TC position in Lend-Lease Sherman?

[Lucky_Strike]

11 hours ago, kohlenklau said:

Thanks for doing it Mark.

I was wondering which of the Shermans it matched to as the A2 just indicated a diesel engine. Tank Commander "looks safer" and we await rocketman's test report to see if any result is obtained beyond cosmetic looks. 

I can see you got your hands full Phil with your map quest and the CMAutoEditor, more than happy to make a few tweaks.

I had to do a bit of Googling and research to figure out which British Sherman matched the M4A2, there's not a single M4A2 in US service, but I think that was how it was, most went to Britain and Commonwealth and USSR all of whom appreciated their reliability and diesel engines.

[rocketman]

12 hours ago, Lucky_Strike said:

That would be worthwhile compared to results with RockinHarry's new mod.

I'm not aware of what mod that is. Please enlighten me.

[rocketman]

12 hours ago, Lucky_Strike said:

That would be worthwhile compared to results with RockinHarry's new mod.

Double post.

Edited December 15, 2022 by rocketman

[Lucky_Strike]

12 hours ago, rocketman said:

I'm not aware of what mod that is. Please enlighten me.

It's the one I mentioned earlier, need to take a look at this thread, the download link is in there somewhere:

@RockinHarry changed the position of Grenadiers riding in the back and the poor guy who is volunteered to be MG gunner so that they are all lower.

@George MC reported that he noticed some difference wrt the protection provided to the MG gunner in this thread:

If you still have any results from your earlier tests it would be worth comparing, especially if you have the scenario or a saved game.

I've made some progress with the mods, PzIV turned into a right pig for some reason, probably because the base model is not in a good way - there are some errors in the models which seem to have knock on effects whenever I try to correct them. It's odd because the models used for the PzIV G late and latest look the same as those used in CMFI, but there the TC is in the right position. The only difference on the surface is the lack of side skirts on the G (Latest) in FI even though the texture is present in the FI data folder. By the time the models gets into RT something has gone wrong with both so that they are using a weird hodgepodge of textures and less standard model data, almost like they were unfinished. Anyway I think I have the solution now so will get them finished v. soon.

[Lucky_Strike]

On 12/13/2022 at 3:46 PM, rocketman said:

How can I get hold of your mods for some testing? Searched CMMODS and found some that Kohl had changed but moved the TC up from a too low position. I'll compare your screenshot to in-game when I get time, but it looks just about right.

Right, I've done the mod for the M4A2 75mm and the PzIVG (Late):

Lend Lease M4A2 75mm

Pz IVG (Late)

Sadly the Pz IVG (Latest) model is really in a bad way - internally there's an unspecified texture attached to a material used in the model which means it can't be exported from Blender without extra tweaking. No matter what extra tweaking I tried all efforts failed to get a result that wasn't bust - throwing it's wheels in the air when moving for eg. This is a limitation of our tool kit that I can't overcome, and without some help from BF there's really no way I can see of correcting this one. Good luck to others who might want to give it a try.

There is a slightly different version of the IVG (Latest) available in CMFI which could be ported to RT but that has other issues - no side skirts armour but all the shadows for them being but one. The TC does sit in the correct position but otherwise it's a buggy model at best. The other existing versions in FB and BN are the same as our RT model so don't waste your time with them.

Anyway here's the finished M4A2 that you were originally concerned about ...

and the PzIVG (Late) to fight against him:

If anyone is feeling adventurous there are a slew of other models that could do with some TLC WRT TC positions ... just from the German armour in RT ...

Hetzer guy looks stuck ...

... get down man FFS!

... another stuck TC

... start digging your grave son

... you're supposed to sit on your seat not stand on it!

... it's called a shield for a reason.

I'd be praying as well!

There are of course all sorts of strange positions especially in HTs and other open-topped vehicles. 

