Variable Combat Odds

Matrix Games Forums

Forums Register Login Photo Gallery Member List Search Calendars FAQ

My Profile Inbox Address Book My Subscription My Forums Log Out

Variable Combat Odds

View related threads: (in this forum | in all forums)

Logged in as: Guest

Users viewing this topic: none

Printable Version

All Forums >> [New Releases from Matrix Games] >> Shadow Empire >> Variable Combat Odds

Page: [1]

Message

<< Older Topic Newer Topic >>

Variable Combat Odds - 7/20/2020 11:52:07 AM

mroyer

Posts: 914
Joined: 3/6/2016
Status: offline

In the following pic, I show two different combat odds for the same exact combat. The only difference between the two (that I am aware of) is the order in which I allocated the attacking forces (the yellow arrows).

Is this an intentional feature of the game that I've not noticed so far? I've never seen anything like that in any other wargame I've played before. Or, is it a game-glitch of some sort - I'm on beta-9.

Perhaps it's just the odds "estimation" is varying, and the true underlying combat odds are the same in both cases?

I'm not complaining ( yet

). Indeed, it's possible I'll even think it's a cool feature once I understand it.

-Mark R.

Attachment (1)

< Message edited by mroyer -- 7/20/2020 11:53:33 AM >

Post #: 1

RE: Variable Combat Odds - 7/20/2020 11:59:12 AM

demiare

Posts: 470
Joined: 6/20/2020
Status: offline

If you're sending much more soldiers then enemy have - they get a attack penalty. This penalty is applied in order, so your firsts units don't get it, while each additional unit receive larger and larger penalty.

No reasons to send 1000 tanks to attack a single soldier - they simply will hinder each other.

(in reply to mroyer)

Post #: 2

RE: Variable Combat Odds - 7/20/2020 12:01:22 PM

demiare

Posts: 470
Joined: 6/20/2020
Status: offline

On first screenshot : your sentinels receive a penalty for outnumbering enemy too much. And as they have very strong attack for early game - penalty for them reducing your chances a lot.

On second screenshot : your sentinels attacking with full power, while useless militia suffer some penalty (not very important as militia is weak by design).

This is what's happening.

(in reply to demiare)

Post #: 3

RE: Variable Combat Odds - 7/20/2020 12:07:21 PM

mroyer

Posts: 914
Joined: 3/6/2016
Status: offline

AhhHA...! I think I like it - thanks for explaining demiare.

I knew odds could actually go down as you over concentrate on an attack, but I never knew the order of attack made a difference. I don't ever remember seeing that before in a game of this scale. A cool touch.

-Mark R.

(in reply to demiare)

Post #: 4

RE: Variable Combat Odds - 7/20/2020 2:15:15 PM

Raagun

Posts: 37
Joined: 7/18/2012
Status: offline

This is cool but really needs UI explanation

(in reply to mroyer)

Post #: 5

RE: Variable Combat Odds - 7/20/2020 2:52:13 PM

demiare

Posts: 470
Joined: 6/20/2020
Status: offline

quote:

ORIGINAL: Raagun

This is cool but really needs UI explanation

If you hover your mouse over icons of combat modifiers - you will get tool-tip describing how it's working. There is a tons of hidden stuff here.

I don't remember what icon is used for outnumbering penalty - guess -1% / -2% one.

(in reply to Raagun)

Post #: 6

RE: Variable Combat Odds - 7/20/2020 5:20:19 PM

lloydster4

Posts: 164
Joined: 6/19/2020
Status: offline

The most significant difference on the attacker side is Command bonuses. The most significant difference on the defender side is Readiness penalties.

Why would the order of attacking units change these? Why would the defender side be affected at all?

(in reply to demiare)

Post #: 7

RE: Variable Combat Odds - 7/20/2020 5:27:43 PM

demiare

Posts: 470
Joined: 6/20/2020
Status: offline

quote:

ORIGINAL: lloydster4

The most significant difference on the attacker side is Command bonuses. The most significant difference on the defender side is Readiness penalties.

Are you speaking about OP example or about generic situation?

quote:

ORIGINAL: lloydster4
Why would the order of attacking units change these? Why would the defender side be affected at all?

Try to imagine situation when 10 humans attacking one (or remember some movies

). Are they capable to attack alltogether? No they would hinder each other. To simulate this game using penalty for outnumbering defenders too much and order of units determine who's going first and who is waiting their for their turn in back rows.

