PDA

View Full Version : Why do my berserk commanders disappear after winning fights


Bananadine
March 30th, 2010, 02:01 AM
A few times now in the same game, I've had a very powerful commander go berserk and win a battle while the rest of his side routed. Two of those times, the commander disappeared after the battle, and thereafter it was as if he had been killed.

This happened, just now, for the second time. The enemy all routed or died, leaving only my commander on the field--but now on the map, some of those routed enemies are apparently still in the province they invaded and then routed from. They are besieging my fortress, and my great commander is no longer on the map. It seems as if my .trn file is broken--as if the battle replay it shows me isn't what "really" happened. What kind of awful bug is this? How can it be avoided?

Lingchih
March 30th, 2010, 02:08 AM
A few times now in the same game, I've had a very powerful commander go berserk and win a battle while the rest of his side routed. Two of those times, the commander disappeared after the battle, and thereafter it was as if he had been killed.

This happened, just now, for the second time. The enemy all routed or died, leaving only my commander on the field--but now on the map, some of those routed enemies are apparently still in the province they invaded and then routed from. They are besieging my fortress, and my great commander is no longer on the map. It seems as if my .trn file is broken--as if the battle replay it shows me isn't what "really" happened. What kind of awful bug is this? How can it be avoided?


He got killed. Deal with it.

Bananadine
March 30th, 2010, 02:29 AM
He got killed. Deal with it.

How did he get killed? I think you don't know. I'm dealing with it by asking a question about it. Does anybody know the answer?

Squirrelloid
March 30th, 2010, 02:59 AM
Sometimes battle replays are wrong, for whatever reason. There is no known workaround, other than asking your opponent if he saw an accurate replay, and getting him to tell you what happened.

Grijalva
March 30th, 2010, 03:13 AM
I noticed this too in a recent game. I had completely routed and defeated the enemy and even won the battle (according to the message) but the enemy remained in actual control of the province on the map.

militarist
March 30th, 2010, 05:05 AM
Today in replay I have 5 tigers and 1 cave dragon alive after battle. Cave dragon with 4 pts of life (he did a great job, keeping 3 grove guards 1 step from my mages for a few turns). After battle I had no him in my army, as well as less tigers.

Sombre
March 30th, 2010, 05:25 AM
Yeah the battle replay bug sucks. I had a battle in a game just recently where I killed a whole load of enemy stuff and both my thug and the support caster survived with ease. But the 'reality' is that both my guys died and very few of theirs did. Hard to see exactly what it was that could have differed so much from what I saw. Maybe a first turn soul slay?

Psycho
March 30th, 2010, 06:42 AM
After 75 rounds of combat all berserked attackers instantly die. If any defenders remained on the field (due to fatigue perhaps), they keep the province. This is not the case here, since you say you were defending, so it's the awful bug striking again unfortunately. But, maybe it was the case for the other battle.

Bananadine
March 30th, 2010, 09:03 AM
After 75 rounds of combat all berserked attackers instantly die. If any defenders remained on the field (due to fatigue perhaps), they keep the province. This is not the case here, since you say you were defending, so it's the awful bug striking again unfortunately. But, maybe it was the case for the other battle.

I don't think so--in both cases, the field was cleared of enemies in the replay I saw. So I guess nobody should have reached the 75-round limit.

It sounds like dudes consider this to be just some random problem. I don't expect to figure it out myself, but still, it's suggestive that for me, this happened twice in the same game, in battles where a single berserker was the last defender standing....

Sombre
March 30th, 2010, 09:20 AM
One thing I would say is the random battle replay bug is WAAAY more common when mods or patch transitions are invovled.

Bananadine
March 30th, 2010, 10:08 AM
In my case, mods are involved. And it happened once before the recent patch transition, and once after it.

fungalreason
March 30th, 2010, 10:39 AM
This may or may not be the cause in this specific circumstance, but I have had cases in the past where having a different version of a mod than the one the server is using causes the replay bug.

Is your commander mindless by chance as well? I think there's a 50 turn limit on them.

chrispedersen
March 30th, 2010, 10:42 AM
Let me ditto Sombre's advice:

The replay bug occurs WAY more often if:

a). Host and client are on two different systems.
b). rollbacks are involved.
c). Patches have been applied mid game
d). Someone does not have the same version of the game 3.23b for example.

I do wish there were a way to validate clients before a turn were submitted (and no, I'm not talking about the -v option).

Sombre
March 30th, 2010, 10:54 AM
Even without different versions of mods, simply having mods on makes the bug more likely.

But yes, different versions means you'll almost certainly get the bug. I wouldn't even call it a bug there, you just have the wrong version.

Bananadine
March 30th, 2010, 12:01 PM
The attacking player just got back to me--he says he saw what I saw. He also found this other thread about the problem: http://forum.shrapnelgames.com/showthread.php?t=34902

In that thread, dudes say what they've said in this thread, plus the additional speculation that a discrepancy in operating system between host and client can also trigger the problem.

It occurs to me that this problem could be happening very often--and we just wouldn't usually notice it, because only a few minor soldiers would have had their fates misrepresented because of it. But of course it's very obvious when a major commander dies in "reality" but not in the replay.

Anyhow, I feel better about the problem now. I'd been worried that there was a mysterious bug that killed powerful berserkers at random. That problem would be very hard to defend against! But now it appears that my powerful berserkers died correctly, and that only the replays I watched were incorrect. That problem is much easier to defend against! In fact it requires no special defense at all. Just equip your commanders so they won't die, and tolerate the fact that you sometimes won't know what they actually did do. :)

(Of course I'll be careful about version numbers etc., but I was already careful about that, and this still happened.)

