Discussion about unrealistic combat in TOAW III

Olorin4

Generalfeldmarschall
Joined
Dec 1, 2007
Messages
251
Reaction score
3
Location
Greece
Country
llGreece
Discussion about possible unrealistic combat in TOAW III

I think this topic deserves its own thread, instead of continuing in the "Stauffenberg speaks" thread. Here, we can continue the discussion about the combat algorithm, started in that thread.

Disclaimer: This thread is not supposed to be a slam fest against the designers. I appreciate all the hard work that is being continuously done on the project. I cannot stress this point enough. What I hope is to help to address what I (and other more esteemed members of the community) regard as possible bugs or shortcomings.
 
Last edited:

Olorin4

Generalfeldmarschall
Joined
Dec 1, 2007
Messages
251
Reaction score
3
Location
Greece
Country
llGreece
The attached screenshot shows the attack that I generated. However, note that the Axis unit had 100% readiness and 150% supply. The Soviet units had 60% readiness and 70% supply - in addition to the huge shock differential (1.25/.8 = 1.56). (I don't know if that's the same as what you were using - it's not clear in your post). As a result, note that the combat odds were 34:5 - less than 7:1. Considering that the fortified status gives a bonus of 8:1 to the defender, the net odds were actually less than one. And the attacker armor strengths were about 24, total, so that ratio was even worse.
But shouldn't anti-armor values also be taken in account? In that case, the bonus from the fortified status is less for the flamepanzers.

Nevertheless, my results weren't quite as bad as yours, although I never could dislodge the defender.
I think the expected result from 2 strong armored divisions attacking a flamepanzer battalion would be to dislodge it, causing serious damage.

I next tried adding the adjacent mech unit to the attack (second screenshot). That should get the odds at least into positive territory, plus it guarantees a flanking bonus.

That attack usually successfully dislodged the defender, as shown in the third screenshot. That makes me suspect that the first attack was not getting the flanking bonus that can be generated from the unit being too small. Apparently, it was big enough for a 10km hex.
Which is all wrong in my opinion. 21 flamepanzers plus a few supporting equipment cannot defend a 10Km front, whatever the circumstances. The ratio is 1 flamepanzer per 500 meters! And the attacking force is 200 T-34's, 50 KV-I's, and 30 KV-II's. The defenders should retreat before combat, but even if this isn't the case, the actual battle should be fairly easy for the attacker. Instead, what usually happens is that many rounds will be burnt, perhaps the defender even managing to hold the hex for many turns against all odds.

I did a few tests of my own in DNO v3.7, involving that flamepanzer battalion and a strong Soviet tank division. Shock was 105% German vs 100% Soviet. Supply and readiness was 100% for the Germans and 50% for the Soviet Mech Corps. Terrain was open. No artillery or air support. The defenders were fortified. I attacked on ignore losses with one Soviet tank division (4td), which had 88 T-34's, 48 KV-I's, 15 KV-II's, 200 B%-7's and 15 T-26's.


Test 1:

1st attack:
failure
Axis losses - 7% (2 flampanzers)
Soviet losses - 12% (36rs, 11 tanks)
used 60% of the turn

2nd attack:
success
Axis losses - 5% (1 flamepanzer)
Soviet losses - 2% (3 tanks, 3 rs)
turn ends


Test 2:

1st attack:
failure
Axis losses - 0%
Soviet losses - 2% (12rs)
used 10% of the turn

2nd attack:
failure
Axis losses - 12% (a few trucks, hmg scouts and tractors)
Soviet losses - 10% (40 rs, 7 tanks)
turn ends


Test 3:

1st attack:
failure
Axis losses -4% (a few trucks, tractors etc)
Soviet losses - 100% (!!!!!)
30% of the turn used


Test 4:
The flamepanzer battalion retreats before combat!


To summarize, from the 10 tests I did:
- in 1 test the tank division evaporated!
- in 3 tests, the flamepanzer battalion held its ground, burning my turn and causing serious casualties to the attacker while suffering minimal damage.
- in 3 tests, the flamepanzer battalion held its ground, causing moderate damage to the attacker and suffering near 0 damage. No combat rounds burned.
- in 2 tests, the flamepanzer battalion retreated, after many rounds and causing serious damage to the attacker.
- in 1 test, the flamepanzer battalion retreated before combat.

I loaded the game in -toawlog mode. I am in no way an expert in reading these files, but I have a few observations:
- when the armor combat begins, only a percentage of the attacking tanks engage in combat. Is this correct?
- in contrast, all of the flamepanzers participated in the combat
- the average effective accuracy of the attacker was 7-12%, but actually only a very few successful hits were possible.
 
Last edited:

ralphtrickey

Member
Joined
Nov 13, 2005
Messages
556
Reaction score
4
Location
Colorado Springs
Country
llUnited States
Can I get a save right before the combat, or a description of which hexes to attack with which units? I'll see if I can explain what's going on there.

