Cav ZoC/AP Mod Testing Workshop

[stockwellpete]

I'm sympathetic with this idea. Sometimes the battles end up looking like a bunch of table top pieces just scattered all over and you couldn't really describe it as having had a main line, or particular cavalry flank, or even wings really. Not always of course, or even most of the time and maybe some battles really ended up like that so it's fine, but it would be interesting and challenging to think about what changes might give the battle a more...I'm not sure, cohesive feel?

It usually happens in close battles when both sides go to over 40% routers as there's simply not enough units left. So lowering auto-lose threshold to something like 35-40% would solve this issue (and would, arguably, be more realistic).

That said, I'm pretty sure, though, that auto-lose is set so high deliberately for gameplay reasons, and wouldn't (and shouldn't) be changed outside of mods.

Exactly.

Richard, would you accept a lower army rout threshold for user made scenarios in future? I am thinking maybe 50% instead of 60%.

[stockwellpete]

I have done a test game this morning against the AI using the latest v1.4, Bretons v Normans and I picked a completely cavalry army for Normans. Goodness was it brutal.

At one point I just sat there munching a biscuit as my cavalry just tore up the Breton position. One unit of Norman lancers flank charged a Breton cavalry unit (that was already fighting another unit of Norman lancers) got the automatic disruption and then broke it as result of combat, it then carried on into the flank of another occupied Breton cavalry unit and did exactly the same thing, and then it charged a mob unit and disrupted it! All in one turn! And all the other Norman cavalry units released by these routs then went charging off into the flank of the Breton infantry position causing further disruptions.

Btw I used the smallest size armies and battlefields to speed up my testing. I'll have another go with this v1.4 in a minute.

[stockwellpete]

I have done a second game with v1.4, Bretons v Normans again.

My first take on this is that the "added extra mandatory CT on charge for non light cavalry vs non light cavalry" is far too much. I think "something going wrong with the charge" is already covered by the RNG when you get a bad roll. In this game I had a Norman lancer unit charge downhill at a Breton lancer unit and it failed its mandatory CT test and disrupted. It then lost the impact combat badly as a consequence and fragmented. The only saving grace being that it then broke off. So then a different Norman lancer unit (with a general) attacked this same Breton unit - and the result was exactly the same! So now I have 2 fragmented Norman lancers.

So at the moment I prefer . . .

Non-light Infantry units charged by non light cavalry this turn a) lose their secondary (NOT Primary) ZoCs this turn only and b) lose 4 AP for the next turn (their turn to move).

additional v1.1 changes:
-fixed logic that was permitting a cav unit to fallback from a previous turns combat and then, now having ap on its turn, charge past the non light infantry it fell back from.

additional 1.2 changes:
-increased AP loss from 4 to 6 so diagonal facing ZoC trap doesn't work either

additional v1.3 changes:
1) extra -1 to ct for non light cav vs non light cav on impact and in melee
2) 33% more casualties in non light cav vs non light cav on impact and in melee

[Athos1660]

Is there a way to know the approximative probability of a cohesion loss with...
- a 'total modifiers' = +1 ?
- a 'total modifiers' = +2 ?
...when the existing morale state is :
- Steady ?
- Disrupted ?
...without any casualties.

= the kind of CT that happens because of a wounded/killed General or in my charge mods.

(a probability or a 'it is twice more likely to...")


Is it possible to use numbers with decimal (1.3 or 2,1) as CT modifiers ?

(edit)
underlined

[rbodleyscott]

Is there a way to know the approximative probability of a cohesion loss with...
- a 'total modifiers' = +1 ?
- a 'total modifiers' = +2 ?
...when the existing morale state is :
- Steady ?
- Disrupted ?
...without any casualties.

= the kind of CT that happens because of a wounded/killed General or in my charge mods.

(a probability or a 'it is twice more likely to...")

https://steamcommunity.com/sharedfiles/ ... =735499412

The numbers are the same for FOG2 Cohesion Tests.

The reasons for the modifiers make no difference, all that matters is the total of all applicable modifiers. (Bearing in mind that some of them can't total more than -1 however many of those modifiers apply)

Is it possible to use numbers with decimal (1.3 or 2,1) as CT modifiers ?

No. Nor for anything else. The engine only handles integers.

