About attack algorithm ,Coordination and Initative ,Is this right?

[charlottep51]

I read the log But I'm still not sure if I'm right


1. If possible, this division will look for a total of 2 widths for every 1 width to fight,

For example, My 20 wide division A will look for A total of 40 wide targets to fight.
For example, now the enemy has 2 x 10 wide and 1 x 20 wide = 40total , which i called "target BOX".

The larger the width of the division, the more effective it is at concentrated fire
(it has a larger target BOX width).

If a small division can't pick a target, it will pick a random target instead (e.g., 4 x 10 wide attacking 2 x 40 wide). The maximum target width can only be 20.



2. Selection of target BOX

A division, instead of selecting the enemy's maximum width, will select the enemy's best division into the target library
The way to measure what makes a Best division is the weight,

Soft attack gains 1 weight per point
and hard attack 1.2 weight per point,

which mean the division that can do the most damage

this Are calculated every hour




3. First round Attack distribution

attack [35% + (Coordination x initiative)], default is 35%,

attacks "the division that can do the most damage"

I Named as "Main target"




4. Second round Attack distribution

Damage of [65% - (Coordination x initiative)], default is 65%,

attacks "all units in the target BOX"
and splits attacks equally



5.Example

For example, the player Me has a 20-wide division with 100 points of attack,
and the target BOX has 40 widths

The enemy has 1 x 20 wide (A) and 2 x 10 wide (B, C),

First round Attack distribution: 35 attacks, hit the "Main target" , let's say A


Second round Attack distribution: 65 attacks, hit A, B, C at the same time,
attacks divided, 65/3=21.6
attacks every division


So, in total


20 width (A) : 21.6+35= 56.6 attack
10 width (B) : 21.6= 21.6 attack
10 width (C) : 21.6= 21.6 attack



6.
superior firepower (upper left line) +5% Coordination
decisive Battle plan (left line) +10% Coordination
Radio +4% Coordination,
tier 6 radar total , +1% Coordination per tier ,


However, the （Coordination x initiative）, that is,
there without initiative, The result of multiplication is 0


Initiative can only be gained from “communication Support company”
+20% initiative at level 1, then +12% initiative per level, total 4 level



Let's say we have a level 6 radar tier, radio technology, no coordination army doctrine , and a full level 4 communication Support company
=35%+56% initiative x10% Coordination=40.6%,
which i think is basically useless, cost a lot of tech ,+5.6% only,For most of the game, the final tech is not available to the player, and the value is only about +2-+3%



We assume a tier 6 radar, radio tech, decisive Battle plan (left line) +10% Coordination, and a full level 4 communication Support company
=35%+56% initiative x20% Coordination=46.2%,

+15.6%
The same
In the actual process
You can get it about +12%-+13%

also not good enough i think
it can augment the player's army when it is already strong enough
But it doesn't support the base attributes when it needed

 

[Cavalry]

Mostly true.
But i don't know where it says the soft attack and hard point. But from diary thread, they say they prioritize lowest Org division or hardness division in the case of AT's targets.

=35%+56% initiative x10% Coordination=40.6%,

I suspect it is 35%+ 1.56 x 10%. Otherwise one doesn't have any coordination bonus if he doesn't use Signal company.

The larger the width of the division, the more effective it is at concentrated fire
(it has a larger target BOX width).

Not sure, I would prefer a smaller box because the more division in the box, the more spread. With 3 divisions, the main target get 57% damage, with 2, the main target get 65%. In 2x40w vs 4x20w, the 20w have a good chance to get 3 or 4 to gang on 1 40w. But when we switch to 10w, I don't now how much chance for 6x10w vs 1x40w. From my current testing game, 10w vs 27w get quite high casualties. I think go low but now lower than 50% width of the other side, otherwise the defense or breakthrough stats is too low, will test more.

And when we say Coordination, I think they means 0.35+ini*Coord "bonus", and this total have a soft cap at 75%.

 

[Corpse Fool]

I don't now how much chance for 6x10w vs 1x40w.

I have some math that can help with this, if you give me more of the scenario in question.

 

[BeauNiddle]

Corpse Fool and I had a thread about this a while back.