< Message edited by demiare -- 7/20/2020 5:28:29 PM >

(in reply to lloydster4)

Post #: 8

RE: Variable Combat Odds - 7/20/2020 7:04:13 PM

eddieballgame

Posts: 676
Joined: 6/29/2011
Status: offline

quote:

ORIGINAL: demiare

quote:

ORIGINAL: lloydster4

The most significant difference on the attacker side is Command bonuses. The most significant difference on the defender side is Readiness penalties.

Are you speaking about OP example or about generic situation?

quote:

ORIGINAL: lloydster4
Why would the order of attacking units change these? Why would the defender side be affected at all?

Try to imagine situation when 10 humans attacking one (or remember some movies ). Are they capable to attack alltogether? No they would hinder each other. To simulate this game using penalty for outnumbering defenders too much and order of units determine who's going first and who is waiting their for their turn in back rows.

That is why I pay little attention to the 'Odds', but focus on the 'Offensive & Defensive Mods' data.
Sometimes, leaving out a particular group makes for better results, particularly in the 'soldiers loss' department.

(in reply to demiare)

Post #: 9

RE: Variable Combat Odds - 7/20/2020 8:06:31 PM

lloydster4

Posts: 164
Joined: 6/19/2020
Status: offline

Demiare, I understand the concept behind the Max-Attacked penalty. But look at OP's example. Why would I be talking about a "generic situation"?

The most significant changes from the 1st to 2nd example are, again: Attacker Command Bonus and Defender Readiness Penalty.

How would these stats be affected by the max-attacked penalty? That doesn't make sense.

In OP's example, Max-attacked penalty is -1% in the 1st screen and -2% in the 2nd. Absolute peanuts.

(in reply to eddieballgame)

Post #: 10

RE: Variable Combat Odds - 7/20/2020 9:54:41 PM

demiare

Posts: 470
Joined: 6/20/2020
Status: offline

quote:

ORIGINAL: lloydster4

Demiare, I understand the concept behind the Max-Attacked penalty. But look at OP's example. Why would I be talking about a "generic situation"?

The most significant changes from the 1st to 2nd example are, again: Attacker Command Bonus and Defender Readiness Penalty.

How would these stats be affected by the max-attacked penalty? That doesn't make sense.

In OP's example, Max-attacked penalty is -1% in the 1st screen and -2% in the 2nd. Absolute peanuts.

Easy. His troops are high in numbers but VERY weak (half of them militia! no tanks, no buggies, no guns - useless infantry only). His major strike power are walkers. Yes, only 10 walkers (equal to 100 men) but they're his main and only power.

In first example walkers get massive attack penalty because of tons of useless bodies send in front of them. This penalty affect only 1% of his attacking forces - 1% counted in manpower, not in a firepower.

In second example his major and only firepower fight without restrictions, while some useless militia guys eating penalty. Yes, there will be more sub-units affected by it (this why penalty is raised to 2%), but they're weaklings in any case.

(in reply to lloydster4)

Post #: 11

RE: Variable Combat Odds - 7/20/2020 9:56:12 PM

DTurtle

Posts: 443
Joined: 4/26/2010
Status: offline

I reported this some time ago in the Tech Support forum.
This was Vic's reply:

quote:

ORIGINAL: Vic

Those odds are just an estimation guys. And it is baded on simulating 1 combat round of battle. Changing the order of the units could result in slightly different results in that simulation since a lot of random rolls are involved and something like the butterfly effect might occur.

So basically there is a somewhat fixed seed for the RNG for the single combat round simulated to get the estimated modifiers. Changing the order of adding units, changes what rolls are made when and therefore gives other results. The biggest variability is always the commander rolls. So don't only rely on the odds shown, but also take a quick look at the modifiers.

All the modifiers are simply multiplied with each other. So the difference between rolling +8%/+10% (multiplier 1.18) or +23%/+23% (multiplier 1.51) on the commanders is very significant. This is especially notable once the total multipliers get quite high.

(in reply to lloydster4)

Post #: 12

RE: Variable Combat Odds - 7/20/2020 10:28:20 PM

demiare

Posts: 470
Joined: 6/20/2020
Status: offline

quote:

ORIGINAL: DTurtle
All the modifiers are simply multiplied with each other. So the difference between rolling +8%/+10% (multiplier 1.18) or +23%/+23% (multiplier 1.51) on the commanders is very significant. This is especially notable once the total multipliers get quite high.