I see what you mean Sombre, about not calling this a bug when it arises due to mod version mismatches--but if that's a way to cause this error, then the game allows you to use mismatched mods when it shouldn't. I consider THAT a bug. Of course this community has a pretty high threshold for calling things "bugs" rather than "WAD" issues or even "features". At any rate, maybe we can agree that it'd be better if the game didn't let you use mismatched versions of itself.

thejeff
March 30th, 2010, 12:30 PM
That's a very old thread. At that time the problem was thought to be strictly a problem of version and/or OS differences. I believe the OS differences part of the problem was fixed awhile ago. And soon after reports of the replay bug happening without OS/version differences became more common. It's not clear whether this was a new problem or previous reports were just dismissed as confusion about OS/version differences.

I know I've seen it in SP, with the client and host being the same and thus using the same version/OS & the same mods. I'm not sure I've seen it without mods, but I usually use mods these days so ...

Mismatched mods would be hard to prevent, since they are just text files. I suppose some sort of checksum could be calculated...
OTOH, mod differences aren't necessarily a problem. The display only mods (Banners, etc) can be completely different without changing gameplay at all.

Sombre
March 30th, 2010, 01:02 PM
the game allows you to use mismatched mods when it shouldn't. I consider THAT a bug. Of course this community has a pretty high threshold for calling things "bugs" rather than "WAD" issues or even "features". At any rate, maybe we can agree that it'd be better if the game didn't let you use mismatched versions of itself.

No, I don't agree. There are some cool tricks which rely on using dms which differ at host and client level. One random example is allowing people to have streamers and standards on or off. I don't see the benefit to the game refusing to work with differing dm contents between players, but along those lines a warning wouldn't be too horrible I guess. I think that's the kind of 'bug' which would be a pain to fix and no-one really cares about too.

Easy enough to just not have mismatched dm versions.

If you're talking about versions of dominions itself, then I do agree.

Bananadine
March 30th, 2010, 01:24 PM
WELL let's put it this way: There are some version discrepancies that cause errors, and some that don't, and I guess the game allows all kinds without protest. Bug!

NTJedi
March 30th, 2010, 02:11 PM
WELL let's put it this way: There are some version discrepancies that cause errors, and some that don't, and I guess the game allows all kinds without protest. Bug!

Wait until you have your best supercombatant die on the battlefield because the battle round reached turn_75. All attackers are auto-killed on battle round_75.

As soon as it reaches 5pm(round_50) all attackers must retreat and as soon as it reaches 9pm(round_75) all attackers are instantly killed. Honestly it should be changed from instantly killed to instantly force an automatic retreat.

Also I'm certain if given the option most players would increase the current auto_retreat and auto_kill settings.

Rookierookie
March 30th, 2010, 02:28 PM
WELL let's put it this way: There are some version discrepancies that cause errors, and some that don't, and I guess the game allows all kinds without protest. Bug!

Wait until you have your best supercombatant die on the battlefield because the battle round reached turn_75. All attackers are auto-killed on battle round_75.

As soon as it reaches 5pm(round_50) all attackers must retreat and as soon as it reaches 9pm(round_75) all attackers are instantly killed. Honestly it should be changed from instantly killed to instantly force an automatic retreat.

Also I'm certain if given the option most players would increase the current auto_retreat and auto_kill settings.
Had that happen. Fun. Enemy routed after 10 turns or so but had many exhausted units and I just couldn't kill them quick enough to win and lost my Pretender.

I've had many a bug-ridden battles against a Shedu in an SP game. In every replay I win the battle but lose my prophet commander, while on the map I lose the battle but the prophet is unharmed.

chrispedersen
March 30th, 2010, 02:30 PM
An auto rout on turn 75 doesn't work, because

a). the units didn't auto route on turn 50 like they were supposed too.

b). Mindless / berserk units shouldn't route anyway.

c). I sure as heck wouldn't want to increase the strength of SC's. I know a lot of people like them - but they are already way too powerful. Armies have way too little utility in late game, as do communions.

thejeff
March 30th, 2010, 03:32 PM
An auto-retreat could be different than an auto-route. It could be just the battle ends and any surviving attacking troops are moved to other provinces as if they had retreated.
It would require a significant code change though. Not likely to happen.

Belac
March 30th, 2010, 04:24 PM
I had the inaccurate-replay bug happen to me in a game recently(definitely mod version discrepancies, as my opponent saw what I saw but other players with scouts in the province saw different stuff entirely). It's mildly annoying unless you were really hoping to see how a complicated strategy worked out, or what went wrong with it.

Squirrelloid
March 30th, 2010, 04:53 PM
An auto-retreat could be different than an auto-route. It could be just the battle ends and any surviving attacking troops are moved to other provinces as if they had retreated.
It would require a significant code change though. Not likely to happen.

If a unit is, say, encased in ice, it shouldn't get to retreat. Allowing them to escape the field is silly. (This is one of the few good ways to actually kill a well-kitted QoW underwater).

Auto-killed on turn 75 is mostly a way of saying 'you haven't managed to retreat yet, so we're going to assume the defenders get to you eventually'.

The one change that should be made is that the turn 75 auto-kill should instead apply to the army which has already routed rather than the attacker. (In situations which it is meant for, the auto-kill will still effect the attacker because they auto-routed at turn 50. But they can't auto-route if the defender already routed (bug), since the game won't let both sides rout. Ie, if one side routes, the other side can't. So if the defender routs before turn 50, they should be auto-killed at turn 75 since its the attacker who will get to kill them at his leisure).

chrispedersen
March 30th, 2010, 09:06 PM
Mostly I agree with you Squirrel, but it still has some defects.

