NM War Resilience potential abuse

[Kensai]

The AGE engine has some known limitations regarding multi-factional conflicts as the one PON simulates. One of the problems we came up is sometimes nations get stuck with unreasonably high NM due to the way war resilience might work for the victor. In most AGE titles, the conflict is between two factions, so the NM ends up a zero sum game between two rivals: the victories of one are the losses of another so when NM raises for the one it lowers for another.

The problem we have encountered in our multiplayer is that since NM is a global value and war resilience works regardless if you are fighting the original nation you got your NM high, sometimes the NM is getting stuck high and stays high. Here's a potential mechanism

(1) Prussia's NM got unreasonably high by just beating a small nation in a regional war, Denmark. When I say unreasonably high I mean at the excess of 190.
(2) Prussia's NM then stuck high because Prussia got involved in a succession of other wars but since, as it seems, war resilience is working both ways, both for the victor and the defeated, it meant that Prussia entered these new wars with its NM stuck in 190. It was extremely easy for Prussia to dispose of its rivals and have a warscore of at least 0, meaning the NM was staying up there for FIVE LONG YEARS as it gave its armies and navies an unrealistic advantage over its rivals

I would like to file this as a bug. I know the engine has by design some limitations that might be hard to eradicate, but if there ever was an engine update for all AGEOD titles, it might be the case to apply a special rule adapted for PON's multi-factional conflicts.

War resilience for the victor should work only for values up to 120 or something, otherwise a nation (especially if played by a human) might abuse (on purpose or incidentally) its high NM won in another conflict to keep it "artificially high" in its future wars. At first instance it sounds as a correct mechanism and it should be like that. But a perennially stuck NM at 180-190 gives a surreal advantage of cohesion (which reigns supreme in all dealings, movement, battle rolls) as it both maximizes it and allows troops and ships to regain it back faster.

I want Pocus' feedback on the following (meaning acknowledging there might be a back door, as if indeed there is, we need to script some changes in our PBEM game; an engine update might come in the future for all AGEOD titles, if they it is fit):

(1) as far as the game allows a theoretical upper limit of 250, especially in PON which has many minor nations that you can conquer, it is imperative that NM falls down quickly enough if it has any value of more than 150. In our game, we set the objective regions of Denmark ourselves (through script) so when the Prussians captured Copenhagen they got this unreal spike in NM to 190, as they had captured the capital of a rival nation. But a spike of 90 for just defeating a minor nation is a tad too much. It has stayed that high for 5 years because of incidental wars of Prussia afterwards which were frankly deadlocks, nobody was winning convincingly (warscore near 0) until Prussia could bear its guns in a few turns and totally annihilate the other side thanks to the gargantuan cohesion bonus. So there is this PBEM-important loophole that a major nation defeats a minor one easily for the NM spike and then cruises on its ultra-high NM to score further victories against its more important rivals. For NM to fall from 190 to 150 (which is still huge by any standards!!) you need something like two years, considering that it usually falls one point each turn.

(2) war resilience for the victor should NOT kick in if the NM of the winning side (positive warscore) is anything above 120-130, otherwise the bonuses on the field from the cohesion, production, supplies advantages are so high that we have some pretty much impossible battle results where a nation attacks a fortified position by the sea, assaults, prevails, and reembarks on just two turn. All while losing little if any cohesion in an unrealistic way. Tests I've ran myself using a scripted decrease of the NM gives totally different results for said battles. Spikes are ok, even up to the theoretical max of 250, as mechanisms of one-time bonuses or ways to achieve instant (major) victories for some scenarios, but a constant and stuck NM of anything outside the range of 75-125 skews the results so much in a way it's not realistic. At least in multi-faction games. So it is imperative it decreases fast if anything above 125 and war resilience does NOT work for higher than that values.

In other words, we have an engine limitation that might be needed to be fixed. Maybe it only applies to PON, RUS, and AJE, which are the only games that essentially have more than 2 factions clashing and NM is not a zero sum game.

[Jim-NC]

There is also a question of if certain countries have higher or lower NM base points at certain times. In our game, Prussia was able to form the NGF and Germany during the middle of the wars. So for example, the NGF is in a war with Japan at this time (1870). During the war, the chain fires for German Unification. This causes the NGF to consume the last 3 minors, and to form Germany. This may or may not have had an effect on their NM or the base level.