Ultimately I guess it may not matter if the engine is calculating risk based on a look-up table - is the crew man standing or sitting, hatch open or closed - so long as that is the same across the board. If other factors like gun shields and hit boxes are coming into play then perhaps some rebalancing is in order. If the model does have an impact on the risk assessment then some of these fellas don't stand a chance.

 

[rocketman]

Thank you! I will look into it soon. But the discussion about if we can mod models to decrease crewmember vulnerability got me thinking. Say that my tests show that we can actually mod how likely casualties are caused - how will that effect PBEMs? If one of the players have a modded vehicle and the other doesn't - whose model is used when the turn simulation is resolved? Since the other player only sees his own side, there is no way of knowing if the other side has a modded vehicle that improves survivability. Or even has modded his opponent's vehicles to be more vulnerable. What is the limit to modding mdr-files? Can armor be reduced? Can optics be blocked? This can potentialle open a can of worms. I like to think that people will refrain from cheating, but in a competitive tournament, maybe someone is inclined to do so. I have no answers since I don't understand the 3d model modding process, or how CM uses models in the simulation, but it is definitely worthy of discussion as it is a means to tilt the balance in one player's favor.

When I get into testing I will probably start a new thread for clarity and link to it from this one.

[Lucky_Strike]

13 hours ago, rocketman said:

Thank you! I will look into it soon. But the discussion about if we can mod models to decrease crewmember vulnerability got me thinking. Say that my tests show that we can actually mod how likely casualties are caused - how will that effect PBEMs? If one of the players have a modded vehicle and the other doesn't - whose model is used when the turn simulation is resolved? Since the other player only sees his own side, there is no way of knowing if the other side has a modded vehicle that improves survivability. Or even has modded his opponent's vehicles to be more vulnerable.

Yes, this is the big question. We do sort of have an official answer to this in as much as we've been told a few times in the past that the underlying game mechanics are not moddable, that is stuff like what this mod does. This is why I said the evidence that the HT MG gunner is marginally better protected was anecdotal. And this is where your tests come in. Personally I don't know, if the engine is calculating probability based on databases and NOT anything physical (is a 3D model physical?) then nothing we do will have an effect. If the hit box of a model is related to the physical placement of the model in 3D space then maybe changing that physical placement will have an effect. I do know, for example, that making trees really wide has no additional effect on their impact on LOS or cover. But that's trees, maybe pixeltruppen models are treated differently, you may find out 

13 hours ago, rocketman said:

What is the limit to modding mdr-files? Can armor be reduced? Can optics be blocked? This can potentialle open a can of worms. I like to think that people will refrain from cheating, but in a competitive tournament, maybe someone is inclined to do so.

Yes we can do all sorts of stuff visually, but the details of stuff like armour thickness, munitions penetration etc are stored somewhere in the app, probably in a bigass database that the engine refers to for its calculations. Our only question is does altering the physical shape/size/position effect hit boxes and consequentially outcome? We just don't know.

13 hours ago, rocketman said:

I have no answers since I don't understand the 3d model modding process, or how CM uses models in the simulation, but it is definitely worthy of discussion as it is a means to tilt the balance in one player's favor.

So far I haven't seen anything that suggests we do have an impact on how the game processes all of its data. But we really haven't done too many detailed tests - most of the folks doing 3D model mods are still just learning the software and what can be made to work, or not, so don't really have much time to focus on the scientific. However @RockinHarry has been testing more of this stuff so he may well have some new insights in time.

13 hours ago, rocketman said:

When I get into testing I will probably start a new thread for clarity and link to it from this one.

A great idea, I will certainly be watching for your results. And if nothing else at least they look less likely to get shot in their bonce!

[RockinHarry]

4 hours ago, Lucky_Strike said:

18 hours ago, rocketman said:

I have no answers since I don't understand the 3d model modding process, or how CM uses models in the simulation, but it is definitely worthy of discussion as it is a means to tilt the balance in one player's favor.

