Should returning be overridden at all?
 

In the thread about bug with golem ignoring returning script it was revealed that many people from time to time run into a problem when AI of a mage considers opposition weak, decides not to use gems at all and so doesn't cast returning. As a result, very often major battles are lost, expensive SC's are slayed and good tactics are ruined.
It seems to me that Returning spells are too specific to be handled as usual spells because they look much more like "Retreat" order than battle spells and "Retreat" orders are never ignored by AI which is absolutely correct.
So the question is: Should scripted "Returning" be overridden by AI if it decides that opposition is "weak" (rather relative term, btw) or not? What do you think?

Re: Should returning be overridden at all?
 

I can see the point. The usual method for handling spells would be moving it higher on the priority category but that wouldnt be a good idea because then AIs might use it when its not wanted.

But are there any other spells that call into a never-overriden category? If we get an exclusion for this spell then wont there be a list of other cases people want excluded? This might be another of those places where we would be opening the door to a load of IF this and IF that statements in the code which has been avoided.

Re: Should returning be overridden at all?
 

I can't think of such spells now. Probably not...
The point is that returning spells (Returning and VoR) are not the battlefield spells in the common meaning and that's the big difference between them and any other spell in the game.

Re: Should returning be overridden at all?
 

Create a flag that says do not override this spell does not change much code.

In that regard, I also think a game by game, user by user do not cast list is a good idea, and does not involve a lot of code. (create, maintain, load, save and use the list probably takes less than 100 line of actual code) That would make save game incompatible with previous versions, but that happened a few times already. :)

Re: Should returning be overridden at all?
 

Well I'm a doofus...I misread the question...so scratch the one (so far) yes vote.

Re: Should returning be overridden at all?
 

So fired. ;p

Re: Should returning be overridden at all?
 

Despite being harmed by this on occasion, I do like that it acts as a counter to Returning.

Re: Should returning be overridden at all?
 

Does returning really need countering? Maybe it's worth starting to think of countering retreat?)
Returning and retreat are different in one major aspect: returning allows to safely escape from assassination attempt and from province surrounded by enemies while retreat doesn't. OTOH, Returning is heavily dependent on the home province, however if you lost your home province, you probably already lost the game as well.:)

Re: Should returning be overridden at all?
 

Retreat has a counter, which is putting troops behind the Retreater...which is countered by Flyers that Retreat, and is countered by Storm....which is countered by Flyers who can fly in Storm....

This game is based on counters, so Returning needs a counter (and no, killing the mage is not a real counter since it works for any effect). Sure, "taking the home province" may seem like a counter, but it really isn't since it implies your opponent has already lost and really doesn't need his stuff countered.

Re: Should returning be overridden at all?
 

How do you put troops behind the retreater when defender gets to move first?

Re: Should returning be overridden at all?
 

Seems like we should have no hope for this to be changed because JK said "yes" :(.

Re: Should returning be overridden at all?
 

In general, you don't. But, I suppose a Round 0 Howl from the SoulStone of the Wolves might do it. I'd have to test it out.

Every counter and then counter counter and the counter counter counter has to end somewhere, right? In this case, the last counter is "be on the Defense."

Re: Should returning be overridden at all?
 

But for returning to work without risk of magic duel, you have to be on defense as well. So I guess you already have your counter here.

Re: Should returning be overridden at all?
 

Not really. Retreat has another counter which is "Assassinate/seduce or take the adjoining provinces." It's hard to do, but it's a decent counter for the guy who has his mages cast a few instants and then retreats them.

Returning's only counters are a small unit that won't set off the AI and not being the attacker and having a chance to kill on Round 1. Take out the first, and it has no real counter since "kill a guy before he can use his X" is the universal counter.

Re: Should returning be overridden at all?
 

Damn, I can't believe that my first post to the forums after years of lurking is to say that:

"I hit YES by mistake, please, do not count my vote" !!

It is very annoying to have the AI override your scripts. I still remember my first time casting "Summon Lamashtas" -- it took me 10 battles to see what the spell does, because my Death ubermage decided not to use gems in the battle!!

Anyway, I use "Returning" a lot with R'Lyeh summoners without ever noticing a single case of being overriden. But if it ever happened it would make me furious: to risk loosing a summoner after 30 turns is a BAD thing.

Re: Should returning be overridden at all?
 

I think it should be capable of being over ridden, same as any other spell. One of the things I like about Dominion's combat system is the way the scripting works as more a "suggestion" than orders. It adds a bit more character to otherwise generic mages to see one get a bit over confident in his abilities and decide he really can take on the entire army by himself.

Besides, you can always force the effect through items if they're that important.

Re: Should returning be overridden at all?
 

Returning should be overridden at times yes. What if your pd uncovers a scout? And what if the only thing you fear is a magic duel. In this case the AI is doing its job in not making you retreat to a scout.

Re: Should returning be overridden at all?
 

Another thing is getting Returning override normal spell, I had my Couatl cast Vortex of Returning instead of Astral Healing after my army annihilated small enemy suicide squad.

Re: Should returning be overridden at all?
 

A related question, does returning work if home province is protected by domes?

My KB says returning should work (sec. 6.14.5). It refers to dom-II so my question is whether or not this was changed in dom-III.

Re: Should returning be overridden at all?
 

Yes, it 100% works

Re: Should returning be overridden at all?
 