Most commonly I see this due to SCs or berserking units.

If both sides have units that berserk, I don't see the logic about which sides routes as mattering.

Now, it might be interesting if each side (attacker first) got hit with a 1000 pt AN attack. One for each side, starting say at turn 65.

NTJedi
March 31st, 2010, 03:21 AM
Auto-killed on turn 75 is mostly a way of saying 'you haven't managed to retreat yet, so we're going to assume the defenders get to you eventually'.


'Defenders get to you eventually' assumes every race of people, no matter how primative, is able to eventually kill any type of creature/being from mythology despite any magic buffs and/or items which is not reasonable. If the "best" military units defending a province are unable to kill a supercombatant after 75 rounds it's clear the regular population or other military from the same province wouldn't make any difference. The supercombatant should then be auto-retreated thus instantly forced into a neighboring province to return another day. Auto-Killed is not only historically inaccurate, logically inaccurate, but it's a crude solution when better and more realistic solutions can easily exist within the game.

Squirrelloid
March 31st, 2010, 03:46 AM
Auto-killed on turn 75 is mostly a way of saying 'you haven't managed to retreat yet, so we're going to assume the defenders get to you eventually'.


'Defenders get to you eventually' assumes every race of people, no matter how primative, is able to eventually kill any type of creature/being from mythology despite any magic buffs and/or items which is not reasonable. If the "best" military units defending a province are unable to kill a supercombatant after 75 rounds it's clear the regular population or other military from the same province wouldn't make any difference. The supercombatant should then be auto-retreated thus instantly forced into a neighboring province to return another day. Auto-Killed is not only historically inaccurate, logically inaccurate, but it's a crude solution when better and more realistic solutions can easily exist within the game.

That's not true at all.

Historical example:
There's documented cases of knights, travelling alone or relatively alone, being dehorsed by bandits whose weapons were otherwise unable to injure him through his armor. Of course, keeping the knight pinned down was pretty trivial since he was without recourse to other aid, so a bandit could run off to the nearest farmhouse, borrow or steal an axe, and use that to punch through the knight's armor. Ie, once you have an opponent helpless, finding a way to penetrate his defenses *somewhere* becomes a lot easier, and you can resort to unusual measures.

Would you care to specify a unit that, once incapacitated, would be impossible to kill for some nation?

Incapacitated foes can be stripped of gear, subject to attacks that simply wouldn't work on a battlefield, or otherwise put to death. No unit is invincible, and having them at your mercy dramatically increases your options for dealing with them. Even the most primitive race in dom3 can make a bonfire, find a way to deprive you of oxygen, or drag you to a place where they can drop exceedingly large rocks on you. And most creatures are far less threatening when stripped of gear, something trivial to accomplish if a creature is unable to act.

Basically, its no longer a question of military capability. Its a question of creativity. And you don't have to be all that creative.

Edit: Also, magic buffs eventually expire.

thejeff
March 31st, 2010, 08:28 AM
Obviously, it can work either way.

An invincible attacker who just hasn't managed to kill off all the defenders before the time limit. Or a trapped and helpless attacker who's tough enough not to actually die before the time limit. Or 2 berserk SCs hacking at each other but unable to do any lasting damage.
Extending the time limit would help resolve some situations, but not all.
Any solution fails in some circumstances: Auto-kill kills those who clearly were in no danger. Auto-retreat can save those who had no way to escape.
Auto-retreat seems a marginally better solution to me, but it also requires code changes and isn't likely to happen. The current auto-kill solution isn't much worse.

As an aside, I have certainly seen both sides rout. I'd always thought it went attacker routs, defender routes, then auto-kill if that doesn't work.

Bananadine
March 31st, 2010, 10:53 AM
Haha this is the problem with forums that use linear thread structures! The topic changes importantly in mid-conversation, but how is anybody supposed to know that without reading every post.

Would you care to specify a unit that, once incapacitated, would be impossible to kill for some nation?

They wouldn't all become incapacitated. A mundane knight traveling in a foreign land would be in big trouble; where would he get his food? And even a giant Jotun mage with the Monolith Armor etc. would eventually have to sleep. But what about a well-equipped wraith lord or a Vastness or a high-level angel of some kind? Those things get tired when they cast spells, if they cast spells, but they wouldn't lose their ability to walk or fly right out of danger just because they had to rout from one battle. And what if only a few hundred peasants happened to be in the province the unit routed to, and the unit was amphibious, invincible to fire and ice, non-sleeping, non-eating, regenerating, and always ethereal, fire-shielded, and blood vengeance...ful? A naturally ethereal undead commander could achieve that much and more, with the right equipment. What would peasants do to a monster like that, even if it did knock itself out by casting spells at them? They'd leave it alone, or they'd all (or almost all) die, that's what!

I guess if you wanted to be "realistic", you'd have routing units end up starting new fights in whatever places they routed to. Your flying, regenerating titan might jump all over the map, fighting battle after battle, until it either won one or legitimately died. So a commander that teleported into enemy territory with no friendly land around could leave a trail of destruction, and would be even harder to kill than it would be now. Not so good for the game! But relatively realistic.

I think there's surely a way to design a more subtle solution to the problems of routing and of overlong battles than the really harsh rules that the game currently uses. Like, if somebody were writing Dominions 4, I think they could do better than what we have. Maybe a unit routing into enemy territory would start one new battle after another, each time it routed, until it won or died--but all its stats would be halved anew with each battle, to roughly represent the "incapacitation" incurred by its hostile, foreign environment. That still wouldn't make a lot of sense for, like, ghosts and non-sleeping demigods and such, but it'd be less absurd than instant death.