Due to the extra damage from going over the enemy defence random fire (small divisions with only one target in the box) are more likely to stack extra hits and thus do slightly more damage in total. Large divisions with multiple target selections will have their spread fire mostly absorbed by the enemy defences - BUT their chosen target is going to get slaughtered.

If you are trying to punch through a thin line then concentration is great. If you are facing a massive org wall with lots of reinforcements then you might want small divisions to hope for the situation where the whole line is deorged at the same time and the stack has to retreat as nobody reinforces in time. (concentrated fire will give the defender more time to bring in reinforcements)

Attached a quote just so you can find the thread. The quote is about four 9w divisions facing four 11w divisions so we could see the difference between small target box vs large target box.

I gave up and drew the tree [I'm just playing with maths at this point. Damn logic puzzles!]

all hit = 1 in 16
3 hits = 4 in 16
2 hits = 6 in 16
1 hit = 4 in 16
0 hits = 1 in 16

large division. 70/30 coordination and 50/50 chance to target A
all hits = 6.25% doing 280% damage = 0.175
3 hits = 25% doing 210% damage = 0.525
2 hits = 37.5% doing 140% damage = 0.525
1 hit = 25% doing 70% damage = 0.175
0 hits = 6.25% doing 0% damage = 0

Averaged damage to primary target = 1.4

Small division. 100% attack on single target but 25/75 chance to target A
all hits = 0.39% doing 400% damage = 0.015
3 hits = 4.69% doing 300% damage = 0.141
2 hits = 21.09% doing 200% damage = 0.422
1 hit = 42.18% doing 100% damage = 0.4218
0 hits = 31.64% doing 0% damage = 0

Averaged damage to primary target = 1 (as you'd expect because on average each unit will shoot one other unit)

And just because. Large division vs 3 targets = 66/33 chance to target A (AB, AC vs BC)
all hits = 19.67% doing 280% damage = 0.5509
3 hits = 39.35% doing 210% damage = 0.8263
2 hits = 29.51% doing 140% damage = 0.4131
1 hit = 9.84% doing 70% damage = 0.0688
0 hits = 1.23% doing 0 damage = 0

Averaged damage to primary target = 1.86


As @Corpse Fool has already pointed out the smaller divisions are spreading 3 damage over the remaining targets and due to the way criticals work that will surge to increased damage in total. The 2.6 the large divisions spread is not only smaller in total but also much less likely to stack and give criticals. But I pity the poor division that gets to be A

 

[GC955]

5.Example

For example, the player Me has a 20-wide division with 100 points of attack,
and the target BOX has 40 widths

The enemy has 1 x 20 wide (A) and 2 x 10 wide (B, C),

First round Attack distribution: 35 attacks, hit the "Main target" , let's say A


Second round Attack distribution: 65 attacks, hit A, B, C at the same time,
attacks divided, 65/3=21.6
attacks every division


So, in total


20 width (A) : 21.6+35= 56.6 attack
10 width (B) : 21.6= 21.6 attack
10 width (C) : 21.6= 21.6 attack



6.
superior firepower (upper left line) +5% Coordination
decisive Battle plan (left line) +10% Coordination
Radio +4% Coordination,
tier 6 radar total , +1% Coordination per tier ,


However, the （Coordination x initiative）, that is,
there without initiative, The result of multiplication is 0


Initiative can only be gained from “communication Support company”
+20% initiative at level 1, then +12% initiative per level, total 4 level



Let's say we have a level 6 radar tier, radio technology, no coordination army doctrine , and a full level 4 communication Support company
=35%+56% initiative x10% Coordination=40.6%,
which i think is basically useless, cost a lot of tech ,+5.6% only,For most of the game, the final tech is not available to the player, and the value is only about +2-+3%



We assume a tier 6 radar, radio tech, decisive Battle plan (left line) +10% Coordination, and a full level 4 communication Support company
=35%+56% initiative x20% Coordination=46.2%,

+15.6%
The same
In the actual process
You can get it about +12%-+13%

also not good enough i think
it can augment the player's army when it is already strong enough
But it doesn't support the base attributes when it needed

We did another thread on this, hereafter there are conclusions:
The formula is:
1 - (35% from define + ( ( 10% from tech + 10% from doctrine + 5% from spirits ) * 1,56 )).
The base initiative is 1,00.