My question Is: Is the base line NM for all nations the same at all times (ie 100)? I know that when a minor is absorbed by the major, the major gets it's stuff (diplomats, money, raw materials, etc.). Does it get change to it's NM, war weariness?

A secondary question is: Does government attributes effect NM changes? Thus if I am a full autocracy vs. a liberal democracy does this effect the changing of NM? What about regional power vs. global? Or maybe conscript army vs. professional?

Could these or other factors explain how a major country can keep NM high for years (above 150)?

[Kensai]

There is also a question of if certain countries have higher or lower NM base points at certain times. In our game, Prussia was able to form the NGF and Germany during the middle of the wars. So for example, the NGF is in a war with Japan at this time (1870). During the war, the chain fires for German Unification. This causes the NGF to consume the last 3 minors, and to form Germany. This may or may not have had an effect on their NM or the base level.

This is totally inconsequential. The conditions of the unification events do not demand peace to be set for them to fire. Their actions do not provide for any kind of NM bonus unless of course there are some objective regions which should give a small bonus (but nothing spectacular). Nothing that shoots the NM to 190 or so. And no, it does not boost any base level. The base level for all nations is 100 and any higher or lower than that should more or less "bleed" toward 100.

I am fairly attentive observer and I believe I have found the culprit in the war resilience mechanism. The problem is two-fold, so I believe there are two places an intervention in the code might be in order, at least in PON:

• regarding conquering the capital (and all objectives) of a nation: the NM bonus should never be so huge as to overshoot 150
• regarding war resilience: it should only work in values below 120, otherwise it could "stick" in an abusive way at high gear as a major nation defeats a minor and then maintains it by continuing wars that can be easily won without even caring for warscore as a score as low as +1 (barely winning!) will make war resilience kick in to hold NM from falling further down

My question Is: Is the base line NM for all nations the same at all times (ie 100)? I know that when a minor is absorbed by the major, the major gets it's stuff (diplomats, money, raw materials, etc.). Does it get change to it's NM, war weariness?

For every nation it's 100. The absorb command absorbs its tangible assets but not its prisoners and problems. The NM and war weariness should remain of the parent (absorbing) nation.

A secondary question is: Does government attributes effect NM changes? Thus if I am a full autocracy vs. a liberal democracy does this effect the changing of NM? What about regional power vs. global? Or maybe conscript army vs. professional?

There is an F1 attribute that directly influences war weariness and war resilience. It's the Nationalism level of your country. See in the manual for further information. But these are simply bonuses/maluses to the central algorithm.

Could these or other factors explain how a major country can keep NM high for years (above 150)?

No. The problem of a "sticking NM" lies to the war resilience algorithm itself, which can be abused, at least in PON where we have many different nations (factions) that may fight and conquer objectives. As NM is a global value, we have to make sure it falls back rapidly to a normal range, otherwise abuses can be seen:

1. Start a war with a minor nation.
2. Conquer it promptly, including the capital. Have your NM shoot to the stars (already a problem! why 190, couldn't 120-130 already be enough?)
3. Start a war with a major nation and just aim for a deadlock, you don't have to acquire a much positive warscore. As little as 1 is enough for war resilience to kick in and then it's trouble for your rival(s).
4. Beat them all with your army which fights as robots at NM of 190, virtually doubling the max cohesion and cohesion regain per turn.

[Jim-NC]

I was writing to the Phils, to see their take on the issue, and if they have placed something in the game.

[Sir Garnet]

Thanks Kensai for this clarified explanation of the potential flaw. We need to see if the potential mechanism postulated is the actual mechanism and if and how it is flawed.

Our Prussian player's report indicates the campaigns were not the overly easy run of victories assumed above. Much better cohesion increases performance - no surprises there. The only question is whether that cohesion is in part sustained by a faulty game mechanism. I don't have a chart showing the detailed progress in increments of Prussian NM correlated with on-map events, nor has anyone offered one, but my recollection as a neutral was that Prussia did seem to get an enormous amount of NM quickly in the early wars declared against it such as the conquest and occupation of Denmark. Maybe there was something about how Denmark's objective in Copenhagen was scripted that generated a large one-time or recurring NM bonus? Have substantial or recurring NM bonuses accrued for capture of any other minor nation capitals?

Kensai's description is based on there being additional factors at work beyond the explanation in the manual, among these a normal decline of NM toward the 100 base. Prussia reports that its NM has been declining while Kensai characterizes it as having been stuck (again, no chart available).