Thanks,
Ralph
ralphtrickey (at) operationalwarfare.com
 
Last edited:

El Cid

Member
Joined
Oct 23, 2004
Messages
235
Reaction score
0
Location
Madrid, Spain
Country
llSpain
I have to agree that TOAW produces some unrealistic results, in cases where the defending units are hard to kill. I have seen it happen a lot with recon units, where the agile equipment becomes very hard to hit and kill.

What happens in the TOAW algorithm? If a unit is hit or fire at 10 times even without getting killed, does it continue fighting? Do individual units ever pass a morale check for each round or hit they get?

There is atable in the manual stating the amount of fires a unit gets per round, but is it the same whether we are talking about turns of half day or of full weeks?

Units made of just armor (either because they start with just armor or because its infantry assets are eliminated) should have their combat resolution revisied in my opinion, and not continue to fight indefinitely without the support from infantry units.
 

Bob Cross

Member
Joined
Aug 6, 2002
Messages
626
Reaction score
3
Location
Houston, TX
Country
llUnited States
But shouldn't anti-armor values also be taken in account? In that case, the bonus from the fortified status is less for the flamepanzers.
Ok, fortified anti-armor values are multiplied by 5. It's the anti-personnel values that are multiplied by 8. The attackers do have some non-armored equipment.

I think the expected result from 2 strong armored divisions attacking a flamepanzer battalion would be to dislodge it, causing serious damage.
Regardless of combat strengths? The defender's defense strength was 5, the combined anti-personnel strengths of the two attackers was 34, and anti-armor strengths were 23. The defender was fortified. Combine the bonuses the defender was getting: 1.56 shock, 1.66 readiness, 2.14 supply, 1.5 proficiency and either 5.00 or 8.00 for being fortified. That adds up to a lot, resulting in very little, if any, actual attack margin.

Which is all wrong in my opinion. 21 flamepanzers plus a few supporting equipment cannot defend a 10Km front, whatever the circumstances. The ratio is 1 flamepanzer per 500 meters!
First, let's get the numbers right. The unit had 48 pieces of active equipment in it: 21 Pz II (flame), 8 HMG Scout, 7 MMG Scout, 3 20mm SPAAG, & 9 251/1. That works out to about 208 meters per piece.

And the attacking force is 200 T-34's, 50 KV-I's, and 30 KV-II's. The defenders should retreat before combat, but even if this isn't the case, the actual battle should be fairly easy for the attacker. Instead, what usually happens is that many rounds will be burnt, perhaps the defender even managing to hold the hex for many turns against all odds.
I don't understand anyone could expect an overrun at odds near even or possibly below even. An overrun should require at least odds of 10:1. The flanking bonus should help, but you still have to have some odds. Here, the strongest unit had attack values of 17 AP, 12 AT against a defense strength of 5, that was fortified. That is not an overrun candidate.

I did a few tests of my own in DNO v3.7, involving that flamepanzer battalion and a strong Soviet tank division. Shock was 105% German vs 100% Soviet. Supply and readiness was 100% for the Germans and 50% for the Soviet Mech Corps. Terrain was open. No artillery or air support. The defenders were fortified. I attacked on ignore losses with one Soviet tank division (4td), which had 88 T-34's, 48 KV-I's, 15 KV-II's, 200 B%-7's and 15 T-26's.
You don't give the actual combat strengths here. But, since you're only using one Tank Division, I'm sure they were well less than even.

You can't just look at the unit sizes and say a division should dislodge a battalion. You have to look at the actual combat strengths.
 

Bob Cross

Member
Joined
Aug 6, 2002
Messages
626
Reaction score
3
Location
Houston, TX
Country
llUnited States
Nevertheless, there are areas we're looking at to address:

1. Perhaps instead of multiplying anti-armor strengths by 5, terrain/deployment bonuses should increase something else. Possibilities are "Chances to hit" and "Chance hit strikes side armor". After all, why should being fortified increase the amount of armor a gun can penetrate?

2. Perhaps increasing the amount of active equipment a defense needs in order to avoid the "flanked" penalty. Problem is, Norm never revealed what the current values are. If Ralph can't find them in the code, we'll have to determine them by trial and error.

3. Currently, a retreat-before-combat is only possible on the first attempt of the turn. This gets reset if the unit retreats from combat, and perhaps in a few other, arcane, circumstances that we aren't yet clear about. Once we figure it out, there may be room for additional resetting conditions, giving more opportunities to overrun the unit after it's been attenuated in combat.

4. Currently, the chance of retreat-before-combat due to the flanking penalty is multiplied by a random number between 1 and the ratio of the attacker/defender active equipment densities. Perhaps that could be stiffened by omitting the random number part.
 

