in totally random either. : )

[Blackcat]

... For some reason casinos keep a very good record of the numbers turning up on each wheel - perhaps they don't believe in totally random either. : )

They have to. Firstly to make sure the person putting the ball into play isn't too "skillfull" and secondly to help prove to the Gaming Commission hat they are running a straight table. Anyone who gambles on a cruise ship needs psychiatric help.

However, corruption in the real world of roulette doesn't have anything to do with CMBN even when roulette is being used as a metaphor for what is going on in the game.

[womble]

Womble, stay away from gambling on games of chance! If you believe that a roulette wheel has a memory you will end up broke very quickly.

You miss my point. The chance of a roulette wheel coming up black is not 1 in 2, ditto red. There are (usually green) 0 and on 'American' wheels 00 as well, so 18/37 or 38 for either red or black. That and the fact that they pay out 35:1 for a 1:36 (or 1:37) chance is where the roulette wheels make their money, because whether people bet consistently on the same colour "Because it's got to come up eventually" or not, the odds of them winning are the same.

[Blackcat]

You miss my point. The chance of a roulette wheel coming up black is not 1 in 2, ditto red. There are (usually green) 0 and on 'American' wheels 00 as well, so 18/37 or 38 for either red or black. That and the fact that they pay out 35:1 for a 1:36 (or 1:37) chance is where the roulette wheels make their money, because whether people bet consistently on the same colour "Because it's got to come up eventually" or not, the odds of them winning are the same.

I did miss your point, I have now taken it. Why any sane person plays roulette or slot machines is beyond me.

[Agua]

Which is why I have managed to lose a grand total of $80 in 15 years of living in midst of a casino boom town. They own the most expensive properties with the most ostentatious decor for a reason: they can afford it.

[womble]

I did miss your point, I have now taken it. Why any sane person plays roulette or slot machines is beyond me.

Anyone playing and expecting, or even hoping to come out ahead, has a screw loose, I'll agree. I can see the attraction of taking the money you might've spent on show tickets/coke/hookers or the like and choosing to blow it on the gee gees or the tables instead: there's a thrill in thinking you might get lucky after all Personally, I'd rather drop a ton on a good restaurant; good food (even if it's a shortcut to the cardiac ward) is healthier than gambling debts.