The manual states "Aggressive" countries such as Prussia/Germany and Sardinia-Piedmont have a positive resilience modifier more than Moderate or Elitist. Also Organized Majors have the highest resilience. Unorganized countries have the least, and more susceptible to war weariness. The modifiers for attributes are in the databse.

"War weariness has an important effect on NM, gradually reducing it even if fighting has ceased and one side is entirely occupied"

"If a nation’s NM is at least 1.5x the lose level and the country still controls its capital, war resilience will help keep NM at that threshold despite war weariness – however, defeats in battle can push it below the resilience level, after which war weariness will operate to push it down toward the lose level. At or below the lose level, each national region of the country loses 1 point of Contentment per turn and unrest will eventually force it to make peace."

The "threshold" here is that 1.5x the Lose level where resilience makes it sticky despite war weariness and note that it takes loss of NM from defeats (which obviously must exceed NM from victories) to push it down to the lose level.

It is a common sense question to ask: can a dev check these internal numbers such as "lose level" and resilience threshold for Prussia and see where they stand?

Kensai opines that war resilience should work only up to 120 or so. Based on prior discussions, I have assumed it is capped around 100 even for the more stellar countries, so if the 1.5xlose war resilience threshold is much higher it would be a surprise. For it to be over 150 the lose level would need to be over 100, which would have very unusual effects.

Isn't 100 the base toward which NM naturally drifts over time without other effects? Kensai says the decline is 1 NM per turn. This seems quite rapid. Is this correct, or what is the mechanism? I have not had the opportunity to have high NM recently or in the absence of events that would make it hard to notice any steady reversion.

As mentioned above re historical cases, such a reversion might interfere with historical behavior which demonstrated an extended time period over which nations might feel themselves superior and capable of anything. Looking at things historically, a circumstance of having a high NM for years is not unreasonable for Prussia/German Empire provided it enjoys a run of successes as it did against a number of countries in successive campaigns. The big attitude adjustment historically came only in 1918.

Reversion to the mean would also encourage players to go to war more frequently to keep NM high. 1 point of NM per turn would mean that say a spectacular 50 NM gain from a war in 1866 would have completely dissipated to 100 well before 1870.

But the actual question is whether NM is accumulating and adjusting and whether resilience is working correctly.

[Kensai]

Any test you will run will show the immense contribution of NM in cohesion. And cohesion in AGEOD games and PON in particular reigns supreme. All other factors equal, NM is a battle winner and this was hugely seen in the Prusso-Austrian war in our game where almost equal forces battled it out. It is no secret for any veteran AGEOD title gamer that taking the other guy's objectives while holding your own raises the NM and make the other guy's fall, eventually destroying him. This is not an issue when two nations fight it out. The problem starts with multi-faction conflicts where war resilience kicks in easily, sticking the NM to the already unreasonable high levels of NM for minor victories.

Also, regarding the NM that was stuck. It WAS stuck, when Prussia was having 0+ warscore, quite easy with all conflicts in the vicinity. It is a mere chance that Prussia has a negative warscore in the current conflict and that's because it is difficult to hurt Japan which is half a world away. It is not stuck anymore but since it started from 190 (for five years now), after almost two years of conflict it is still super high 150. Once warscore is positive, it seems that war resilience kicks in and the NM remains at where it was. Prussia almost managed to do this in our game.

War resilience should either not work for the victor above 120-125 NM or engage only when warscore is really high. To the excess of 50 or so.

[Pocus]

As you can see, War Resilience can't activate itself if morale is above 100...

Code: Select all

      // War resilience, a compensatory mechanism triggering for some nations, they are very difficult to beat except if hammered fast
      if Faction.HST_IsAtWar and (Faction.Morale < 100) and (Faction.Morale <= Faction.MoraleLose * 1.5) then
      begin
        for i := 1 to Faction.NationResilience + Faction.FM_GetFacModValue(_fmdWarResilience) do
          if Dice(100) > Faction.Morale then
          begin
            Faction.Morale := Faction.Morale +1;
            Faction.History.ChgOverTurnsData(_hisMoraleResilience, GS.cuTurn, 1);
          end;
      end;


[Jim-NC]

What about war weariness then? How fast does it drop back a high NM to 100?

[Pocus]

WW code is quite simple, crude even You lose 1 NM per WW per turn if at war.

[bjfagan]

Despite Kensai's comments to the contrary, and ignoring my statements, Prussian NM in the above (alleged abuse) example did go down. But due to victories over the different countries that attacked Prussia in succession, not together, helped to keep NM from falling. NM was somwhere in the 180's when I invaded Great Britain and took some areas there, pushing NM back up.