Olorin4

Generalfeldmarschall
Joined
Dec 1, 2007
Messages
251
Reaction score
3
Location
Greece
Country
llGreece
You don't give the actual combat strengths here. But, since you're only using one Tank Division, I'm sure they were well less than even.

You can't just look at the unit sizes and say a division should dislodge a battalion. You have to look at the actual combat strengths.
4th Tank Division had the following strength values: APers: 18, AArm: 12, Def: 14, Recon: 31%
Flamepanzer battalion: APers: 3, AArm: 1, Def: 4, Recon: 29%

From the toawlog:

Anti armor combat begins.
Defender anti armor strength: 254
Attacker anti armor strength: 1067
Defender anti personnel strength: 1100
Attacker anti personnel strength: 1485
Defender survivability: 794
Attacker survivability: 939

(I guess all modifiers are taken into account here, recon, deployment etc)

These are the generic values for the unit as a whole. What about individual equipment? Don't we have to take those into account?
 
Last edited:

Bob Cross

Member
Joined
Aug 6, 2002
Messages
626
Reaction score
3
Location
Houston, TX
Country
llUnited States
4th Tank Division had the following strength values: APers: 18, AArm: 12, Def: 14, Recon: 31%
Flamepanzer battalion: APers: 3, AArm: 1, Def: 4, Recon: 29%
About what I expected. 18 vs 4 is 4.5:1 without even adding the effect of the fortification. That's not an overrun candidate even if it hadn't been fortified. With fortification, it's not even 1:1 odds.

Look, as I demonstrated in my example, if you get even moderately decent odds it can be dislodged. Two armor divisions + one mech division were usually good enough.
 

Bob Cross

Member
Joined
Aug 6, 2002
Messages
626
Reaction score
3
Location
Houston, TX
Country
llUnited States
2. Perhaps increasing the amount of active equipment a defense needs in order to avoid the "flanked" penalty. Problem is, Norm never revealed what the current values are. If Ralph can't find them in the code, we'll have to determine them by trial and error.
Note that we do know the level of equipment a defense needs in order to start suffering the density penalty. At 10km/hex it's 250 pieces. So, we can guess at the "ideal" level to perhaps be in the 125 piece area. That's about a regiment's worth, as one would expect for 10km hexes.
 

Olorin4

Generalfeldmarschall
Joined
Dec 1, 2007
Messages
251
Reaction score
3
Location
Greece
Country
llGreece
Ralph, did you open the save yet? I am looking forward to your conclusions, if any.
 

Bob Cross

Member
Joined
Aug 6, 2002
Messages
626
Reaction score
3
Location
Houston, TX
Country
llUnited States
Note that we do know the level of equipment a defense needs in order to start suffering the density penalty. At 10km/hex it's 250 pieces. So, we can guess at the "ideal" level to perhaps be in the 125 piece area. That's about a regiment's worth, as one would expect for 10km hexes.
Note that Ralph did locate in the code what the value was for "minimum active equipment necessary to avoid the flanking penalty". It's 10% of the above value ("minimum active equipment necessary to incur a density penalty"). So, for 10km hexes, it's 25. That explains why the Flame Panzer unit (with 48 pieces) wasn't being flanked.

We felt that it could be safely increased somewhat, so we've increased it to 15%. So the value for 10km hexes will be 37.5 pieces.
 

Telumar

Member
Joined
Jun 9, 2004
Messages
1,690
Reaction score
6
Location
niflheim
Country
llGermany
Note that Ralph did locate in the code what the value was for "minimum active equipment necessary to avoid the flanking penalty". It's 10% of the above value ("minimum active equipment necessary to incur a density penalty"). So, for 10km hexes, it's 25. That explains why the Flame Panzer unit (with 48 pieces) wasn't being flanked.

We felt that it could be safely increased somewhat, so we've increased it to 15%. So the value for 10km hexes will be 37.5 pieces.
It would be nice if such things could be noted anywhere in the manual, i.e. in the appendix where there are already some calculation examples, maybe along with other "arcane and esoteric formulas" :clown: - i think designers and thus scenarios can only benefit from such transparency.
 

ralphtrickey

Member
Joined
Nov 13, 2005
Messages
556
Reaction score
4
Location
Colorado Springs
Country
llUnited States
Ralph, did you open the save yet? I am looking forward to your conclusions, if any.
I looked at it, and at the values that Bob posted from the logs. I've got a few more tests I want to run, but the results are inline with the data. The FlamPanzers are hard to hit, so it's not unexpected that there are very few of them that get destroyed in a turn. There is also a lot of other equipment with that unit, enough to ensure that it doesn't get 'flanked.' I got distracted by playing with Elmer for a bit, I broke a couple of things, and fixed a couple of bugs. After I get them fixed up, I'll take another look and do a more detailed analysis.

Ralph
 
Top