Back from the testing range with interesting findings about trees

[Bulletpoint]

I previously did some tests of 60mm mortars against AT guns in forest. The tests showed that the guns gain very substantial protection from being in light forest tiles with 1 tree, surrounded by 8 similar tiles.

But my tests did not show if this protection came from the type of tile itself, or the actual trees.

I now ran more tests, to find out what happens when guns are in an "orchard" with only trees, or when they are placed in light forest tiles without any trees at all.

I won't describe the whole testing setup again, but here are the raw numbers:

Guns in pure grassland: 107 killed, 37 wounded

Guns in light forest + 1 tree per square: 35 killed, 27 wounded

Guns in naked light forest tiles, no trees: 93 killed, 30 wounded

Guns in the orchard, 1 tree per tile: 52 killed, 35 wounded

Conclusions

So, it seems it's the trees themselves that protect against 60mm mortars, not the forest terrain type. However, there also seems to be a small cover effect from the light forest tile. It could be a statistical fluke, because I only ran 4 tests with 5 target guns in each test.

Also, I noticed something very interesting: Sparse trees provide a LOT of concealment to AT-guns. In the tests with naked grassland, and tests with forest tiles without trees, ALL the AT-guns were spotted at 1 km range, in ALL the tests, after a maximum of 30 seconds.

But NONE of the guns were spotted in ANY of the 'orchard' tests. So, an AT-gun sitting in a plain grass square with 1 tree, and having one single square in front of it that also has one tree, becomes basically unspottable at this range, at least within a 7 minute timeframe. I suspect the concealment bonus of forest tiles is only applicable to infantry, and gun crewmembers, not the actual gun.

Recommendations

• If you have to kill an AT-gun in light forest, with at least one tree, expect to spend three times more ammo than against a gun in the open. Basically, don't bother with it unless you have no other way of knocking it out.
   
• If you're defending, try placing your gun in forest, and if no real forest is available, an orchard will still cut your losses from mortar fire by about 50 percent and make it massively more difficult to spot the gun.
   
• Forget what you read about "tree bursts" when playing this game.

Edited September 19, 2015 by Bulletpoint

[womble]

Note also that this may (until further testing says otherwise) only apply to 60mm mortars. It probably applies to lighter mortars, since the protective effect of trees will have at least some component of "keeping the origin of the explosion away from the targets" (as has been mentioned in a previous thread). It would be interesting to see whether the height of the trees makes a difference: do taller trees have a greater protective effect, and do even bushes causing above-ground detonation do the same thing?

I'm intrigued that the single tree between the gun and the observer ("combined" with the tree in the AS the gun's in) makes such a difference in concealment. IME just putting an ATG at the very edge of a tree'ed area (i.e. next to a tree that's at the edge of the patch of woodland) makes no difference whatsoever in the concealment of the gun, and the target is spotted as easily as if it were on a tile of that type with no tree.

Edited September 19, 2015 by womble

[Bulletpoint]

Note also that this may (until further testing says otherwise) only apply to 60mm mortars.

Yep. But I think it is quite important, since it's such a staple US weapon.

I'm intrigued that the single tree between the gun and the observer ("combined" with the tree in the AS the gun's in) makes such a difference in concealment. IME just putting an ATG at the very edge of a tree'ed area (i.e. next to a tree that's at the edge of the patch of woodland) makes no difference whatsoever in the concealment of the gun, and the target is spotted as easily as if it were on a tile of that type with no tree.

I've been thinking a bit more about this issue, and I'm wondering if it's a consequence of trees breaking line of sight. Without LOS, there's no spotting possible at all, no matter how long you wait. So maybe this onlly happens because the mortar teams were firing straight at the guns, and the lone tree was exactly between the mortar team and the gun.

This would imply that spotting is always calculated from the centre of the square where the spotting team is, and the LOS is drawn to the centre of the square where the enemy is. regardless of where individual soldiers in each team are located within the squares.

So, if I'm standing in the middle of an 8 metre square, and looking at you standing in the centre of another square, and there's a single tree exactly between us, of course I won't be able to see you. But my buddy who's standing 4 metres to my right should be able to spot you. But in the game, it seems all team members are placed exactly in the middle of the square for spotting purposes?

It also seems that the physical width of the gun is not taken into account either. It only occupies a mathematical point. So, it can hide behind a tree despite being several times wider.