Of course you can get round that by changing all of the modifiers to be an order of magnitude larger, but you would need to mod more of the code to fit the revised values.

[Schweetness101]

My first take on this is that the "added extra mandatory CT on charge for non light cavalry vs non light cavalry" is far too much. I think "something going wrong with the charge" is already covered by the RNG when you get a bad roll.

yeah I agree. The auto ct is an interesting idea from a bottom up design standpoint sort of, but I don't think it is really necessary, and is way too much when paired with the extra casualties and -ve CTs for non light cavalry combat.

So at the moment I prefer . . .

yeah, I think the ZoC and inf AP changes are good for now. The main question for the cav mod then I think, before combining it with the flank angle mod, is what exact values to use for the modifiers. Currently the mod is:

Impact:
Casualties: Increased 33%
Morale Modifier: -1

Melee:
Casualties: Increased 33%
Morale Modifier: -1

(ie they are the same for impact and melee)

Do you think they should be the same and we should just move forward with those values, or should we consider something like (just as an example):

Impact:
Casualties: Increased 40%
Morale Modifier: -2

Melee:
Casualties: Increased 10%
Morale Modifier: -1

to emphasize the greater casualty and morale impact of the cavalry vs cavalry charge specifically, kind of in the spirit of Athos' idea, but without adding any extra pre charge CT phase or anything like that?

On the one hand, making the impact and melee phases have different modified values adds more complexity and more rules to remember, and also begins to resemble 'bottom up' design, which is anathema to FOG2's design in general as far as I know, but might give a bit more realistic feel to the simulation if it's not just that cavalry combat resolves faster, but that the impacts are particularly more deadly.

In the end what matters more is the result (cavalry combat resolved more quickly) than the realism of the mechanics, so unless doing different values on impact and melee adds something extra to the final result I'd say just do the same value increases for both impact and melee, but I could see it going either way.

[Schweetness101]

one way to ask the above question is, how long should non light cav vs non light cav combat take in general?

With vanilla CT and casualty figures, doing a 10 vs 10 of armoured muslim lancers(did this test as a series of 1v1 duels without sending now freed up units to flank so of course takes longer than a real game):

first fragmentation on red turn 3, broke on blue turn 3
another disruption on red turn 4 and another on blue turn 4
second break on red turn 5, another disruption on bue turn 5
3rd break on red turn 6 and another couple disruptions, and some more disruptions on blue turn 6
3 more fragmentations on turn 7, but still only 3 of 20 units have broken
4 more broken on turn 8, mostly autobreaks, and game ended 20% to 50% on blue turn 8

and with the 1.3 figures of 33% casualty increase and -1 ct modifier on impact and melee:

by end of second turn there were 2 disruptions and 2 fragmentations
by end of red turn 3 there were 2 breaks, 2 fragmented units and 2 disrupted units, and by end of blue 3 there were 3 broken and another fragmented
by end of turn 4 there were 5 broken units, 2 disrupted, one fragmented
another 2 broke on turn 5 and one more fragmented, leaving 7 units total broken and only 3 duels remaining (so on turn 5 have same number broken as vanilla on turn
9 were fragmented by end of turn 6 and the only remaining duel had a fragmented unit

of course the armor and superior quality of the tested unit has an effect, and there would need to be more testing with non lancers and non superior quality cav to see how much faster or slower it is comparatively. A lot of the reason for that is the probability distribution of a dice role, where getting a -1 CT from 11 to 10 is not increasing your odds of breaking as much as a -1 to CT from 7 to 6, just for example.

Because, as mentioned, in a real game you would be getting flanks in once some units had broken, the above is not a perfect simulation but just used to give an idea. Based off that sort of thing, would you recommend greater or lesser changes to the casualty and CT modifiers? I guess really we'd have to test in real games to know, but let me rephrase as, is this sufficient change to go ahead and combine with the flank angle mod and start testing it all together?

[Athos1660]

Is there a way to know the approximative probability of a cohesion loss with...
- a 'total modifiers' = +1 ?
- a 'total modifiers' = +2 ?
...when the existing morale state is :
- Steady ?
- Disrupted ?
...without any casualties.

= the kind of CT that happens because of a wounded/killed General or in my charge mods.

(a probability or a 'it is twice more likely to...")

https://steamcommunity.com/sharedfiles/ ... =735499412

The numbers are the same for FOG2 Cohesion Tests.