First target damage range:

Without SIGN
0,35<->0.45 for MA, MW and all non specified doctrines;
0,35<->0,5 for MW+ad hoc spirit and SF-L-*;
0,35<->0,55 for GBP-L.

With SIGN39 (*1.32)
0,35<->0.48 for MA, MW and all non specified doctrines;;
0,35<->0,55 for MW+ad hoc spirit and SF-L-*;
0,35<->0,61 for GBP-L.

The maximum value is 0,66 (Union of GBP-L, 1942 Radar and SIGN44). The +10% GBP-L unique bonus gives an attack about 18/20% more "concentrated". The conclusion of the thread is: "actually the effect looks pretty nice, though still not as shattering as the old meta".
This is the thread: https://forum.paradoxplaza.com/forum/threads/optimal-meta-using-simulation-graphs-included.1501778/

 

[charlottep51]

We did another thread on this, hereafter there are conclusions:
The formula is:
1 - (35% from define + ( ( 10% from tech + 10% from doctrine + 5% from spirits ) * 1,56 )).
The base initiative is 1,00.


First target damage range:

Without SIGN
0,35<->0.45 for MA, MW and all non specified doctrines;
0,35<->0,5 for MW+ad hoc spirit and SF-L-*;
0,35<->0,55 for GBP-L.

With SIGN39 (*1.32)
0,35<->0.48 for MA, MW and all non specified doctrines;;
0,35<->0,55 for MW+ad hoc spirit and SF-L-*;
0,35<->0,61 for GBP-L.

The maximum value is 0,66 (Union of GBP-L, 1942 Radar and SIGN44). The +10% GBP-L unique bonus gives an attack about 18/20% more "concentrated". The conclusion of the thread is: "actually the effect looks pretty nice, though still not as shattering as the old meta".
This is the thread: https://forum.paradoxplaza.com/forum/threads/optimal-meta-using-simulation-graphs-included.1501778/

Corpse Fool and I had a thread about this a while back.

Due to the extra damage from going over the enemy defence random fire (small divisions with only one target in the box) are more likely to stack extra hits and thus do slightly more damage in total. Large divisions with multiple target selections will have their spread fire mostly absorbed by the enemy defences - BUT their chosen target is going to get slaughtered.

If you are trying to punch through a thin line then concentration is great. If you are facing a massive org wall with lots of reinforcements then you might want small divisions to hope for the situation where the whole line is deorged at the same time and the stack has to retreat as nobody reinforces in time. (concentrated fire will give the defender more time to bring in reinforcements)

Attached a quote just so you can find the thread. The quote is about four 9w divisions facing four 11w divisions so we could see the difference between small target box vs large target box.

I have some math that can help with this, if you give me more of the scenario in question.

Mostly true.
But i don't know where it says the soft attack and hard point. But from diary thread, they say they prioritize lowest Org division or hardness division in the case of AT's targets.


I suspect it is 35%+ 1.56 x 10%. Otherwise one doesn't have any coordination bonus if he doesn't use Signal company.


Not sure, I would prefer a smaller box because the more division in the box, the more spread. With 3 divisions, the main target get 57% damage, with 2, the main target get 65%. In 2x40w vs 4x20w, the 20w have a good chance to get 3 or 4 to gang on 1 40w. But when we switch to 10w, I don't now how much chance for 6x10w vs 1x40w. From my current testing game, 10w vs 27w get quite high casualties. I think go low but now lower than 50% width of the other side, otherwise the defense or breakthrough stats is too low, will test more.

And when we say Coordination, I think they means 0.35+ini*Coord "bonus", and this total have a soft cap at 75%.

Thank you guys very much!

Your math post are very, very good,I learned a lot from it

 

[Xiziz]

Just asking to clarify, but using signals when you have a lower width than your opponent seems like it is just a waste of IC and a support slot?

AI uses mostly 18-27 width division templates(with some still going 14/4 and the exception of ITA with 10-12cw armour) so signals on 27+ cw only?

 

[GC955]

Just asking to clarify, but using signals when you have a lower width than your opponent seems like it is just a waste of IC and a support slot?