One of the most compelling reasons that Returning should be capable of being overridden is that (to my knowledge) no other spell is immune to overriding. Also, it's no more or less "faulty" than all sorts of other AI decisions, whether ill-advisedly following a script, ill-advisedly breaking it, or just extra-script casting, any of which may devastate a battle plan or otherwise cause ruin.

Taking into account the AI's foibles is a dimension that has to be learnt along with everything else.

Re: Should returning be overridden at all?
 

No, but some spells are prevented from being cast without scripting. This is really just the reverse.

Saying, "the AI makes bad decisions, live with it", is not a good argument against improving it where it can be improved, even if it'll still make other bad decisions.

Re: Should returning be overridden at all?
 

If I script Returning it usually means I have a really good reason for it. And I'd be more happy with it working against a scout than not working against teleporting in S9 Magic Dueling Statue.

Re: Should returning be overridden at all?
 

Yes, I absolutely agree. Let it be working always but not "sometimes".
And the worst thing with it is that it is nearly unpredictable when it works or not

Sadly, this thread seems to have no point because I really wanted this to be changed but as eight people (two of ten made a mistake) and one of the devs among them say it is fine, probably, there's nearly no chance of this being changed.

OTOH this problem seems to be rather popular and 26 people of 34 voters say it should not work this way. Maybe, it is enough to at least hope for something? :)

Re: Should returning be overridden at all?
 

ano there are reasons why the returning and scripting in general operates this way.

I believe in previous dominion versions alot of players moaned/complained about constantly wasting gems because there scripting was followed big spells were cast when they faced a small opposing force. Obviously in MP, players used to take advantage of this, hit them with a small force, via teleport/cloud trapeze, trigger the big spells in the magic phase. Then smash there army in the combat phase when they have no gems to cast there big spells. Even a hidden scout following your army does not help in this scenario.

Also, if returning is 100% then the astral golem increases vastly in power as there are fewer counters to it. So it would after be nerfed if returning was made 100% effective. It has been well known for along time that the astral golem's big vulnerability is magic duel and thus has to be used with due care.

Re: Should returning be overridden at all?
 

Astral golem is not that ultimate SC and maybe even not sc at all. It is rather vulnerable in fact even to good troops. And it is no better than well equipped sneaking raiding thugs who ruin your day. These guys also can hardly be dealt with and no one says it is bad.
As for AI algorithm, I think it is just fine because I was one of those people who liked leaving the whole enemy army without gems with a sudden magical attack (Call of the Winds the most often) in Dom PPP. It was a common tactic those days.
And as you could see, I'm not arguing about AI algorithm in general, just about one spell (or two, taking VoR into account) which is indeed very different from any other.

Re: Should returning be overridden at all?
 

I guess since there is a "AI do not cast" spell list, it is logical to have "AI do not overide" list. And if you really like unpredictability, I suppose we can have troops randomly ignore orders (like mages charge into front 1/10 of the time). That surely would spice things up.

Re: Should returning be overridden at all?
 

I have actually already had this in Darwins Zorro! So it already exists in the game. It has to do with casting quickness. Wether its a bug or not I do not know. Probably start a thread when I have more time to discuss it.

Re: Should returning be overridden at all?
 

I think this means a mage has no more spell to cast and decides to "stay behind troops". If you give it "cast spells" order, I have never seen a mage move.

Re: Should returning be overridden at all?
 

No, my 6 mages where firing piercers, scripting went:-

Squad 1.

Mage 1 :- Aim, Quicken Self, Hold/Firex3, Fire&Flee3turns.
Mage 2 :- Quicken Self, Hold/Firex4, Fire&Flee3turns.
Mage 3 :- Quicken Self, Hold/Firex4, Fire&Flee3turns.

Squad 2.
Same.

In theory I should have had 6 14AN range 50 missile attacks with 25'ish precision every round. Which I did.:) But instead of holding still at the back of the battlefield, they had, 'suicidal creep', which means after firing walk a few squares forward. They continue to do it until they meet enemy troops and get killed. They do not even try to stay behind my troops.

Obviously, this makes such a strategy with any bow/crossbow weapon useless.

Re: Should returning be overridden at all?
 

I had a similar thing with W9 blessed yavana archers. The idea was to have them shoot 3 times in 2 turns, but instead they moved a little forward every other turn.

Re: Should returning be overridden at all?
 

Maybe their target was out of range?
In the cases like this the first thing that should be done is looking into debug log (run dom with -ddd option).

Re: Should returning be overridden at all?
 

Targets were in range. One turn they just shoot at the targets and the next they first shoot and then move a little forward. They don't keep their remaining AP to be able to shoot twice next turn, they use them all up by moving forward.

Re: Should returning be overridden at all?
 

Whats happening is when your unit is under the effect of quickness, you seem to have a few action points left over, so you automatically use them up and move forward.

Great if you have a kitted out thug/SC as you want to be up close and personal but very bad for missile weapons. It does not effect spell casters who are quickened, as casting 1 spell uses up all action points.

Its just a problem with missile weapons/quickness combo.

Re: Should returning be overridden at all?
 

Understood. Probably, that's a bug.

Re: Should returning be overridden at all?
 

Back to magic duel and returning. A seraph on defence having armor of virtue and casting phoenix pyre to live through magic duels gets magic dueled. He returns to home province with a new effect on him - Dragon master (3).

Re: Should returning be overridden at all?
 

Nice bug:)