That might still be overpowered though. The defender would still have to defend his land against both primary attackers, and enemies routing in from other provinces. So maybe the defendant could be compensated for that increased burden, with an increased chance of reward due to attack by routing enemies. Maybe, for instance, a commander routing into enemy land could have, say, a 50% chance of starting a new battle there, a 25% chance of instantly dying, and a 25% (with a boost for low morale?) chance of defecting to the enemy side, maybe with any commander status removed. And if it did fight after routing, but then died, maybe minor local shamans or whatever would convert its corpse and equipment into gems that the defender would get to keep.

There are a million ways for it to work! I think the system we have now is good enough, silly as it can seem at times, but there are better ways. If not what I suggested, then something else. But only in the imaginary Dominions 4. :)

NTJedi
April 2nd, 2010, 03:17 PM
Would you care to specify a unit that, once incapacitated, would be impossible to kill for some nation?

First the attacking unit is NOT incapacitated... not sure where you pulled this idea. Turn_75 does not equal incapacitated. The attacking unit is still actively healthy and moving on the battlefield during turn_75 due to berserk or some other natural trait. And when an angel with a tower shield, regeneration, and it's natural awe cannot be harmed by the "best" military the natural population of peasants have no reason or intention of risking their lives. Heck if a town discovered a hundred navy seals were killed by some beast which resisted point blank C4 explosions then OBVIOUSLY the citizens of the town will let the beast control that piece of the beach.
Sure something can eventually be done to initiate another attack, but that would require a separate (game_turn)strike from the military not the general population! You are mistaking the effectiveness of general population as compared with an actual military and mage strike force.

[/QUOTE]

Squirrelloid
April 2nd, 2010, 04:03 PM
Would you care to specify a unit that, once incapacitated, would be impossible to kill for some nation?

First the attacking unit is NOT incapacitated... not sure where you pulled this idea. Turn_75 does not equal incapacitated. The attacking unit is still actively healthy and moving on the battlefield during turn_75 due to berserk or some other natural trait. And when an angel with a tower shield, regeneration, and it's natural awe cannot be harmed by the "best" military the natural population of peasants have no reason or intention of risking their lives. Heck if a town discovered a hundred navy seals were killed by some beast which resisted point blank C4 explosions then OBVIOUSLY the citizens of the town will let the beast control that piece of the beach.
Sure something can eventually be done to initiate another attack, but that would require a separate (game_turn)strike from the military not the general population! You are mistaking the effectiveness of general population as compared with an actual military and mage strike force.

[/QUOTE]

99% of the time when i see a combat go to turn 75, its because the last remaining enemy units *couldn't* retreat. (Encased in ice, passed out and unable to get below 100 fatigue due to fatigue causing effects, etc...). Ie, incapacitated.

In the specific case of berzerk, eventually that berzerker rage ends. Generally, berzerkers are exceptionally tired when they come down from their rage. Its not unreasonable to say 'turn 75' is when this happens, and its not a far cry from being incapacitated.

Any other unit should have managed to retreat by turn 75.

NTJedi
April 2nd, 2010, 05:39 PM
99% of the time when i see a combat go to turn 75, its because the last remaining enemy units *couldn't* retreat. (Encased in ice, passed out and unable to get below 100 fatigue due to fatigue causing effects, etc...). Ie, incapacitated.

In the specific case of berzerk, eventually that berzerker rage ends. Generally, berzerkers are exceptionally tired when they come down from their rage. Its not unreasonable to say 'turn 75' is when this happens, and its not a far cry from being incapacitated.

Any other unit should have managed to retreat by turn 75.

From my experience it's been when the defender could not retreat while the supercombatant was killing the exhausted defenders/crippled defenders. I cannot even recall a battle where exhausted attackers or fleeing attackers were autokilled. As a result the perfectly healthy supercombatant is auto_killed instead of a more realistic solution of auto_retreat so the battle can continue later. Thus we witness an almighty legendary creature killed due to lack of game time!
Ideally the game should provide an instant retreat instead of an instead auto_kill as well as a game setting to adjust the auto_kill game turn. I'm sure many would increase the auto_kill game turn to at least turn_100.

thejeff
April 2nd, 2010, 05:57 PM
A time limit on berserk would also work nicely.

Squirrelloid
April 2nd, 2010, 08:09 PM
99% of the time when i see a combat go to turn 75, its because the last remaining enemy units *couldn't* retreat. (Encased in ice, passed out and unable to get below 100 fatigue due to fatigue causing effects, etc...). Ie, incapacitated.

In the specific case of berzerk, eventually that berzerker rage ends. Generally, berzerkers are exceptionally tired when they come down from their rage. Its not unreasonable to say 'turn 75' is when this happens, and its not a far cry from being incapacitated.

Any other unit should have managed to retreat by turn 75.

From my experience it's been when the defender could not retreat while the supercombatant was killing the exhausted defenders/crippled defenders. I cannot even recall a battle where exhausted attackers or fleeing attackers were autokilled. As a result the perfectly healthy supercombatant is auto_killed instead of a more realistic solution of auto_retreat so the battle can continue later. Thus we witness an almighty legendary creature killed due to lack of game time!
Ideally the game should provide an instant retreat instead of an instead auto_kill as well as a game setting to adjust the auto_kill game turn. I'm sure many would increase the auto_kill game turn to at least turn_100.

I've already maintained the side that routed should be the one to be auto-killed, so if its defenders who have failed to rout from the field then I agree, the attacker shouldn't be autokilled.

chrispedersen
April 2nd, 2010, 08:22 PM
SC's are too powerful. Instead of autokilling on 75, it should be autokill on 50.