 

 

Edited September 19, 2015 by Bulletpoint

[MOS:96B2P]

<Snip> I'm intrigued that the single tree between the gun and the observer ("combined" with the tree in the AS the gun's in) makes such a difference in concealment. IME just putting an ATG at the very edge of a tree'ed area (i.e. next to a tree that's at the edge of the patch of woodland) makes no difference whatsoever in the concealment of the gun, and the target is spotted as easily as if it were on a tile of that type with no tree.

This placing of the ATG has great tactical significance.  It would be interesting if the testing could nail down if placing the ATG at the edge of a woodline does much good or not.  Maybe we should be placing them in the second action spot back into the woods?   

<Snip> I've been thinking a bit more about this issue, and I'm wondering if it's a consequence of trees breaking line of sight. Without LOS, there's no spotting possible at all, no matter how long you wait. So maybe this onlly happens because the mortar teams were firing straight at the guns, and the lone tree was exactly between the mortar team and the gun.  <Snip> 

 

IIRC the tree is randomly placed within the action spot when your building a map.  I think it would be easy to move the ATG over a bit and/or re-set the tree to see if this changes anything and test your idea of the loan tree exactly between the mortar and spotter.  It sounds like the mortars were being used in direct lay?     

[Bud Backer]

This placing of the ATG has great tactical significance.  It would be interesting if the testing could nail down if placing the ATG at the edge of a woodline does much good or not.  Maybe we should be placing them in the second action spot back into the woods?   

   

I think this is discussed on Bil's blog here:

Trees and Cover

[MOS:96B2P]

I think this is discussed on Bil's blog here:

Trees and Cover

Thanks Bud.  It figures Bil would have already covered that.  That is a great blog he has.  I keep meaning to take some time and read through the entire blog from beginning to end.  Instead I just read part that I think I need at the time.  Someday soon ............

 

[womble]

This placing of the ATG has great tactical significance.  It would be interesting if the testing could nail down if placing the ATG at the edge of a woodline does much good or not.  Maybe we should be placing them in the second action spot back into the woods?     

That's exactly what you should be doing, and not just because of the tree thing that's being discussed. I can assure you that ATGs at the edges of woodlines are generally easy targets to acquire.

IIRC the tree is randomly placed within the action spot when your building a map.  I think it would be easy to move the ATG over a bit and/or re-set the tree to see if this changes anything and test your idea of the loan tree exactly between the mortar and spotter.  It sounds like the mortars were being used in direct lay?     

There are two sorts of tree placements: single tree in the middle of the AS; n trees randomly placed, where n = 1-3. To move the ATG around inside its AS you'd have to change their facing.

 

 

Yep. But I think it is quite important, since it's such a staple US weapon.

 

Didn't say it wasn't important, just reminding of the narrowness of the parameters when your language started getting more generalised.

 

I've been thinking a bit more about this issue, and I'm wondering if it's a consequence of trees breaking line of sight. Without LOS, there's no spotting possible at all, no matter how long you wait. So maybe this onlly happens because the mortar teams were firing straight at the guns, and the lone tree was exactly between the mortar team and the gun.

This would imply that spotting is always calculated from the centre of the square where the spotting team is, and the LOS is drawn to the centre of the square where the enemy is. regardless of where individual soldiers in each team are located within the squares.

So, if I'm standing in the middle of an 8 metre square, and looking at you standing in the centre of another square, and there's a single tree exactly between us, of course I won't be able to see you. But my buddy who's standing 4 metres to my right should be able to spot you. But in the game, it seems all team members are placed exactly in the middle of the square for spotting purposes?

It's evident that this isn't the case, and BFC have also asserted it's not. It's true for estimations of LOS/LOF from action spots that haven't yet been reached to AS where there's no enemy spotted; those calculate centre-to-centre. But when you're measuring LOS from a unit's current location, every eyeball's location is checked.

However, if the target hasn't yet been spotted, the LOS check will be to the centre of the AS where the ATG is. And if there's a tree there (exactly in the centre; on the point the LOS check is being made to), maybe the LOS check is affected some how.

But I'm convinced from experience and BFC's assurances consistent with my experience that it's not that it doesn't use the full observational baseline of the observing element. Assuming they're all looking and there aren't other obstacles to seeing. 

