Formations - Progress

[navwarcol]

ORIGINAL: dazkaz15

I'm not sure if the problem I see when the footprints overlap is just about casualties.

I think the surrender mechanics might be a bit out as well.

I usually try to get a complete encirclement with a lot of the isolated units entrenched in forests, or other high cover areas, and not only are they not taking casualties, there not surrendering either.

I know you are trying to get a realistic casualty figure, but does this include realistic surrendering figures.

If my tactics are a lot different than those used historically, like I said I like to surround my enemy so when they are routed they can't get away, would there not be a lot different figures than those used historically.

What I'm trying to say is I hope you're not forfeiting a plausible a historical outcome for the sake of historical figures, because It won't work in a game like this that allows you to do tactics that were not used historically.

I would much rather the results where plausible than historical.

I use the same tactics Dazkaz15. I have the same concern... we already know the historic results, but I hope it is able with different tactics to have different results.

[Arjuna]

OK fair enough. I'll look into it.

[Phoenix100]

When I played We Fight and Die Here yesterday I managed to surround a TD unit (intel said 4 tanks) on the very top of Dom Butgenbach and I had infantry, tanks and arty slogging away at it for over two hours. It was dug in, entrenched. Mist. The footprints overlapped, even the counters, I think, at various points. It showed no casualties for two hours, then suddenly 1 tank down. A few minutes later it bolted about 300 metres, routing, and then showed as 'destroyed'. It seemed plausible, though not sure if it was. Gritty unit, dug in, hulls down. Held out because it had no options. As soon as they broke and were out in the open they were knocked out. Probably they'd run out of ammo by then too. There's a narrative you could write to cover it, at any rate. It was facing odds over 10 to 1 though. I suppose I'd expect it to have run sooner, or surrendered, even if I hadn't got a hit. Look forward to seeing what happens in that sort of situation with the new build (this week...)

[BigDuke66]

"Held out because it had no options."
That point is interesting, what if the unit has no choice but to fight or surrender?
If their position is entrenched and moral high I wouldn't wonder if they would hold out even when pounded by every barrel around them, a good entrenchment level should keep their casualties and suppression down, the unit would simply keep their heads down and wait for the other side to expose them self so the encircled unit could finally take some good shots at the enemy.
Altogether this case would mean a slow grind till the encircled unit is finished, I guess only thing that would help are some really heavy artillery pieces making it clear that holding out is no option.

In the other extreme I would expect to see a unit exposed, low on moral and getting pounded out of every direction surrender quiet fast.

In both cases the values of the officer and the ammo level should also be used for the calculation, even the best units throw the towel when lead poorly and running out of ammo.

Could a more drastic effect on the moral level achieve this?
Maybe by considering what enemy presence is around a unit and how strong it is and counting against it the presence of friendly units, so that 2 or 3 battalions could hold out longer when being encircled together instead of each for them self.

[Arjuna]

One thing I did today was revise the code that applied modifiers to hit prob for units being close together. Basically I have added an extra level ( previously it applies a modifier where the range was less than 100m and anopther less than 50m ). It now has one for less than 10 - ie overlapping. I increased the hit prob for all three cases and I removed the restriction that this code was only invoked if the target was in cover. It now applies everywhere except for urban terrain.

I made quite a number of tweaks to the fire, suppression and retreat/rout code. I removed some code from the retreat algorythm which was in effect penalising a force twice for recent cas. this was having the effect of making it more likely that a force routed. I also paired back some of the modifiers reducing the variation range for these. I played through thew tutorial a couple of times - the last time , all the way through. There is now more cases of retreating and less of routing. So wodin can be pleased with that one. [:)] I need to run a series of run throughs of the tutorial, record the stats, average the results and then decide whther its good enough.

But the first full run through resulted in the Americans taking 20% lossess over the four days. Given their posture and force ratio that's about right. Whereas with the previous patch bombardment would account for say 60 to 70% of cas, it now accounts for 28%, with AArm 24% and APer 38%. This result may be low or high compared to the average, so I'll wait till I have these before declaring it ready.

[Phoenix100]

Do you control the arty much yourself when you do this, Dave? Just asking as I always find much higher arty casualties if I control some of it myself to any substantial extent. That might be a different point - how effective is the AI at bringing in arty fire? Quite often I take control because I can't see anything happening to break up attacks, clear objective rings etc. I used to think it would be 'most realistic' if I let the AI control all my arty, but, frankly, I now find it difficult to draw most scenarios if I do that.

The enemy arty is a bane in this gamne, because it's so effective. But that's obviously the way it was in real life, if you even read the stats for percentages killed. I wouldn't want it any differently. A drop from 70 to 28 seems a lot, though, as you imply, depends what happened in the actual fight.

Less routing more retreating sounds like it will make a lot of us happier - until we notice the enemy, surrounded, outgunned and hanging on, retreating a few hundred yards, coming back at it after a pause, again and again...never dying... etc.....I'm sure you can't win, Dave.
Looking forward to this new build.

[Arjuna]

These testing games are all done AI v AI with no human input other that by me to start the thing running.

Yes I know I can't win. It's writ large in the sky, in my genes and now in this forum. T'was ever thus. [:)]

[RockinHarry]

ORIGINAL: Arjuna

One thing I did today was revise the code that applied modifiers to hit prob for units being close together. Basically I have added an extra level ( previously it applies a modifier where the range was less than 100m and anopther less than 50m ). It now has one for less than 10 - ie overlapping. I increased the hit prob for all three cases and I removed the restriction that this code was only invoked if the target was in cover. It now applies everywhere except for urban terrain.