AI uses mostly 18-27 width division templates(with some still going 14/4 and the exception of ITA with 10-12cw armour) so signals on 27+ cw only?

The Coordination effect is relevant for bigger division, because of the capability to find the "best target" and strike it. On a little division (<18 width) the effect is lesser or insignificant: a 10 width division cannot strike the "best target" because of its engagement width (10*2=20), so the coordination is irrelevant and it works almost as before NSB.

So, summarizing my opinion:

<18 width No Coordination => SIGN useless
18<->27 Coordination effect is present => SIGN serviceable, worthwhile if you have enough Support Equipment
>27 Coordination effect is magnified => SIGN is almost mandatory

 

[Xiziz]

The Coordination effect is relevant for bigger division, because of the capability to find the "best target" and strike it. On a little division (<18 width) the effect is lesser or insignificant: a 10 width division cannot strike the "best target" because of its engagement width (10*2=20), so the coordination is irrelevant and it works almost as before NSB.

So, summarizing my opinion:

<18 width No Coordination => SIGN useless
18<->27 Coordination effect is present => SIGN serviceable, worthwhile if you have enough Support Equipment
>27 Coordination effect is magnified => SIGN is almost mandatory

Thanks, that was just what i thought too based on the data presented.

 

[Cavalry]

>27 Coordination effect is magnified => SIGN is almost mandatory

The bonus is quite small, for example level 3 1942 signal has only +44% initiave, and the total is only about 55% vs 50% on no signal company. So on 1 40w vs 4x20w, the concentrated target get 67.5% attack with signal, and 62.5% in the case no signal. Mean 8% more concentrated attack, but the rest of the attack don't waste. A support arty looks a good try and true than spend research on at least 3 signal techs

The main bonus of signal, the planning speed is destroyed by the general'ability "Staff offense"

 

[Corpse Fool]

The Coordination effect is relevant for bigger division, because of the capability to find the "best target" and strike it. On a little division (<18 width) the effect is lesser or insignificant: a 10 width division cannot strike the "best target" because of its engagement width (10*2=20), so the coordination is irrelevant and it works almost as before NSB.

So, summarizing my opinion:

<18 width No Coordination => SIGN useless
18<->27 Coordination effect is present => SIGN serviceable, worthwhile if you have enough Support Equipment
>27 Coordination effect is magnified => SIGN is almost mandatory

This depends more/entirely on relative size in comparison to the enemy, than the raw value of your size. If your formations are smaller then theirs coordination doesnt get to do anything.

 

[GC955]

The bonus is quite small, for example level 3 1942 signal has only +44% initiave, and the total is only about 55% vs 50% on no signal company. So on 1 40w vs 4x20w, the concentrated target get 67.5% attack with signal, and 62.5% in the case no signal. Mean 8% more concentrated attack, but the rest of the attack don't waste. A support arty looks a good try and true than spend research on at least 3 signal techs

The main bonus of signal, the planning speed is destroyed by the general'ability "Staff offense"

The "almost" of my opinion meant that, if you want to amplify at maximum the Coord. effect, you must use the SIGN and use it in a bigger division, but there are other parameters that we don't take into account, for example the enemies combat width. As stated before, the coordination has an impact, (e.g. GBP-L now is an interesting option, SIGN now is ""useful"" while before was practically useless), but not revolutionizes the game.

 

[GC955]

This depends more/entirely on relative size in comparison to the enemy, than the raw value of your size. If your formations are smaller then theirs coordination doesnt get to do anything.

I agree with you, but I see the issue from the attacking point of view: if I use a 40/41/45 combat width is improbable that I encounter a defender with an higher width, so the coordination helps. Moreover, in the other thread we see (if you remember) that NSB bears up small width defenders, so a "meta" situation will be 1*40w vs 8*10w/4*20w, not viceversa.

Then we can discuss between strike divisions one by one or spread the damage to push off the ORG wall all together, but it's another matter.