We can all repeat our points = ).

But in retrospect, having it be something you could specify at game start would be cool, just given that almost everyone agrees that Tarts are the end-all be-all, I can't see why you would want to do anything to *increase* their power.

Rookierookie
April 3rd, 2010, 01:37 AM
The issue happens because the defenders have already routed. They lost the freakin' battle, and somehow manage to win because they couldn't run away.

The simplest solution is really to make the side that routs first lose the battle, regardless of whatever happens afterwards, and auto-kill the losing side.

Sombre
April 3rd, 2010, 07:19 AM
So remove rout retreats entirely?

NTJedi
April 6th, 2010, 02:16 PM
SC's are too powerful. Instead of autokilling on 75, it should be autokill on 50.

It's wrong to leave an unrealistic and auto_kill game function within the game when better measures can be taken to deal with SC's being too powerful.
This can be done if the developers were to add new spells which cause insanity on the target. Each time the target is successfully struck by the spell it adds 3 insanity. Another idea is adding a few new summon creatures which horrormark those which kill it. Another idea is a ritual spell which randomly teleports one commander, mage or SC to a random province. I'm working on a mod which can provide the first two suggestions, but I wish I could mod the autokill of turn75 to something of turn125.

Bananadine
April 6th, 2010, 02:44 PM
I cannot even recall a battle where exhausted attackers or fleeing attackers were autokilled.

One time a Sacred Statue teleported onto my army and I had some kappas and they dried up and couldn't leave and the Sacred Statue ran out of time before managing to mind-burn them all so it gave up and died.

I wish I could mod the autokill of turn75 to something of turn125.

That would be cool. Hm it's just a couple of numbers right? Maybe a Dominions developer could confirm for us that it's actually stored in the game program as just a couple of numbers, and then somebody could hunt them down in a hex editor and manually patch them. Probably not though

Sombre
April 6th, 2010, 03:41 PM
This can be done if the developers were to add new spells which cause insanity on the target. Each time the target is successfully struck by the spell it adds 3 insanity
......
I'm working on a mod which can provide the first two suggestions

How are you doing the insanity spell?

chrispedersen
April 7th, 2010, 09:45 AM
SC's are too powerful. Instead of autokilling on 75, it should be autokill on 50.

It's wrong to leave an unrealistic and auto_kill game function within the game when better measures can be taken to deal with SC's being too powerful.
This can be done if the developers were to add new spells which cause insanity on the target. Each time the target is successfully struck by the spell it adds 3 insanity. Another idea is adding a few new summon creatures which horrormark those which kill it. Another idea is a ritual spell which randomly teleports one commander, mage or SC to a random province. I'm working on a mod which can provide the first two suggestions, but I wish I could mod the autokill of turn75 to something of turn125.

I couldn't disagree more. I don't know what *unrealistic* has to do with an imaginary game with blood sacrifice, magic, and unicorns.

While I think your ideas of new spells are great, I would still be opposed to removing the autokill. Have you ever tried killing a properly kitted chayot, grigory, or regenerating sphinx? It can be bloody hard. One of the tactics I *like* to do is to get them to berserk and auto kill them.

I can equally argue that the problem is not the time limits - its wrong to leave unrealistic and autokill SP's!

Bananadine
April 7th, 2010, 10:30 AM
I couldn't disagree more. I don't know what *unrealistic* has to do with an imaginary game with blood sacrifice, magic, and unicorns.

It's also a game about land, peasants, government, taxes, religion, and spears. The game is unrealistic and realistic, to great extents both.

People naturally care about consistency, so when a superhero that can survive being trampled by ten elephants during a fire storm gets killed by an invisible clock, dudes are annoyed. People naturally care about drama, so, well, same thing. I think consistency and drama are more relevant here than realism is.

Rookierookie
April 7th, 2010, 10:58 AM
SC's are too powerful. Instead of autokilling on 75, it should be autokill on 50.

It's wrong to leave an unrealistic and auto_kill game function within the game when better measures can be taken to deal with SC's being too powerful.
This can be done if the developers were to add new spells which cause insanity on the target. Each time the target is successfully struck by the spell it adds 3 insanity. Another idea is adding a few new summon creatures which horrormark those which kill it. Another idea is a ritual spell which randomly teleports one commander, mage or SC to a random province. I'm working on a mod which can provide the first two suggestions, but I wish I could mod the autokill of turn75 to something of turn125.

I couldn't disagree more. I don't know what *unrealistic* has to do with an imaginary game with blood sacrifice, magic, and unicorns.

While I think your ideas of new spells are great, I would still be opposed to removing the autokill. Have you ever tried killing a properly kitted chayot, grigory, or regenerating sphinx? It can be bloody hard. One of the tactics I *like* to do is to get them to berserk and auto kill them.

I can equally argue that the problem is not the time limits - its wrong to leave unrealistic and autokill SP's!

If logical consistency doesn't matter in a game, why don't we make it so that the army that routs first always wins the battle, more powerful spells cost less gems, and building temples decrease your dominion?

Illuminated One
April 7th, 2010, 11:43 AM
I don't know about consistent, in the hard sense everything can be made consistent, but having some sort of realism that you can relate too is important.
I.e. ok, we have unicorns but everything works the way we're used to makes (maybe) a good unicorn game. But we have unicorns so we are unrealistic anyway so humans reproduce through laying eggs and then throwing the eggs into the wastebin and making children the normal way (what, why eggs then? Rule 7-3i of course) and the law of ant-ropy states that the world will be gradually filled by ropes... I'm still wondering if these mushrooms were edible...