Don't forget armor supremacy effects. -50...-80% is a HUGE modifier. This is why mechanized forces are trump even being so expensive.

Full encirclement & entrenchment in very dependable location (like a ruins) also more important then commander rolls too.

Sure, commanders are great but don't rely on them completely or sometimes you will get a surprise.

(in reply to DTurtle)

Post #: 13

RE: Variable Combat Odds - 7/21/2020 12:46:24 AM

lloydster4

Posts: 164
Joined: 6/19/2020
Status: offline

quote:

ORIGINAL: demiare

quote:

ORIGINAL: lloydster4

Demiare, I understand the concept behind the Max-Attacked penalty. But look at OP's example. Why would I be talking about a "generic situation"?

The most significant changes from the 1st to 2nd example are, again: Attacker Command Bonus and Defender Readiness Penalty.

How would these stats be affected by the max-attacked penalty? That doesn't make sense.

In OP's example, Max-attacked penalty is -1% in the 1st screen and -2% in the 2nd. Absolute peanuts.

Easy. His troops are high in numbers but VERY weak (half of them militia! no tanks, no buggies, no guns - useless infantry only). His major strike power are walkers. Yes, only 10 walkers (equal to 100 men) but they're his main and only power.

In first example walkers get massive attack penalty because of tons of useless bodies send in front of them. This penalty affect only 1% of his attacking forces - 1% counted in manpower, not in a firepower.

In second example his major and only firepower fight without restrictions, while some useless militia guys eating penalty. Yes, there will be more sub-units affected by it (this why penalty is raised to 2%), but they're weaklings in any case.

I feel like you're not even attempting to understand the situation

(in reply to demiare)

Post #: 14

RE: Variable Combat Odds - 7/21/2020 9:17:12 AM

demiare

Posts: 470
Joined: 6/20/2020
Status: offline

quote:

ORIGINAL: lloydster4

I feel like you're not even attempting to understand the situation

No, instead you don't understand that modifiers in estimation are only global one. In first example walkers sub-unit receive -50% or more penalty for outnumbering - sure, they're only 1% of OP attacking forces (this why we see -1% penalty for outnumbering), but outnumbering penalty is personal for units. Without half or even more Sentinel's firepower his attacking forces can't do anything against tanks.

Lol, I played with bunch of Sentinels few times. In similar situations you NEED them to attack enemy first. Same will be with tanks - poor idea to send a tank unit last if you're outnumbering enemy. There is too high difference in firepower between Sentinels and starting units.

< Message edited by demiare -- 7/21/2020 9:19:22 AM >

(in reply to lloydster4)

Post #: 15

RE: Variable Combat Odds - 7/22/2020 4:35:20 AM

lloydster4

Posts: 164
Joined: 6/19/2020
Status: offline

quote:

ORIGINAL: demiare
No, instead you don't understand that modifiers in estimation are only global one. In first example walkers sub-unit receive -50% or more penalty for outnumbering - sure, they're only 1% of OP attacking forces (this why we see -1% penalty for outnumbering), but outnumbering penalty is personal for units. Without half or even more Sentinel's firepower his attacking forces can't do anything against tanks.

The max-attacked value for motorcycles is 6, so the max-attacked penalty would apply to the 7th received attack--scaling up to 50% on the 12th attack.

In OP's example, the attacker has 16 sub-units and the defender has 7. If the attacks were evenly distributed, then each motorcycle would only receive 3-4 attacks on the first round.

While it is possible that one of the enemy subunits could get attacked 7 times, is is exceedingly unlikely. Hence why the estimated max-attacked penalty is 1-2%.

There is something else that is affecting the odds.

(in reply to demiare)

Post #: 16

RE: Variable Combat Odds - 7/22/2020 2:45:10 PM

DTurtle

Posts: 443
Joined: 4/26/2010
Status: offline

quote:

ORIGINAL: lloydster4

There is something else that is affecting the odds.

In the example posted the largest difference comes from worse readiness rolls for the defender. Because of that the overall modifier goes from -53% to -63%, lowering the modified strength from 19 to 15 (roughly 20% less).
The other large difference comes from the leader rolls for the attacker, increasing the modified strength a bit from 83 to 97 (roughly 15% more).