I made quite a number of tweaks to the fire, suppression and retreat/rout code. I removed some code from the retreat algorythm which was in effect penalising a force twice for recent cas. this was having the effect of making it more likely that a force routed. I also paired back some of the modifiers reducing the variation range for these. I played through thew tutorial a couple of times - the last time , all the way through. There is now more cases of retreating and less of routing. So wodin can be pleased with that one. [:)] I need to run a series of run throughs of the tutorial, record the stats, average the results and then decide whther its good enough.

But the first full run through resulted in the Americans taking 20% lossess over the four days. Given their posture and force ratio that's about right. Whereas with the previous patch bombardment would account for say 60 to 70% of cas, it now accounts for 28%, with AArm 24% and APer 38%. This result may be low or high compared to the average, so I'll wait till I have these before declaring it ready.

That sounds very promising! Looking forward to test, hopefully next public beta with this! [:)]

[Alchenar]

It's nice to read about the progress being made.

I just want to re-iterate that my specific concern stems from the Joe's Bridge scenario - you have a 30 man static flak unit that will rout almost immediately but can't run and won't surrender and doesn't take casualties - so unless you are very precise with your orders and get a bit of luck, that flak unit which should simply be overrun and disintegrate without much hesitation can actually hold up the Allied attack for the entire evening as units can't bypass it.

It would be soothing to know that the changes are being tested against that circumstance (and if other players also see this issue occur).

[wodin]

Thanks Dave!

28% Sounds right if there isn't much Arty in the scenario and alot of Inf fighting close combat. I think alot of Arty kills where pre battle bombardment I reckon which happens pre scenario most of the time as a scenario usually starts at the time of the Inf\AFV assault. The tutorial I think if I remember properly is a scenario that doesn't have mass Arty so again I think 28% is sound.

In a long 10 day scenario with loads of Arty I expect that to rise substantially though. You also need to remember if we say there aren't enough small arms casualties they have to be taken off something else and that is Arty casualties..otherwise we will have way more casualties than is historical..really it's down to maths.

[navwarcol]

I would agree with Wodin. I think it must be sometimes tiresome for the designer to be asked to tweak this, then that, only to find that when he tweaked 'that' it messed 'this' back up... I feel for Dave and the many other designers who sometimes must feel they are going in circles....but these products are worth it in the end. For myself, a product such as this one, virtually guarantees I will buy the next from Panther, as well.

[Arjuna]

ORIGINAL: Alchenar

It's nice to read about the progress being made.

I just want to re-iterate that my specific concern stems from the Joe's Bridge scenario - you have a 30 man static flak unit that will rout almost immediately but can't run and won't surrender and doesn't take casualties - so unless you are very precise with your orders and get a bit of luck, that flak unit which should simply be overrun and disintegrate without much hesitation can actually hold up the Allied attack for the entire evening as units can't bypass it.

It would be soothing to know that the changes are being tested against that circumstance (and if other players also see this issue occur).

How about this?

[Arjuna]

And just to prove the point.

[Arjuna]

And success goes to the Irish Guards. I used the just overhauled successive lines formation for the attack. It worked a treat. [:)]

The 3rd Irish Guards across the river have just had their advance guard pushed back by the German counterattack coming from the north. But I'm confident they will hold.

[Arjuna]

I finished the six runs through of the tutorial. The cas results are:

• US
  • Pers = 15% ave - 11% min - 19% max
  • AFV = 17% ave - 13% min - 24% max
  • Gun = 10% ave - 7% min - 13% max
• German
  • Pers = 50% ave - 42% min - 62% max
  • AFV = 54% ave - 30% min - 65% max
  • Gun = 54% ave - 42% min - 64% max

According to Dupuy (Numbers Predictions and War) the attacking Americans should have lost 17% pers, 20% AFV and 7% guns. So we're pretty close. Within the acceptable margins I reckon. The losses for the defending Germans were higher that I expected but the particular scenario is a difficult for the Germans. So I am going to leave things with the combat sytstem as they are now and go back to the ormations.

[dazkaz15]

Thanks for all the hard work Dave.

Im looking forward to playing a dynamic game again.
It was starting to feel more like WW1 than WW2.

[Alchenar]

ORIGINAL: Arjuna

How about this?

Thanks!

I have to admit that given the choice my preference would be for the game to get the tempo of operations right before making sure that the end-result casualties look right (though I know the two are related).

[Phoenix100]

tempo of ops? what do you mean, Alch?

[wodin]

ORIGINAL: Alchenar

ORIGINAL: Arjuna

How about this?

Thanks!

I have to admit that given the choice my preference would be for the game to get the tempo of operations right before making sure that the end-result casualties look right (though I know the two are related).

I do always feel I have to push my troops alot harder and faster than they should be to take the OBJ's in the set time..always clicking rest off..rarely do I feel I can rest troops..and push them to exhaustion, if I didn't I'd not be able to get the OBJ's on time.

[navwarcol]

ORIGINAL: wodin

ORIGINAL: Alchenar

ORIGINAL: Arjuna

How about this?

Thanks!

I have to admit that given the choice my preference would be for the game to get the tempo of operations right before making sure that the end-result casualties look right (though I know the two are related).

I do always feel I have to push my troops alot harder and faster than they should be to take the OBJ's in the set time..always clicking rest off..rarely do I feel I can rest troops..and push them to exhaustion, if I didn't I'd not be able to get the OBJ's on time.

Yes.. can just see the little pixeltruppen "Damn General Wodin... never see HIM out here marching all night with no sleep....!"