[Balance]Automatic retaliation and titans/top tiers units

[Galdred]

I think the unlimited retaliation is a bit too much. It made sense in operational and above games (where attacks were more abstract and involved assaulting a hex), but with the weapon ranges, and different power levels of the units, it creates some very weird, and IMO, unwanted behaviours (from the AI, but it is playing perfectly reasonably in this matter) :

firing at a unit gives it a free retaliation attack if it is in range. Thus, firing at a unit with very high firepower you have very little chance of damaging (eg a titan) is just a free attack you are conceding to it. This results in the AI not targeting my titan at all (in the scenario Obedience).
Note that I could not get to the end of the mission due to my installation crashing repeatedly around turn 5 (I guess there are some anti titan units near the objective).

Firing with 4 units on a titan would give it 5 times its normal firepower, without making a dent in it, so the AI is correct in not doing that, but it doesn't work too well at this scale(grand tactical) IMO.
Even for non titan units, this can be a problem (basically, it makes higher tier units impervious to lower tier ones, as them attacking would just increase the firepower of the high tier unit a lot).

I would suggest limiting the retaliation attacks one way or another. There are lots of way to do so : decreased efficiency for retaliations, no retaliation at all, 1 retaliation attack only, or only 1 retaliation attack if the attack had not been used last turn, which would make it a bit like overwatch.
I would favor the last solution, or even a true overwatch mode(if the attack had not been used), but I cannot really tell which would work better without testing. I don't think this should apply for assault (it would not take too much time to the titan to stomp the units, and more generally, units make themselves more vulnerable in an assault).

I have not understood the way ammunition work yet, but even if it is the limiting factor to retaliation, then we would have another problem (the titan wasting all volcano ammunitions while retaliating against worthless targets), while not solving the core issue (the firepower of the titan would still be multiplied by the number of units attacking it).

Note that this problem much predates Armageddon, as it was already present in Battle Isle ( which is pretty similar in scale).

Another problem is that it creates a weird effect when it comes to unit size :
5 platoons of 10 guys firing at a target will induce 5 reaction fire, while an ork mob of 50 Boyz would only take 1 reaction fire.
I don't think it is very consistent with the scale ( and it penalizes tanks a lot against AT ork bands as they typically have a small unit count).

[Kerensky]

Thanks for your insights!

Just to let you know, we have some measures in place to cushion this effect. Several of the heaviest and most powerful weapons have a trait attached to them that does not let them fire in retaliation events. This not only prevents excessively powerful retaliations, but also preserves ammo on extremely heavy weapons from being wasted on things like Ork infantry fodder. Even so, as you noted, the AI refuses to engage the Titan in the Obedience scenario, which is a pretty serious concern. This is not only a result of the Titan's firepower, but also it's defense ratings. AI does not like to attack if it sees it's prediction doesn't result in damage.

Right now the focus is still on early game and early unit interactions, but Titan balance and AI effects is something we are aware of.

[Galdred]

But even in low tier weapons, it causes severe balance problems, as it makes bigger mobs inherently much stronger than smaller ones(the 50 strong ork band vs 5*10 squads of orks firing at a single target exemple in my previous post).
That makes imperial doctrine look stupid (They should definitely copy the orks and go for 50 men strong platoons), and make unit much harder to balance ( as unit power is no more a function of defense and offence) :
A 50 strength unit only gets 1 volley of returning fire when 5 10 strength units get 5, while firing at the same target.
A 50 strength unit retaliation will be much stronger than the 5*10 strength units because assuming the attack causes 5 casuaties, it will retaliate at 45, then 40,then 35, while the 5*10 St units will retaliate at 5, then not at all, then 5 again.
This is partly mitigated by the 50 st unit could be easier to break (but I suppose casualties induced suppression depend on the ratio cas/initial size of the unit).

The high body count effect can also be mitigated by either only allowing 1 retaliation strike, or making each one less powerfull (like halving the returning fire for each retaliation, which would cap retaliation at 2*base retaliation strength).

[Kerensky]

While I do agree the sheer number difference is potentially very unbalancing, there are a number of systems in place to even this out. Rate of fire and accuracy (lower for large formations, higher for small formations) are key stats to control this strength difference. In fact the internal debate on how to best use these to achieve balance has been quite heated. Suffice to say that just because Orks number in the 30-40 range and Imperials number 15-20, their net strength evens out much closer than their face value unit count would initially indicate.

[Galdred]

What about toning down accuracy for each subsequent retaliation? Given that retaliations are what make these units uneven, anything that diminish the efficiency ( or cap the number of ) multiple retaliations would diminish the urge to have other elements to balance larger squads.

In Age of Wonders 1, units could retaliate in melee an unlimited number of times too, which made high level heroes able to solo armies quite well ( I did this for the last missions of the campaign). They ended up making retaliation eat action points from the next turn to tone it down in AoW2 ( I don't think it is a very elegant solution, though), which kind of acts like a hard cap (the problem being that you end up not controlling your hero much after he gets attacked).

[IainMcNeil]

Reducing accuracy for each retaliation is an interesting idea. Each attack could reduce the accuracy by 25%, so the first retaliation is 100%, the second 75%, the third (75%*75%) and so on.

This would give an added gameplay effect of making a bigger difference which unit you attacked with first.