The reasons for the modifiers make no difference, all that matters is the total of all applicable modifiers. (Bearing in mind that some of them can't total more than -1 however many of those modifiers apply)

Is it possible to use numbers with decimal (1.3 or 2,1) as CT modifiers ?

No. Nor for anything else. The engine only handles integers.

Of course you can get round that by changing all of the modifiers to be an order of magnitude larger, but you would need to mod more of the code to fit the revised values.

Very informative. Thank you very much

[stockwellpete]

yeah, I think the ZoC and inf AP changes are good for now. The main question for the cav mod then I think, before combining it with the flank angle mod, is what exact values to use for the modifiers. Currently the mod is:

Impact:
Casualties: Increased 33%
Morale Modifier: -1

Melee:
Casualties: Increased 33%
Morale Modifier: -1

(ie they are the same for impact and melee)

Do you think they should be the same and we should just move forward with those values, or should we consider something like (just as an example):

I think we should go with this as it is straightforward. The only question for me is whether casualties in both impact and melee could be increased by 50% instead of 33%. Reducing the time a cavalry unit gets to auto-rout I think will help us with the broader aim of speeding up cavalry melees in relation to infantry melees (given that cavalry units are often only 240 men.

[stockwellpete]

Because, as mentioned, in a real game you would be getting flanks in once some units had broken, the above is not a perfect simulation but just used to give an idea. Based off that sort of thing, would you recommend greater or lesser changes to the casualty and CT modifiers? I guess really we'd have to test in real games to know, but let me rephrase as, is this sufficient change to go ahead and combine with the flank angle mod and start testing it all together?

Yes, I think we can combine them now. As I was saying, the smallest format of the game allows you to test things very quickly. Breton v Norman is quite a good match-up for our purposes, I think.

[Schweetness101]

Because, as mentioned, in a real game you would be getting flanks in once some units had broken, the above is not a perfect simulation but just used to give an idea. Based off that sort of thing, would you recommend greater or lesser changes to the casualty and CT modifiers? I guess really we'd have to test in real games to know, but let me rephrase as, is this sufficient change to go ahead and combine with the flank angle mod and start testing it all together?

Yes, I think we can combine them now. As I was saying, the smallest format of the game allows you to test things very quickly. Breton v Norman is quite a good match-up for our purposes, I think.

ok cool and where are you thinking of starting with the flank angle mod in the combination mod? ie combine cav zoc with a flank angle mod version with no flanks at all, flanks for cav vs inf, flanks but without autodrops and just poa modifiers etc...

[Athos1660]

(...) to emphasize the greater casualty and morale impact of the cavalry vs cavalry charge specifically, kind of in the spirit of Athos' idea, but without adding any extra pre charge CT phase or anything like that?
(...)

btw my idea has never been to "emphasise the great casualty (...) impact of the cavalry vs cavalry charge specifically".
The values for casualties for both impact and melee in my mods are Vanilla ones.
My only idea is to test the morale cohesion et the physical cohesion/density of the unit during the charge.
Thus the use of only CTs.

You guys want to tremendously increase casualties to shorten the times of cav fights.



___
I don't know whether 'physical cohesion of the elements of a group' can be said in English...

[Athos1660]

(double post...)

[stockwellpete]

ok cool and where are you thinking of starting with the flank angle mod in the combination mod? ie combine cav zoc with a flank angle mod version with no flanks at all, flanks for cav vs inf, flanks but without autodrops and just poa modifiers etc...

Flanks without autodrops but with POA modifiers is how I am thinking at the moment. I am not 100% sure about POA's against units that are not engaged with the enemy but you can leave them in as per vanilla now. I just feel that an unengaged group of fighters would turn to face any enemy that approached them but we can sort that later if we feel a change from vanilla is necessary.

[Schweetness101]

ok cool and where are you thinking of starting with the flank angle mod in the combination mod? ie combine cav zoc with a flank angle mod version with no flanks at all, flanks for cav vs inf, flanks but without autodrops and just poa modifiers etc...