It's unrealistic in the first place to have a rout occur exactly on round 50 (that's about 5 minutes of combat) or any other fixed round for that. Quite simply put why would a soldier who is totally uninjured or fatigued and all his comrades are so too suddenly decide to run away and leave his wife and daughter to demon summoners?
There would be better ways to prevent endless battles (icrease fatigue by 5 on round 50, by 10 by round 75 by 15 by round 100 ..., let an SC rout if he's too worn out and not berserk...). Of course the AI taking over after round 5 and SCs being so strong would be even more of a problem but then you change that...

Graeme Dice
April 7th, 2010, 11:44 AM
The word your looking for is not realism, but verisimilitude. The rules of the game should obey their own internal self-consistency. A time-limit that arbitrarily kills units that would otherwise be invulnerable isn't particularly consistent. It is definitely necessary to have some point where battles end though, as otherwise it's possible that some combinations could run indefinitely. Two feebleminded sacred statues staring each other down for example.

Gregstrom
April 7th, 2010, 01:50 PM
The game is in fact extremely realistic - it's just that no-one ever noticed that all units in the game are members of the New Amalgamated Union of Gribblies, Humanoids, Thugs and Yetis (NAUGHTY). Among the NAUGHTY Articles are extremely strict rules on the duration of melee activities any NAUGHTY Member takes part in over any given month. The NAUGHTY rules specify exactly how long the attacking and defending sides can stay involved in combat, and punishment for breaching said rules is carried out on the spot. This is why attacking armies leave so promptly - they're actually going off for the union-mandated tea break.

Pretenders beware - argue with the unions and the NAUGHTY people will come and get you!

chrispedersen
April 7th, 2010, 02:05 PM
Gregstroms point is exactly.. on point.

You could arbitrarily say that combat makes the world unstable and that each round increases the likelihood of death.

Now, if it were me, I'd have horrors show up increasing in numbers and severity. But I have no problem abstracting that part of the equation.

I mean, if you think about it, no one has complained that undead simply go poof. Why should you complain therefor if an SC goes poof?

similarly you don't complain that we abstract the turns - the entire months turn is abstracted to one click. We don't see the armies progressing.

The preeminent question is a balance question, not a versimilitude question.

If you are really saying that you can't stand the versimilitude (vs wanting to make sc's more powerful) then you would be fine with any solution that kept the 75 turn limit autokill, so long as it had a reasonable backstory.

And I don't think that is really what anyone is saying. I think they are just using the versimilitude gripe as a way of attacking the turn limit.

Squirrelloid
April 7th, 2010, 02:07 PM
Its perfectly plausible to rationalize an autokill at some fixed time - ie, units eventually are too fatigued to fight, undead units need to replenish their dark energy or whatever, etc... So after some number of rounds they simply collapse or shut down and the defenders can do what they will (since they control the land).

The game has rules. Autokill is one of them. Its a perfectly balanced rule, especially since the rules are designed to give the attacker a fair shake at leaving before it happens by auto-routing. The *only* issue is that the game won't let the attackers route if the defenders have already routed. To avoid this, there are two alternatives: (1) let both sides route, (2) instead of auto-killing the attacker, auto-kill the side that routed (keeping auto-route for the attacker at turn 50 - presumption for the defender).

As a game rule issue, that's the end of the discussion.

Once we start talking about removing autokill because its 'unrealistic', we start playing with game balance (autokill is often the only way to kill a pimped-out SC), which means we have to try and make lots of other changes and that leads nowhere good (and probably means we have to do things like remove enc. 0 from the game and/or make SCs a lot weaker - possibly eliminating the role entirely). We have to start rebalancing some spells and abilities (berzerk becomes stupidly powerful without autokill). Removing autokill is not a fix, it forces a total game redesign.

thejeff
April 7th, 2010, 03:04 PM
Is it actually true that attackers can't route if defenders have already routed?

I know I've seen the reverse recently. Attackers route, then defenders.

Squirrelloid
April 7th, 2010, 03:24 PM
Is it actually true that attackers can't route if defenders have already routed?

I know I've seen the reverse recently. Attackers route, then defenders.

Particular squads can rout from both sides, but once one side has globally routed (you see the 'x's forces have routed' message) the other side cannot globally rout, and may not be able to trigger unit routs (not sure about the second).

Auto-route is a global rout, and thus fails if the defenders have globally routed.

Psycho
April 7th, 2010, 04:23 PM
Is it actually true that attackers can't route if defenders have already routed?

I know I've seen the reverse recently. Attackers route, then defenders.

It is true and it enables a situation when an immortal (think vampire lord thug) behind the pd has the possibility to destroy the whole attacking army to the very last mage. When pd routes, the attacking army will not route no matter how many casualties inflicted to them.

thejeff
April 7th, 2010, 04:40 PM
I definitely have an example of both sides routing.
I can post the turn if you'd like. The attacker routes first with the message, then he has a berserk Ice Devil and my forces can't kill him and the routing (fatigued or bound) troops fast enough so they route. Now with the "Atlantis has routed" message alternating with the "Pangaea has routed" one.

I didn't count turns, so it's possible one or the other of us routed for different reasons. I can run it with debug on, but that'll take awhile, so it'll have to wait.

I wonder if it only works in that order? If the defenders route first (for % of hp reasons, or whatever) then since the attackers are supposed to autoroute first, the game doesn't bother to try to route them. Whereas if the attackers have routed and the battle keeps going long enough it's happy to try the defenders next?