 

[Senor Bigbits]

The bonus is quite small, for example level 3 1942 signal has only +44% initiave, and the total is only about 55% vs 50% on no signal company. So on 1 40w vs 4x20w, the concentrated target get 67.5% attack with signal, and 62.5% in the case no signal. Mean 8% more concentrated attack, but the rest of the attack don't waste. A support arty looks a good try and true than spend research on at least 3 signal techs

The main bonus of signal, the planning speed is destroyed by the general'ability "Staff offense"

For when you use SF, this is correct. However, L-GBP gives you 10% more coordination, which means the difference between, if we're using 1942 tech with 20% coordination, goes from a 45% split to a 36% percent split.

And, also, to clear up some misconceptions, it was stated in the diary that 'best target' was changed from 'lowest org' to 'whatever target that unit can make the most attacks against'.

 

[Senor Bigbits]

To further iterate the potential that may lie with coordination and SIGN, if one goes down any non-L-GBP doctrine and doesnt choose SIGN, with merely the radio + radar techs, 55% of attacks will be split between all targetable divisions.
With L-GBP and 1942 SIGN, only 36% of attacks get split across all targetable divisions, with the vast majority going towards the target that it can deal the most possible attacks to. Considering this is in regards to large divisions (which means the damage thats dispersed is even more thinly spread due to the higher targeting range, and thusly 'wasted' against units with high defense/breakthrough), that can be a very large amount of attacks.

 

[Corpse Fool]

(which means the damage thats dispersed is even more thinly spread due to the higher targeting range, and thusly 'wasted' against units with high defense/breakthrough)

If we assume that these larger formations like 2x45w in a 90w battle (such that EW=CW) and we have an arbitrary total of 1000 attacks between them, we can get a table that looks sort of like this...

Enemy	Best 35%	Other 35%	Best 50%	Other 50%	Best 66.2%	Other 66.2%
2x45w	1x675 (15 defs/w)	1x325 (7.23~ defs/w)	1x750 (16.67~ defs/w)	1x250 (5.56 defs/w)	1x831 (18.47~ defs/w)	1x169 (3.76~ defs/w)
3x30w	1x567~ (18.9 defs/w)	2x217~ (7.23~ defs/w)	1x667~ (22.23~ defs/w)	2x167~ (5.56 defs/w)	1x775~ (25.83~ defs/w)	2x113 (3.76~ defs/w)
4x22.5w*	1x513~ (22.8 defs/w)	3x163~ (7.24~ defs/w)	1x625 (27.78~ defs/w)	3x125 (5.56 defs/w)	1x747~ (33.2~ defs/w)	3x85~ (3.78~ defs/w)
5x18w
6x15w
7x12.86~w*
8x11.25w*
9x10w**	1x423 (42.3~ defs/w)	8x73~ (7.3~ defs/w)	1x556 (55.6~ defs/w)	8x56~ (5.6~ defs/w)	1x700~ (70~ defs/w)	8x38~ (3.8~ defs/w)
10x9w**

* notes decimal widths that aren't really practical to acquire, actual results will differ slightly.
** marks the point where over stacking penalties occur (ignoring tactics), so actual def/w needs to be higher to account for the penalty.

I didn't really have the energy to fill out the full table, but we can still see the general trend that the secondary targets require about the same defs/w regardless of their total width, while the primary target requires increasingly more defs/w as their size decreases.

The reason I wanted to make a table like that, is so I could make a graph like this...

Which shows the amount of hits you'd expect to be scored against different levels of defenses, for the different coordinations and using the 10w. What I wanted to show is that getting to enough defense to minimize the hit rate of the splash damage is going to have a much greater impact on total damage reduction than trying to stack enough defense to completely minimize hit rate, including the primary target.

 

[GC955]

Which shows the amount of hits you'd expect to be scored against different levels of defenses, for the different coordinations and using the 10w. What I wanted to show is that getting to enough defense to minimize the hit rate of the splash damage is going to have a much greater impact on total damage reduction than trying to stack enough defense to completely minimize hit rate, including the primary target.

If I understand correctly, you say that is more effective for the defender to have more divisions than better divisions, due to the fact that the spread damage per width is constant, while the concentrated one is inversely proportional to the combat width, right?
Apologize me if I misunderstand.