As a side note: experimenting a little, I couldn't find any direct mention of a firepower decrease due to overwhelming firepower in the detailed combat logs. It is probably reflected elsewhere of course. In comparison, an overstacking penalty is applied evenly to all units, and should definitely be avoided.

Edit: See my post below, the firepower decrease is applied.

< Message edited by DTurtle -- 7/23/2020 5:26:41 PM >

(in reply to lloydster4)

Post #: 17

RE: Variable Combat Odds - 7/22/2020 2:53:30 PM

demiare

Posts: 470
Joined: 6/20/2020
Status: offline

quote:

ORIGINAL: DTurtle
In comparison, an overstacking penalty is applied evenly to all units, and should definitely be avoided.

Yes, it's a funny quirk of overstacking penalty calculation.

It's calculated based on units order & sub-units count, but when you trigger it for example with last MG battalion - it's effect applied to sub-units inside that battalion by their TYPE and not ID.

So in result EVERY MG sub-unit in your whole attacking army receive a penalty.

At least it's my understanding on how it's working after doing similar set of experiments. After discovering this I'm trying to increase diversity of models used in my assault army (by including stuff like mech.arty / assault guns / etc), so I can dump a penalty on them when I really need to encircle powerful enemy army.

(in reply to DTurtle)

Post #: 18

RE: Variable Combat Odds - 7/22/2020 3:55:47 PM

lloydster4

Posts: 164
Joined: 6/19/2020
Status: offline

quote:

ORIGINAL: DTurtle
In the example posted the largest difference comes from worse readiness rolls for the defender. Because of that the overall modifier goes from -53% to -63%, lowering the modified strength from 19 to 15 (roughly 20% less).
The other large difference comes from the leader rolls for the attacker, increasing the modified strength a bit from 83 to 97 (roughly 15% more).

This is correct. But I can't understand why these would change. I assumed, perhaps falsely, that those were both fixed pre-conditions.

Is this a quirk of the prediction system, or can I actually increase my leader rolls by fiddling with unit order?

(in reply to DTurtle)

Post #: 19

RE: Variable Combat Odds - 7/22/2020 4:22:49 PM

demiare

Posts: 470
Joined: 6/20/2020
Status: offline

quote:

ORIGINAL: lloydster4

Is this a quirk of the prediction system, or can I actually increase my leader rolls by fiddling with unit order?

AFAIK "real" rolls will happen only when you begin combat. Before it they're only predictions - this why they're changing.

(in reply to lloydster4)

Post #: 20

RE: Variable Combat Odds - 7/23/2020 5:25:39 PM

DTurtle

Posts: 443
Joined: 4/26/2010
Status: offline

quote:

ORIGINAL: demiare

Yes, it's a funny quirk of overstacking penalty calculation.

It's calculated based on units order & sub-units count, but when you trigger it for example with last MG battalion - it's effect applied to sub-units inside that battalion by their TYPE and not ID.

So in result EVERY MG sub-unit in your whole attacking army receive a penalty.

At least it's my understanding on how it's working after doing similar set of experiments. After discovering this I'm trying to increase diversity of models used in my assault army (by including stuff like mech.arty / assault guns / etc), so I can dump a penalty on them when I really need to encircle powerful enemy army.

Some more testing results:
Overstacking (too many units from too few sides - the stack points shown in the top right): Negative multiplier to ALL units, no matter what order they attacked in, no matter the type of unit attacking.
Too much firepower: Applied in combat when several units attack the same unit. Applied in the order the units attack. Note, that this is NOT necessarily the order the units were added to the attack queue by the player! Tanks for example will usually attack before infantry, even if added later. It will go up to a -80% multiplier. This looks to be evaluated after every attack on a unit.

quote:

ORIGINAL: lloydster4
This is correct. But I can't understand why these would change. I assumed, perhaps falsely, that those were both fixed pre-conditions.

Is this a quirk of the prediction system, or can I actually increase my leader rolls by fiddling with unit order?

Its a quirk of the prediction system. The prediction is made by simulating a single combat round. Changing unit order changes something with the order the rolls are made, which results in different results shown.

Fiddling with the unit order shouldn't make any difference for the actual attack itself. However, I have not extensively tested this.

(in reply to lloydster4)

Post #: 21