Psycho
April 7th, 2010, 05:47 PM
That's possible. I can't remember any such situation when the attacker routed first. This all sounds like one of those WAD bugs.

chrispedersen
April 7th, 2010, 07:26 PM
I believe I've seen it the other way as well.

Illuminated One
April 7th, 2010, 08:21 PM
If you are really saying that you can't stand the versimilitude (vs wanting to make sc's more powerful) then you would be fine with any solution that kept the 75 turn limit autokill, so long as it had a reasonable backstory.

The point is reasonable backstory. I don't doubt that you can make something up but ... your backstory would have to be more believable than this:
They are all just living inside a computer simulation with arbitrary rules designed for an unknown purpose, cynics say it's just for the entertainment of an extradimensional species, and the 50-turn rout 75-turn autokill rule was included to reduce hosting and coding times.

That two angels duking it out for 50 rounds makes the world "unstable" in some way could be but two drunken vikings in a barfight?

And I don't think that is really what anyone is saying. I think they are just using the versimilitude gripe as a way of attacking the turn limit.

No...? Why should anyone complain about the turn-limit if it wouldn't be a problem to them? Oh, yeah, I can loose to turn limit. But I can also win, no? Only that winning by using "gamey strategies" is somewhat stale even more so if you are actually forced to use them.
It's only because the game is not balanced that you can even say you need the autokill because you can't kill SCs in another way. Anyway do you really think the game is balanced, I think you said something else about Hinnom?
The game is only balanced in the sense that there is no game winner unit and there are a lot of options. People find their own way of playing. Saying that implementing a more beautiful solution than turn limit and providing other ways to deal with SCs will kill balance is just like saying that planting new trees will make the forest chaotic. It was chaotic all along the only difference is that the guides don't apply anymore...

It is true and it enables a situation when an immortal (think vampire lord thug) behind the pd has the possibility to destroy the whole attacking army to the very last mage. When pd routes, the attacking army will not route no matter how many casualties inflicted to them.

It's certainly possible unless they changed it.
I've had a thug routing after he had routed the enemy (he routed after chasing the last two crippled militiaman to the other side of the map and was even faster from the field in the end).

Rookierookie
April 7th, 2010, 10:14 PM
Gregstroms point is exactly.. on point.

You could arbitrarily say that combat makes the world unstable and that each round increases the likelihood of death.

Now, if it were me, I'd have horrors show up increasing in numbers and severity. But I have no problem abstracting that part of the equation.

I mean, if you think about it, no one has complained that undead simply go poof. Why should you complain therefor if an SC goes poof?

similarly you don't complain that we abstract the turns - the entire months turn is abstracted to one click. We don't see the armies progressing.

The preeminent question is a balance question, not a versimilitude question.

If you are really saying that you can't stand the versimilitude (vs wanting to make sc's more powerful) then you would be fine with any solution that kept the 75 turn limit autokill, so long as it had a reasonable backstory.

And I don't think that is really what anyone is saying. I think they are just using the versimilitude gripe as a way of attacking the turn limit.
Give me a reasonable backstory to this:
SC attacks province.
Defending army routs.
Some defending units can't run because their fatigue is over 100.
After 75 turns, berserker couldn't kill all defenders and is killed.
SC fatigue is 0 at all times and HP is always full.

chrispedersen
April 7th, 2010, 10:43 PM
I can create reasonable backstory all day long - so can a lot of people. Kind of immaterial because it wouldn't be official, and it would be subjective; ie., what you find perfectly reasonably, I may not.

chrispedersen
April 7th, 2010, 10:51 PM
It's only because the game is not balanced that you can even say you need the autokill because you can't kill SCs in another way. Anyway do you really think the game is balanced, I think you said something else about Hinnom?

Quite the opposite: I can say we need autokills because SC's aren't balanced.

I am against the removal of the autokill because it makes the game *less* balanced.

Hinnom as a nation - is an option. Game administrators have the ability to remove it or not.

Players, have the ability to compensate either in the game setup or in diplomacy.

Autokills are a game mechanic that are not subject to modification.

As for finding a better way to balance SC's; or a better way of implimenting autokills - I'm certainly not opposed to it. However given the developers statements the discussion it rather bootless.

Sombre
April 8th, 2010, 06:02 AM
So is 'defending' autokill.

Psycho
April 8th, 2010, 07:50 AM
It is true and it enables a situation when an immortal (think vampire lord thug) behind the pd has the possibility to destroy the whole attacking army to the very last mage. When pd routes, the attacking army will not route no matter how many casualties inflicted to them.

It's certainly possible unless they changed it.
I've had a thug routing after he had routed the enemy (he routed after chasing the last two crippled militiaman to the other side of the map and was even faster from the field in the end).

Yes, he would route at 50 turn limit. That's why I wrote that it's possible (not certain) to destroy the whole attacking army - you still have to do it in 50 turns time.

chrispedersen
April 8th, 2010, 01:33 PM
So is 'defending' autokill.

Ya? So? Its still a free country.

Sombre
April 8th, 2010, 02:10 PM
Er,.. ok. So it's only ok for you to point that out?

Gregstrom
April 8th, 2010, 03:55 PM
That's pretty much the way America did the 'Free World' thing, so... sure.

chrispedersen
April 8th, 2010, 07:49 PM
Er,.. ok. So it's only ok for you to point that out?

Actually Sombre, I'm not really interested in getting in a flame war with you, especially as you're dead on on a lot of topics especially modding and code.

That said, your previous post left the distinct impression you were once again policing the threads and saying what points of views were valid and which were not; something I reject.

