A spotting test

[poesel]

I wanted to do some spotting tests in forests since I'm still a bit unsure how far you can see into wooded areas.

To do that I need a 'base' test first with as few variables as possible. I choose FOs because they have the same size in the US and german versions, have binocs and are not that trigger happy.

But this test already produced some weird results which I don't have a good explanation.

Setup:

1x DR FO, 5x US FO

Flat terrain, 200m x 1000m, good weather, Iron

DR on baseline, US every 200m (US1 = 200m, US2=400m, …)

No orders were given, all units were oriented towards the enemy during setup.

US1 spots DR: 28s

US1 IDs DR: 63s

DR spots US1: 22s

DR IDs US1: 30s

This goes as expected after roughly a minute the FOs have spotted each other.

(continued in next posting because of 5 image limit)

Overview:

[poesel]

I've left the units as is for 10 minutes. No orders were given. Most units were completely prone, in some one soldier was kneeling. I can't see a difference between kneeling and non-kneeling units in respect to spotting.

I'll start with the FO farthest away. He spots the friendlies 200m ahead. Nothing else is ever spotted so I guess there is a cut-off between 200-400m. Will test later.

Next team spots the friendlies ahead and after - all good.

Same here:

Now here it starts getting weird: this team did not spot anything for the whole of 10 minutes lying in the grass. It is spotted by team 3 and since LOS is reciprocal team 2 must also be able to see 3.

That should IMHO not happen. All other teams have spotted inside the first two minutes so the chances are not that bad. No spotting in 10 minutes looks like a bug to me.

Another weird thing: team 1 spots the germans and team 4 - but not team 2 & 3!? This happened around minute 5 (don't know exactly) but has not changed afterwards.

This may happen but IMHO is very, very unlikely.

Is this a one-off bug? I think not. There must have been a lot of spotting checks during the time. If just one succeeds we have a contact or ID.

Now its already late here but I will have to rerun the tests to make sure (or someone else does (here was a smiley but it also counts against the 5 picture limit...))

The scenario is here:

http://dl.dropbox.com/u/8811801/0%20Sehtest%20I.btt

The savegame from minute 10 here:

http://dl.dropbox.com/u/8811801/Sehtest%20I.bts

[c3k]

poesel71,

Thanks for sharing this. I'll take a look at it. Obviously, we've got to see if there are any differences in experience, chain of command, posture, etc. for US team 1. Something _seems_ to be different.

Thanks,

Ken

[Amizaur]

I think that after doing some "synthetic" tests, with as few variables as possible, and estabilishing some general results, the next step would be to try to recreate one of bizzare cases of strange spotting problems reported here in forums.

Because it may be that the spotting mechanism is absolutely OK in general, but some details of terrain and geomtetry are causing unexpected glitches that only happen in specific situations (it may depend on terrain type, distance, angles, height, maybe light, or the way the spotting checks are calculated, who knows...).

So after checking the spotting performance in general, I would try to set units and tanks in the very same places of the very same map as shown on pictures and try see if the same thing happens. If it happens, then I would try to investigate what specifically causes this, or send the save file to devs maybe they have better tools to analyse it.

---> poesel71:

The "team A spotted team B but team B didn't spot team A at all" phenomena is strange, maybe should not happen in the game. But I think it's not completly unrealistic in terms of real-world. On edges of effective spotting distance, it could happen that member of team A using binocs accidetally spotted an almost-invisible team B member. After it spotted it, it knows where to look and it can maintain contact with it, while being very carefull to not be seen by him. Then it can observe guys from Team B for hours, and they - having very little chance and being less lucky - may not notice him. But I'm not sure something like that is and/or should be modelled in the game.

Do you know maybe, if the light conditions - especially the sun direction - affects the spotting ? Is this "unequal spotting performance" direction-related or maybe random ? Are your teams positioned "equally" regarding the sun direction, I mean sun being from the side ?

[Crushingleeek]

good work. what about the difference in spotting times between DR and US?

[Bradley Dick]

I think there's a lot more abstraction than the graphics engine would lead us to believe. The heavy wooded tile looks like you could drive over it but it's absolutely impassable. I would love to see what some rigorous testing comes up with.

[Vanir Ausf B]

There is. For example, the rocky terrain type looks as flat as concrete, but it provides cover significantly better than pavement or dirt (I have tested it).

[Bradley Dick]

There is. For example, the rocky terrain type looks as flat as concrete, but it provides cover significantly better than pavement or dirt (I have tested it).

Isn't pavement still providing some cover or did they stop abstracting things like lamp posts since CMx1? I could be wrong but I think a road is zero cover and everything goes up from there.

[Vanir Ausf B]

I don't know if pavement is zero cover, I just know rocky provides more cover despite being just as flat visually.

[poesel]

Obviously, we've got to see if there are any differences in experience, chain of command, posture, etc. for US team 1. Something _seems_ to be different.

Experience for all units is regular. None of the units are in C2 as there is no superior unit. Posture I don't know - I have placed them during setup and no other orders were given.

Most were prone, some had soldier(s) kneeling. I have not collected data on that aspect but I can't see that the kneeling units are spotting / being spotted better than the others.

My main point with this test is that it looks(!) like spotting is sometimes not reciprocal.

I've run the test again (see above).

After 10 minutes ('!' where non-reciprocal):

DR spots US1

US1 spots DR, US3(!), US4(!)

US2 spots DR(!), US3

US3 spots US2, US4, US5

US4 spots US3, US5

US5 spots US3, US4

Not much to worry one might think. But after minute two there was no change in spotting. So for 8 minutes nothing happened - US1 sees 3&4 but not reverse. In my understanding of the system spotting should have happened at some time.

Save is here:

http://dl.dropbox.com/u/8811801/Sehtest%20002.bts

Now lets test if they start moving:

I give US2-US5 a MOVE order to the adjacent AS. Now the US units do all spot at least 200m in each direction and have a sound contact for 400m. DR can now ID to 400m.

Next test: QUICK

Same as above. Not much changes for the US. DR gets a sound contact to 600m and an ID for a few seconds.

Movement in the open extends the spotting range by ~200m. Fast moves add another ~200m. Spotting of moving units works IMHO as expected.

Lets see what happens when I up the number of units. I put 3 US FOs per 200m line on the field and start again.

Result:

Some units still can't see others. On average however spotting looks good.

I pan around with the camera to trace the bad spotters. Hmm, theres always some higher grass in the direction of a unit that can't be spotted. I had left the standard gras ground as is and thought this is all decorative...

Next test:

The 200m marks are sand. I change everything in between to water (this looks really weird now) and rerun the test (http://dl.dropbox.com/u/8811801/Sehtest%20003.bts after on minute).

Lo and behold: spotting works for everyone! So what I thought as simple 'gras' is 'scrub' ('Gestrüpp') and is modelled as such in the game.

Wow! I bow my head before BFC! This is also simulated and so units lying behind scrub can't be spotted by other units at the same height. Now if you could only make us see all this magic we wouldn't ask so many stupid questions here!