[Bulletpoint]

Didn't say it wasn't important, just reminding of the narrowness of the parameters when your language started getting more generalised.

Fully agree. This test is only about 60mm mortars. Would be interesting to go back and test 81mm too.

If treebursts work (and the "increased distance from point of impact theory" is correct), then I suppose at some calibre, mortars should start to get more deadly against targets under trees, instead of less deadly as seems to be the case with the 60mm mortars.

However, if the "trees provide a general cover modifier" theory is correct, then larger calibre mortars should also be less deadly against targets under tree cover than against targets in the open, but still relatively more effective than smaller calibres.

Edited September 19, 2015 by Bulletpoint

[Bud Backer]

Thanks Bud.  It figures Bil would have already covered that.  That is a great blog he has.  I keep meaning to take some time and read through the entire blog from beginning to end.  Instead I just read part that I think I need at the time.  Someday soon ............

 

I went through much of it when I discovered it, but he has added more since then, and I wasn't experienced enough to appreciate fully everything at that time so I have to re-read it all. It's a treasure-trove of info. 

[Rinaldi]

Orchards tend to be deceptively frustrating resistance nests. The sparse nature still provides half-decent cover, and more importantly a masked route of retreat. I wouldn't stick around after the initial fusillade if I can avoid it however.

Edited September 19, 2015 by Rinaldi

[Bulletpoint]

Yeah Bil's blog is great, but mostly for actual real-life tactics (that can often be used in the game). He's not really dealing with the quirks of the game engine.

 

[womble]

Fully agree. This test is only about 60mm mortars. Would be interesting to go back and test 81mm too.

If treebursts work (and the "increased distance from point of impact theory" is correct), then I suppose at some calibre, mortars should start to get more deadly against targets under trees, instead of less deadly as seems to be the case with the 60mm mortars.

However, if the "trees provide a general cover modifier" theory is correct, then larger calibre mortars should also be less deadly against targets under tree cover than against targets in the open, but still relatively more effective than smaller calibres.

I reckon it'll depend on whether the "terrain save" of the base terrain type has any directional parameters to it, and how, exactly, BFC have chosen to model the dispersal patterns of the casualty-causing elements. I'm pretty sure they model the discrete chunks of shrapnel flying about, albeit in a somewhat abstracted manner. I don't know whether they've modelled the actual lobes of the burst patterns; I can't remember anyone doing "mass casualty" tests remarking on how their dense field of soft targets fell over in certain patterns when plastered with plummeting HE... Regardless, the trunks of the trees will probably still get in the way of the modelled shrapnel pieces if the trajectory intersects. I don't expect BFC have modelled the production of secondary missiles (and concomittant ablation of primary missile effect apart from the bits that hit tree elements stopping).

[Bulletpoint]

I reckon it'll depend on whether the "terrain save" of the base terrain type has any directional parameters to it, and how, exactly, BFC have chosen to model the dispersal patterns of the casualty-causing elements. I'm pretty sure they model the discrete chunks of shrapnel flying about, albeit in a somewhat abstracted manner. I don't know whether they've modelled the actual lobes of the burst patterns; I can't remember anyone doing "mass casualty" tests remarking on how their dense field of soft targets fell over in certain patterns when plastered with plummeting HE... Regardless, the trunks of the trees will probably still get in the way of the modelled shrapnel pieces if the trajectory intersects. I don't expect BFC have modelled the production of secondary missiles (and concomittant ablation of primary missile effect apart from the bits that hit tree elements stopping).

Maybe we won't need to take it to that level of complexity to test airbursts.

Basically the overall effect of tree bursts could be demonstrated if troops in trenches under tree cover are shown to take higher casualties than troops in identical trenches without the trees. Did anyone test this scenario before?

Edited September 19, 2015 by Bulletpoint

[Bil Hardenberger]

Yeah Bil's blog is great, but mostly for actual real-life tactics (that can often be used in the game). He's not really dealing with the quirks of the game engine.

 

Yeah, I'll leave that to the Gamey SOB's, you know Doug and his ilk.  

Seriously, the intent of my Blog is to address the application of real world tactics in the CM games... it is my firm belief that if you learn the real world tactics, and know how to apply them in CM you will be successful more times than not.

Edited September 23, 2015 by Bil Hardenberger