IlluminatedOne made several new statements to which I wished to respond. New content hence new discussion. Some of his points were contrary to what I actually believe. For example, I indicated that I'm favor of NT's spells that add insanity; I suggested that games should have an option to set what the autokill duration should be. I just don't believe that they are relevent to the essential points. Are SC's too limited or too strong? There are *very* few effective midgame and endgame tactics, and increasing the strength of SC's will only narrow effective strategies.

Finally, when posts get into the area of discussing people, as opposed to topics they are really falling afoul of the spirit and letter of the usage agreements, and I'd like to suggest that courtesy, is the more appropriate response.

Sombre
April 9th, 2010, 09:49 AM
Actually Sombre, I'm not really interested in getting in a flame war with you, especially as you're dead on on a lot of topics especially modding and code.

What does that have to do with anything? You're the only one 'flaming' by suggesting I'm trolling to provoke a flame war.

That said, your previous post left the distinct impression you were once again policing the threads and saying what points of views were valid and which were not; something I reject.

I don't see where you're getting that at all. I just pointed out a logical problem in your post.

I find the accusation that I try to 'police' what other people say insulting and amusing in equal measure. Since when is disagreeing with someone when they're wrong and pointing out the faulty logic and inaccuracies in their posts 'policing'?

IlluminatedOne made several new statements to which I wished to respond. New content hence new discussion. Some of his points were contrary to what I actually believe. For example, I indicated that I'm favor of NT's spells that add insanity; I suggested that games should have an option to set what the autokill duration should be. I just don't believe that they are relevent to the essential points. Are SC's too limited or too strong? There are *very* few effective midgame and endgame tactics, and increasing the strength of SC's will only narrow effective strategies.

Nice recap, but nothing to do with my interjection.

Arguing against the theory behind proposed changes, that's fine, that's discussion. Pointing out that the changes aren't going to happen, also fine, that's pragmatism. But using the fact that the changes aren't going to happen as part of your argument against them, that's silly. Your argument in favour of autokill is equally 'bootless' if you believe nothing will change. You level that charge against them, I turn it back on you. Recapped.

Finally, when posts get into the area of discussing people, as opposed to topics they are really falling afoul of the spirit and letter of the usage agreements, and I'd like to suggest that courtesy, is the more appropriate response.

You're the only one discussing people as far as I can see. But what would I know, I'm just a flame war starting troll trying to police the forums.

Annette
April 9th, 2010, 10:26 AM
Please, let's get this discussion back on-topic.

Rookierookie
April 9th, 2010, 10:28 AM
Er,.. ok. So it's only ok for you to point that out?

Actually Sombre, I'm not really interested in getting in a flame war with you, especially as you're dead on on a lot of topics especially modding and code.

That said, your previous post left the distinct impression you were once again policing the threads and saying what points of views were valid and which were not; something I reject.

IlluminatedOne made several new statements to which I wished to respond. New content hence new discussion. Some of his points were contrary to what I actually believe. For example, I indicated that I'm favor of NT's spells that add insanity; I suggested that games should have an option to set what the autokill duration should be. I just don't believe that they are relevent to the essential points. Are SC's too limited or too strong? There are *very* few effective midgame and endgame tactics, and increasing the strength of SC's will only narrow effective strategies.

Finally, when posts get into the area of discussing people, as opposed to topics they are really falling afoul of the spirit and letter of the usage agreements, and I'd like to suggest that courtesy, is the more appropriate response.
Are you aware that by accusing the people who are against the current autokill mechanism of being mindless SC proponents, and by saying things like "using the versimilitude gripe as a way of attacking the turn limit", you already stand on the wrong side of trolling?

Annette
April 9th, 2010, 10:41 AM
I'm assuming we were posting at the same time, Rookierookie. But again, please return to the topic so I don't have to close this down.

NTJedi
April 9th, 2010, 05:51 PM
While I think your ideas of new spells are great, I would still be opposed to removing the autokill. Have you ever tried killing a properly kitted chayot, grigory, or regenerating sphinx? It can be bloody hard. One of the tactics I *like* to do is to get them to berserk and auto kill them.

I can equally argue that the problem is not the time limits - its wrong to leave unrealistic and autokill SP's!

AutoKill is a game function which has an extreme result which is not verisimilitude which was explained extremely well by Graeme Dice. Leaving the AutoKill would be like watching the move "Clash of the Titans" yet they kill the Kraken simply by having too many crippled units which could not flee the shore thus the Kraken is killed because it ran out of time. When I'm playing a game if some almighty powerful Kraken exists as seen within the movie... I don't want it killed because it ran out of time when logically it should be allowed to retreat and return another day. I can see the hero returning with the medusa head only for some villager to say, "Oh we killed the kraken because the mighty beast simply ran out of time trying to kill all the illusions some mages casted by the shore. No weapons could pierce it, no spells could harm it... the beast simply died due to taking too long to kill the illusions."

kennydicke
April 9th, 2010, 05:59 PM
My computer was KIA recently but, in the days before, I had noticed an up-tick of battle replays that didn't match-up in SP. None of the battles had lasted long enough for autokills to happen.

I noticed somebody pointed out that mods contribute to this, is it related to the number of mods or their size? I had CBM, CPCS, Bogarus Scout, and 3 nation mods enabled at the time. Would there be less problems if I combined them?

Does berserking units contribute? Every battle with a discrepancy had berserked units but not every battle with berserked units had a discrepancy.

It happened a lot more UW; maybe it was because I started UW? I moved onto land very early and was having the same problems but, it seemed, noticeably less so.

Finally, maybe there was no up-tick and I was just looking for it?

chrispedersen
April 9th, 2010, 07:40 PM
I suspect its the latter. Late game, I seem to get this error in virtually every game.