So far I haven't seen anything that suggests we do have an impact on how the game processes all of its data. But we really haven't done too many detailed tests - most of the folks doing 3D model mods are still just learning the software and what can be made to work, or not, so don't really have much time to focus on the scientific. However @RockinHarry has been testing more of this stuff so he may well have some new insights in time.

My edits on crew member positions in halftracks was just minimal and I made sure that no intersections with vehicles "outer" armored type geomety takes places. I assume vehicles use sort of a hitbox used for various things and I think more detailed data could possibly be stored in META data section. Since one can´t see those hitboxes in Blender graphically, one can´t exclude that any crew position shift "can" have ill effects on their protction (from vehicle armor). That´s from my game testing and observations only. No new insights since then.

[rocketman]

I came up with a simple test that I think proves that the position of the TC matters a lot and thereby that the 3d model for infantry/TC is also the hitbox. I went back to @kohlenklau's fixes for FI and used the Firefly one. Unmodded the TC was barely visible and if my hypothesis was right, he would be impossible to hit with rifle-fire. I put 10 Fireflies at 50 m range from sniper and scout units (submachine guns). I ran the test with modded and unmodded. The unmodded variant suffered no casualties in one turn. What would be an easy target. With the modded visible TCs, they suffered 5 casualties in 7 seconds. So I don't think much more testing is needed.

It would now be interesting to see what else can be modded in Blender. Can optics be blocked? Can the breach to the gun be blocked. Can armor be made full of holes?

Back to my earlier statement about if this makes cheating in PBEMs possible - it seems like it might be (unfortunately) but that needs to be tested as well.

[Falaise]

On 12/22/2022 at 5:48 PM, rocketman said:

I came up with a simple test that I think proves that the position of the TC matters a lot and thereby that the 3d model for infantry/TC is also the hitbox. I went back to @kohlenklau's fixes for FI and used the Firefly one. Unmodded the TC was barely visible and if my hypothesis was right, he would be impossible to hit with rifle-fire. I put 10 Fireflies at 50 m range from sniper and scout units (submachine guns). I ran the test with modded and unmodded. The unmodded variant suffered no casualties in one turn. What would be an easy target. With the modded visible TCs, they suffered 5 casualties in 7 seconds. So I don't think much more testing is needed.

It would now be interesting to see what else can be modded in Blender. Can optics be blocked? Can the breach to the gun be blocked. Can armor be made full of holes?

Back to my earlier statement about if this makes cheating in PBEMs possible - it seems like it might be (unfortunately) but that needs to be tested as well.

very interesting
In the test you performed @kohlenklau took an existing position in a version of cm to replace. Except it is possible that it is coded like this (I say this but I have no idea how to comment it works)
The ultimate test would be to use a fully cloaked, lowered tank leader modified in Blender to see if he is under fire or not

[rocketman]

I ran some tests using @kohlenklau 's Firefly mod for FI - the one that raises the TC to a normal position (slightly vulnerable) from the completely lowered position (stock model) where he can't be hit by fire, but probably can't spot much either. I used this for a proof of concept for all mods that changes the TC/HT gunner position by changing the model in Blender.

In single player mode, the mod loads and the TC is easily hit while in the stock version he is near invincible.

In PBEM - if the player that starts the scenario HAS the mod but the oppo doesn't, then the mod IS NOT loaded and doesn't affect the game.

In PBEM - if the player that starts the scenario DOES NOT have the mod but his oppo DOES, then the mod is loaded in in effect. My guess is that the second player who inputs is moves/orders is the one that triggers the simulation of the coming turn and his 3D models are used.

This testing is pretty tedious so I can't say with certainty if this applies to all cases, but it is what makes sense to me, unless Battlefront chimes in on the issue.

So, in conclusion, if you have any of these mods and plan to play a PBEM, let your oppo know. Problem is that fog of war etc means that you'll never know if a tank will appear in the scenario or not so maybe the prudent thing is to only use these for single player games.