 

[sdplissken]

You can’t use coordination if your engagement width is less than 1/2 combat width of your opponent. Hence, a 10w will have 0 coordination facing 24w divisions. Your 20w divisions can’t coordinate an attack against a 45w division etc

also there is coordination targeting on defense, if you are within EW

 

[Corpse Fool]

you say that is more effective for the defender to have more divisions than better divisions

Yes,

due to the fact that the spread damage per width is constant, while the concentrated one is inversely proportional to the combat width

I'm not really sure what this means. I'm probably just having a brainfart, I feel like that should make sense to me, but it doesn't.

But to try to detail more of my thought process. In the Before Times, larger formations had the benefit of an increased concentration of attacks, and defenses. Defenses totals for the entire force also didn't really matter, only the level of defenses that each particular formation had. Yes, having multiple formations did lower the chances of the enemies stacking more than one of their formations against you, but someone was always going to be getting targeted and having more, smaller formations that had less defenses meant there was a greater chance of those attacks being undefended. Reman's video that shows 1x40w v 2x20w, and the 40 wins is due specifically to this phenomenon. The same amount of attacks are being thrown at either side, but the 40 has more defenses in the formation that is getting hit, to reduce the damage those attacks deal.

That is no longer true. Having more formations will generally now lower the amount of attacks that each formation will suffer, so they won't suffer quite as much amplified damage as they did before due to lacking in defense. Another graph!

This graph is largely the same as before, but it adds a green line for the way that damage would have been spread out in the Before Times. We can see that with the yellow line, it more or less halves the total damage it suffers at 92 defenses, while the green one doesn't cross 200 hits until 356 defenses. You would have needed around 4x as much defenses in the Before Times to have the same level of 'damage reduction'. I suppose what I'm trying to get at here is that stat concentrations don't matter as much as they did before.

You can’t use coordination if your engagement width is less than 1/2 combat width of your opponent. Hence, a 10w will have 0 coordination facing 24w divisions. Your 20w divisions can’t coordinate an attack against a 45w division etc

also there is coordination targeting on defense, if you are within EW

Rather than 'can't use', I think it is more won't use. I personally don't think you should want to use coordination, but I suppose this is still up for debate. But it isn't when you're less than 1/2 the width of the enemy, it is whenever your formation is any bit smaller than the enemy. Since EW is 2xCW, The EW of a 39w is 78, which is too small to fit 2x40w that would be 80w. So it only fits a single one and coordination isn't being used.

 

[sdplissken]

Yes,

I'm not really sure what this means. I'm probably just having a brainfart, I feel like that should make sense to me, but it doesn't.

But to try to detail more of my thought process. In the Before Times, larger formations had the benefit of an increased concentration of attacks, and defenses. Defenses totals for the entire force also didn't really matter, only the level of defenses that each particular formation had. Yes, having multiple formations did lower the chances of the enemies stacking more than one of their formations against you, but someone was always going to be getting targeted and having more, smaller formations that had less defenses meant there was a greater chance of those attacks being undefended. Reman's video that shows 1x40w v 2x20w, and the 40 wins is due specifically to this phenomenon. The same amount of attacks are being thrown at either side, but the 40 has more defenses in the formation that is getting hit, to reduce the damage those attacks deal.

That is no longer true. Having more formations will generally now lower the amount of attacks that each formation will suffer, so they won't suffer quite as much amplified damage as they did before due to lacking in defense. Another graph!
View attachment 787898
This graph is largely the same as before, but it adds a green line for the way that damage would have been spread out in the Before Times. We can see that with the yellow line, it more or less halves the total damage it suffers at 92 defenses, while the green one doesn't cross 200 hits until 356 defenses. You would have needed around 4x as much defenses in the Before Times to have the same level of 'damage reduction'. I suppose what I'm trying to get at here is that stat concentrations don't matter as much as they did before.



Rather than 'can't use', I think it is more won't use. I personally don't think you should want to use coordination, but I suppose this is still up for debate. But it isn't when you're less than 1/2 the width of the enemy, it is whenever your formation is any bit smaller than the enemy. Since EW is 2xCW, The EW of a 39w is 78, which is too small to fit 2x40w that would be 80w. So it only fits a single one and coordination isn't being used.

Why wouldn’t you want to use coordination. It far easier to get overruns when units don’t reinforce. And yeah, you get what I mean I am correct improperly worded English is my second language. That’s exactly what I was explaining.

coordination de orgs divisions more quickly.