If there is a problem with NM at all, it is the amount it goes up or down for taking objective locations or special events.

[Pocus]

I understand that, but we can't really and realistically rebalance what all objectives give, plus it would not affect ongoing games.

I have changed the code though, this should help significantly maintains a reasonable morale. Before, you had a chance equals to (Morale-100) % to lose 1 NM, per turn. Now this test repeats itself until it is failed, so it will lead to the loss of several points a turn sometime.

[Jim-NC]

So in theory, you could go from 199 NM to 100 in 1 turn if the "dice rolls" go bad. And you should expect to always lose several points this way. This seems harsh to the player who just won something.

[Kensai]

I have changed the code though, this should help significantly maintains a reasonable morale. Before, you had a chance equals to (Morale-100) % to lose 1 NM, per turn. Now this test repeats itself until it is failed, so it will lead to the loss of several points a turn sometime.

This is terrific! Just what I thought would be a good course of action.

So in theory, you could go from 199 NM to 100 in 1 turn if the "dice rolls" go bad. And you should expect to always lose several points this way. This seems harsh to the player who just won something.

Jim-NC, once the NM falls to something hugely high but already less unrealistic (such as 150) the dice roll will have more chance to fail than to succeed, meaning prematurely stopping each turn. Once you go to something reasonable such as 120-125 (which is already high!!) I guess the chances are that your NM will remain high as the roll will mostly fail.

Anyway, given the serious consequences NM has on various factors in combat and supplying, I feel that there is an urgent need to set it to something reasonable. For how long should people be partying after a major victory? In our game for FIVE LONG YEARS Prussian forces were so psyched up that it had almost an arcade resolution of conflicts... nations cannot not and should not be in war without serious effects for such a long time. Pocus' solution fixes this and should be welcome. At least for PON.

[Kensai]

Btw, the opposite does not seem as important (a very low NM to rise rapidly towards 100) as it seems this happens anyway quite fast. Moreover I have noticed that when you get such a low NM (even <10) the effects are not so dramatic (as if cohesion cannot get too low) so it might work well with the current algorithm. Was this opposite rising algorithm a non-stopping one, as the one you now change, Pocus?

After all, the losing side will be getting its "war resilience" firing up once the winning side is reduced to less than 1.5 difference.

[Vezina]

I understand that, but we can't really and realistically rebalance what all objectives give, plus it would not affect ongoing games.

I have changed the code though, this should help significantly maintains a reasonable morale. Before, you had a chance equals to (Morale-100) % to lose 1 NM, per turn. Now this test repeats itself until it is failed, so it will lead to the loss of several points a turn sometime.

I would have to agree with Jim-NC on this. I think the test for multiple points is fine, as waiting some 90+ turns for the morale to drop from 190 to 100 is excessive. However, with no limit, this seems excessive as well. I would think the code as you have changed it, with a hard limit of 5 or 10 "1 NM drops" per turn would be fair. That way, i.e. losing 10 a turn, a country at 190 morale could enjoy a few months of heightened success, but it would come back down to normal fairly quickly without continued success.

[Kensai]

Yeah, but it could work on the other way as well. Lucky dice rolls might keep your NM high nonetheless. In any way, chances are that once your NM drops to 150, which is still very high, the dice roll begins to be cast in your favor. I feel that any turn above 150-160 in NM in multi-faction games is already too high... the abstracted cohesion boost is excessive.

[Vezina]

I get that and agree for the most part, but in a strategy game like this you need to be able to rely on things changing gradually and/or non-randomly. There's no way to strategically account for a 48 NM drop for no other reason than bad dice rolls. With a hard limit, you may lose 10 NM in one unlucky turn, but the difference between 184 and 174 is nowhere near the scale of 190 to 150 in one turn.

[Kensai]

Hey, at least it will be a LOWER drop than the number you gained instantly. If you can raise it so high in a single turn just by capturing a capital, you should at least theoretically be able to lose some of it by next fortnight.
Btw, losing a dice 48 turns in a row is quite the feat, considering that the more you go toward 100 the higher the probability... up to almost 1 in 2)

[Vezina]

I agree it's a feat, but being a player of Axis and Allies that rolled 15 infantry dice without rolling a single 1...I know it can happen. And the gain was expected so it can be accounted for. A random loss (or gain for that matter) can't.