[thepuffin]

Retaliation is definitely not a very enjoyable mechanic at the moment, especially with the power of heavy, fast-firing artillery like the Griffon. One mortar unit can end up "retaliating" against an entire front's worth of units in a single turn, and then fire itself in the AI turn. I realise you don't want to have to ask the user every time whether they should retaliate, but when you assign the Fire Support attribute to artillery you are giving them unrealistic power.

I suggest limiting Fire Support retaliation to a range of a single hex.

[Rudankort]

We need to separate retaliation from support fire here. Retaliation just means that when you attack enemy unit, it shoots back. To be quite frank, I don't think this mechanic is realistic on any scale, be it operational or tactical. A unit cannot suddenly get several times more shots than when not attacked, at least not with its full power, as is the case in Panzer Corps and many other games. However, this mechanic is important for gameplay, and ultimately results in more realism than lack of it. Removing retaliation with a crappy unit and then attacking at will (wide-spread tactics in Heroes and other similar games) is not realistic, and breaks the balance. When you know you will be retaliated, this forces you to carefully pick location (cover, range, los for indirect weapons) for your units and order in which they attack. Nor can you just concentrate the fire of all your units on a single enemy and wipe it out before it can do anything. With all-ranged combat this becomes even more important than in Panzer Corps like games.

As for support, there is definitely some food for thought here. For example, for heavier support units, like artillery vehicles, we could try to limit support fire to just one instance, while lighter support units might have more.

[Galdred]

Actually, it is much easier to justify retaliation at the operational level, when there is no really ranged attack : most of the "time" in a turn is spent moving, getting resupplied, or preparing for an assault. The assault itself consumes a tiny fraction of the time, so an unit being assaulted several time would insdeed have no problem fighting back each attacker (now, the problem would be the impossibility to coordinate multiple assaults of several units at the same time, but doing so would drastically increase the complexity of the game, almost to The Operational Art of War level, so the PanzerCorps system abstraction feels just fine).

On the other hand, on the tactical scale, when units already start at range fron each other, the attacks do not represent the whole supplying, preparation, and assault process, only the simple act of acquiring a target and firing. That is, actual combat represent more or less 100% of the attacking action (that is why retaliation from assault is less than a problem for me, as in an assault, preparing and moving to the target position become relevant again).

Now from a gameplay point of view, I agree that the Heroes of Might and Magic system is also far from ideal (I mean, the optimal way to assault is indeed to split an unit 1/99 so that the single element attacks first and die from retaliation, while the 99 left attack without retaliation. That could be remedied by not several ways not using the retaliation if there is overkill, or using only a fraction of it).

But many games at the tactical level do without automatic retaliation at all :
Steel Panthers, John Tiller's Campaign Series, XCom, Jagged Alliance, Final Liberation, Chaos Gate, same for miniature or board games : Descent : Journeys in the Dark, Warhammer 40K, epic...

Retaliation from melee is a more divisive subject :
Steel Panthers doesn't differentiate 1 range assault from other attacks, Warhammer 40k epic 40K , FInal Liberation and Chaos Gate all allow the defender to kill the attacker in melee, BUT Final Liberation gives attacker bonus/ defender malus to subsequent attacks ( so yes, you can destroy a warlord titans in melee with enough Gretchins), while epic (whether it is Space Marines or epic 40k) and Chaos Gate all allow combined attacks (3 attackers attacking at the same time).

You still have to carefully pick the location when attacking in these games : It still costs you one action, which is very precious, and the unit will be fired at by the target or other units on its next turn, and all of these games have reaction fire (you spend one action, and shoot at the first unit doing an action in range on its turn whether it is attacking, or moving).
In order to avoid the first player to wipe out second player, these games use something else, on top of overwatch fire :
Most (XCom, Final Liberation, SteelPanthers, Jagged Alliance, Chaos Gate) have units start far apart enough so that almost no one is in range on turn 1 ( so being first player does not allow you to fire with everyone), other only allow a small subset of your units to be activated at the same time (Final Liberation has such a mode, Epic works like this too), but that would make asynchronous play a slog.

But anyway, it is not an issue worth changing things so far along development, as the whole maps and units are built around this system, that is why I ended up recommending just toning down accuracy for subesquent retaliations.
It is an interesting mechanics indeed, but it makes the game feels more abstract, more puzzly, while the games I mentioned have a simulation feel to them which makes them feel more "real" for a lack of a better term.

[Kerensky]

To be honest, accuracy as a stat already bothers me quite a bit. Every time I feel units aren't performing well, I always seem to boost their baseline accuracy rating. Base accuracy is easy to lower on paper for unit balance, but true battlefield conditions with terrain, range, and LOS obstacles make accuracy drop like a rock.

Besides, isn't there a morale system also reducing accuracy? Morale, as I understand it, is that coloring of your unit strength from white to red, and as this gets further from white and closer to red, you suffer initiative and accuracy penalties. And morale definitely drops when you attack a unit repeatedly.

[Rudankort]

Galdred
Thanks for in-depth reply, there many things I could comment on, but maybe we better discuss this in a more calm atmosphere after release. For now, you are quire right - the decision has been made, but some penalties after several attacks on the same unit could be considered (I was thinking about initiative), to give "mass attack" some additional advantage. Not just for balance, but to give the player more tactical options.