Flanks without autodrops but with POA modifiers is how I am thinking at the moment. I am not 100% sure about POA's against units that are not engaged with the enemy but you can leave them in as per vanilla now. I just feel that an unengaged group of fighters would turn to face any enemy that approached them but we can sort that later if we feel a change from vanilla is necessary.

and same for cav and inf? or let cav flank just one another? or leave cav flanks the same as vanilla?

my preferred is to continue to permit cav to get flanks in on one another and on infantry, giving cav a kind of special flanking status, but maybe it should be more complex. There is a lot of wiggle room between +50 poa and +200 POA and auto drop.

[stockwellpete]

and same for cav and inf? or let cav flank just one another? or leave cav flanks the same as vanilla?

I think we will eventually have to test both - no flank attacks at all just with POAs for both cavalry and infantry - and no flank attacks for infantry v infantry (just POA's) but flank attacks retained for cavalry v cavalry. I am not sure what we do about cavalry v infantry in this second variant (just POA's I assume?).

[Schweetness101]

should I make it so that make it so movement/AP loss to infantry only happens if the cav charged them from the front? Those are the only relevant ones for the inf cav trap.

[stockwellpete]

should I make it so that make it so movement/AP loss to infantry only happens if the cav charged them from the front? Those are the only relevant ones for the inf cav trap.

Yes, OK.

[stockwellpete]

Another issue I have with cavalry melees is the pursuit mechanism where, at the moment, a victorious cavalry unit will chase off its routed or evading cavalry opponent, but will often ignore the ZOC's of other enemy units (infantry or cavalry) in the vicinity. So they can look like a bit like a skier slaloming its way across the battlefield. I don't like that effect very much, particularly when you get units from both sides doing it in the same melee and then they have to turn back around to engage with each other again. That slows things right down. So I think enemy ZOC's (infantry or cavalry) should stop a cavalry unit pursuing a defeated or evading enemy.

Again I would make the point that I made in the flank attack mod that "in "real life", soldiers tended to fight as contingents or "battles" (in the medieval period), not as separate units", so I don't think individual cavalry "units" should be pursuing enemy "units" out of a melee if that melee is still going on around them. Obviously the permutations of unit configurations in melees are endless, so the trick is to come up with something simple that works in most cases. One idea I have (and it may be no good) is that a victorious cavalry unit (that is no longer in melee with anything else) first of all moves into the square of the defeated enemy unit. If that brings it into contact with another enemy unit then it stops there immediately and begins to fight that next enemy unit. If there is no enemy unit to fight from that new square then it may pursue its defeated enemy subject to the ZOC restrictions I have mentioned before. I have no idea at all if this is easily moddable, but the purpose of it would be to make cavalry melees more decisive (by that I mean it would help to resolve all issues in one particular area of the battlefield), so that the victorious side would then have a chance of supporting their army in the centre.

Further to this post I made a few days ago, I still stand by the first paragraph completely, but I think the idea contained in the second paragraph does not go quite far enough. I have highlighted the key point above in bold text.

For example, you have 2 Norman lancers fighting 2 Breton lancers in a big open space, and one Breton lancer routs as a result of melee combat. Under the idea suggested above the victorious Norman lancer unit would first of all occupy the space previously held by the defeated Breton lancer and then check to see if there was another enemy to fight before it could pursue. The only other enemy unit in the vicinity is the other Breton lancer unit which is now adjacent (but obviously facing the other way and fighting the other Norman lancer). So unless the victorious Norman lancer has the ability to make a 90 degree turn in these circumstances to join the melee against the remaining Breton lancer it has no viable target and would pursue (defeating the intention of the modification). The alternative, which is much simpler, is that the victorious Norman lancer should check to see if there is another melee opponent before it moves at all to pursue. In this example, the victorious Norman lancer would have another opponent to fight and so would join the melee with the other Norman lancer against the remaining Breton lancer. There would be no impact phase in this situation, just a 2v1 melee. (I think this idea might also be applicable to shock infantry as well.)

I think the big advantage of this idea is that it would speed up overall cavalry melee resolution considerably. You wouldn't have situations where units from both sides end up pursuing vanquished units and then have to turn around and come back at each other again. In a greater proportion of cases, it would be a short, sharp fight where one side or the other would win the cavalry battle outright and then have the opportunity to mount a genuine flank attack on the centre.

[stockwellpete]

I noticed one small thing in a test. An enemy skirmisher evaded one of my cavalry units but moved into a position where it could be rear charged by another cavalry unit of mine on the same turn. It did not suffer automatic cohesion loss on the charge.