[Kensai]

Maybe. But it's still a balancing issue. The huge NM can allow for extreme crazy, almost unrealistic gestures of combat. It is important to experience a war simulation, not an arcade game. Pocus' solution is ideal because it allows for short-lived sparks without distorting reality too much.

[bjfagan]

If the chance of dropping 10 NM in one turn is very low, then there is no problem in coding a cap on the number of dice rolls possible. Vezina is correct, the possibility of a big drop is present, so a limit to the amount it can drop in one turn is justified. Cap it at 10 NM points in one turn.

[Moriety]

If the chance of dropping 10 NM in one turn is very low, then there is no problem in coding a cap on the number of dice rolls possible. Vezina is correct, the possibility of a big drop is present, so a limit to the amount it can drop in one turn is justified. Cap it at 10 NM points in one turn.

I've yet to have a war.
Surely a natural linear rise in morale if you are winning is fine, countered by a natural decline when losing, and also based upon each turn. I think morale has always been spot-on in Ageod games so it's been odd to read this post about 150 morale two years after a war finished. Surely an AI Yes/No gate check for "war ended or not"? should then trigger a peace time bit of AI maths to slowly return the morale to normal?
God, I wish I'd listened to my maths teacher, I'd love to dive in....

[Jim-NC]

I say cap it at 10/turn during war time, but not during peace. Otherwise, you could be winning the war, and then all of a sudden lose all your NM gain from a single bad turn.

And if you need an example of a bad turn, ask the British of our MP game what happened in Smyrna! The "chance of surrender is so low, that we don't have to worry about it" surrender took everyone by surprise, and was not repeatable after several attempts by different people to explain the results.

[glennbob]

I say cap it at 10/turn during war time, but not during peace. Otherwise, you could be winning the war, and then all of a sudden lose all your NM gain from a single bad turn.

And if you need an example of a bad turn, ask the British of our MP game what happened in Smyrna! The "chance of surrender is so low, that we don't have to worry about it" surrender took everyone by surprise, and was not repeatable after several attempts by different people to explain the results.

Oh God, don't. The agony. Unbearable. All my Scots. Gone.

[Moriety]

A tailored decline is needed surely?

Scots? I asked one of my uncles if he could give me copies of the family photo's of our ancestors, many of whom are Scots and Irish. It would be good if we could show photo's of our ancestors who fought in the wars- especially that most stupid war of all- WWI, and where this game ends. I saw a picture of very proud Great, or Great Great Uncle in his kilt, but no idea of the regiment, but it was WWI. I must ask again.
One of my youngest cousins just completed his training in the Royal Navy. I told him he'd just made it, as the next round of cuts will have him in a rowing boat.

[Kensai]

I say cap it at 10/turn during war time, but not during peace. Otherwise, you could be winning the war, and then all of a sudden lose all your NM gain from a single bad turn.

And if you need an example of a bad turn, ask the British of our MP game what happened in Smyrna! The "chance of surrender is so low, that we don't have to worry about it" surrender took everyone by surprise, and was not repeatable after several attempts by different people to explain the results.

Excuse me, but bad luck works both ways. Your bad luck might be the adversary's good luck and vice versa. I don't recall saying "we don't have to worry about it" with the old algorithm and certainly don't think the Ottoman player saw it as "bad luck" at all. It WAS a bad luck for Britain but in the spot of Britain could have been anyone. I have studied dozens of different battle results (from pretty good to disastrous) by just replaying the same turn in my recent tests of NM. Luck is everywhere, every check in the game is a dice roll. But the game should be mostly won through strategy and tactics not some unbalanced ultra-high inhuman ability to fight with surreal cohesion values. This should not end up an arcade.

Once you reach 120 (still a high value!) that decrease roll will fail four out of five anyway. A pretty decent percentage.

[aries613]

Maybe. But it's still a balancing issue.

[aries613]

Hello,everyone!

[Pocus]

Ok, a cap of 10 will make content everyone.

[Kensai]

Pocus, one other curiosity. How does war resilience work if a nation fights two rivals at the same time? If one of these two rivals has more than 1.5x the NM of the first nation, does war resilience get blocked? I suppose yes, which makes it even more important to decrease high NMs as soon as possible as it influences the war ability of a nation even when it doesn't make sense. In our game Japan is winning the war against Prussia at the moment (positive warscore) but since this latter travels on high NM from its previous wars it seems it stops Japan's war resilience even if there is no reason why that should happen.