Automaton AI Customization

[AyinDygra]

I put forth this idea (buried in my megapost (http://forum.square-enix.com/ffxi/threads/14365-My-Ideal-Expansion?p=190149&viewfull=1#post190149)), that would allow Puppetmasters a great deal more control over their automaton's decisions.

The concept is Automaton AI cards.
Soul Plates can be turned into Artificial Intelligence program cards for Automatons. These are equipped like attachments in 2 new slots (not attachment slots) one in the head, one in the body, and they change the AI of the Automaton, giving players control over the abilities used by the automaton and access to new abilities and traits.

These AI Cards are separated into two groups: Weapon and Ether Control Cards. While they only drop from Archaic Machinery (Gears, Chariots, Ramparts), many Soul Plate cards can be used in the creation of AI cards to enable other special properties. Ether Control Cards can only be created from soul plates taken of enemies with mage jobs. (AI cards are added to the Automaton's attachment list, or something exactly like it, only for cards.)

I'll copy-paste/hide the post here for further discussion in this thread, rather than the megapost.



From my perspective, most of the issues I have with the automaton AI would be fixed with something like this. Maneuvers would be used to manage attachments, not convoluted AI. It also opens up a whole new world of AI cards that give new special attacks, spells, job traits and innate status effects.

[Jar]

or just make the AI not suck..

[xiozen]

Great Post!!! But I think it would require a complete overhaul of the system currently in place... perhaps if you can somehow seriously downscale it so that it incorporates many of the current in-game mechanics without requiring so many major adjustments--dunno... but good ideas nevertheless.

[AyinDygra]

or just make the AI not suck..

I'm afraid it will always be a horrible design concept to force Puppetmasters to use Maneuvers that do not benefit the automaton to trigger weaponskills and spells.

Turning weaponskills, abilities, and spells "On and off" to completely separate the function of Maneuvers from their control over automaton attachments is one of the only logical solutions.

This goes beyond an automaton curing status ailments before curing the master's HP. It's setting up automatons to work with us, not against us. Right now, using certain maneuvers to trigger certain attachment effects wrongly informs the automaton to do things we don't want them to do. These mixed signals are unnecessary.

The system I proposed also added a new feature to the automatons that opens up a whole new world for SE, without adding new attachments for every little thing (we're out of room for new attachments in many configurations)... so adding AI cards that add new abilities to the automatons is a great idea to me. (I'd also love my automaton to be able to make use of mechanical enemy abilities like Restoral and Homing Missile, or have new job traits added.)

(Edit: my system here doesn't 100% preclude the use of Maneuvers to influence the Automaton's decisions, it just stops the automaton from doing things we don't want them to do ever)

[Theytak]

I'm afraid it will always be a horrible design concept to force Puppetmasters to use Maneuvers that do not benefit the automaton to trigger weaponskills and spells.

Turning weaponskills, abilities, and spells "On and off" to completely separate the function of Maneuvers from their control over automaton attachments is one of the only logical solutions.

This goes beyond an automaton curing status ailments before curing the master's HP. It's setting up automatons to work with us, not against us. Right now, using certain maneuvers to trigger certain attachment effects wrongly informs the automaton to do things we don't want them to do. These mixed signals are unnecessary.

The system I proposed also added a new feature to the automatons that opens up a whole new world for SE, without adding new attachments for every little thing (we're out of room for new attachments in many configurations)... so adding AI cards that add new abilities to the automatons is a great idea to me. (I'd also love my automaton to be able to make use of mechanical enemy abilities like Restoral and Homing Missile, or have new job traits added.)

(Edit: my system here doesn't 100% preclude the use of Maneuvers to influence the Automaton's decisions, it just stops the automaton from doing things we don't want them to do ever)
I'd rather they just make the AI not suck than have to learn some new, excessively over-complicated system for an already complex job. I'm 100% sure it would also be easier for SE to just make the AI not suck by fixing the 3-4 issues it has, than to totally overhaul the puppet AI system.

Good effort, though. You got the spirit, you've just gotta work on your level of expectations a bit.

[AyinDygra]

Not to sound overly defensive of my idea, but technically, it's not that big of an overhaul...

Each on/off switch simply adds a check in the automaton logic: where it would normally have done something, now it doesn't, because you said don't ever do that, or "only do that under these simple/specific circumstances."

On the surface this system resembles, but only scratches the surface of, the Gambit system that I'd have otherwise suggested but is clearly too much to add to FFXI. The AI Card system only adds the most basic of options, with the Drain/Aspir tiers thrown in for a little bit more control, and the ability to select specific maneuvers to trigger specific weaponskills (rather than relying on elements that often have conflicting results between weaponskill selection and attachment effects)

All of these checks already exist, from my understanding, including some other variables the automaton takes into consideration at regular (adjustable) intervals. I don't think it would be too hard to learn how to turn abilities and spells on and off, or to remember which maneuver you selected to trigger a weaponskill.

Also, this suggestion actually requires and allows for the current maneuver control scheme to continue in existence, so it's not a whole new system to learn. My addition simply allows for fine-tuning of the choices open to the AI at certain points.

(I also like how my suggestion opens up future possibilities for the job that are currently impractical to implement due to limited attachment slots and elemental capacities, so it "kills two birds with one stone" in effect.)

[Tetsujin]

you're making it sound a lot less complicated than it is. Referencing/reworking/gutting old code, adding new menus, questions of balance and implementation ; the on/off suggestion has been made before, I don't think it a reasonable solution. It also takes the mind out of the automaton. Part of mastering the job is learning how to work with your puppet (despite having a rather frustrating AI that needs some fixing) and working with your puppet includes accepting that it has a psuedo-mind of its own.

[Theytak]

the biggest issues with your idea stem from the fact that SE has stated they aren't going to change the automaton menu layout (in response to the "give us the ability to pre-set/save automaton attachments" idea that we've wanted forever), so adding more to it isn't going to happen (and thus, shouldn't be bothered with in future ideas), while adding those cards to the regular attachment slots means that you lose out on other important attachments for the sake of fixing something that shouldn't need a piece of gear to fix in the first place. This is the same argument used against "Well we could add an attachment that fixes it" or "There's already attachments that fix that (hai2uscanner)" in spite of the fact that the things being fixed are things that shouldn't even be issues. We shouldn't have to use up attachment slots to fix our AI, the AI should just get fixed.

[Camate]

It might be a little early to mention this since the September version update hasn’t even taken place yet, but the development team has already begun preparing for level 99.

To get a glimpse of the direction we are headed with automaton adjustments, the below are the main concepts:

1. Revamp all attachment effects
2. Revamp automaton AI
3. Addition of job abilities that will cure automaton status ailments

We will be bringing you all the detailed info some time after the September version update on the forum and test server.

[esoR]

/hurray!!!

[SpankWustler]

2. Revamp automaton AI

Feels good man.

[Zhronne]

It might be a little early to mention this since the September version update hasn’t even taken place yet, but the development team has already begun preparing for level 99.

To get a glimpse of the direction we are headed with automaton adjustments, the below are the main concepts:

1. Revamp all attachment effects
2. Revamp automaton AI
3. Addition of job abilities that will cure automaton status ailments

We will be bringing you all the detailed info some time after the September version update on the forum and test server.
my my... impressive!

[Chamaan]

I want to be happy, I really do. But I've been hurt before... I'm just not ready to say... it... you know IT...

I guess the only thing we can complain about for the moment is that this fix isn't coming this month. We'll see what they do to it.

[xbobx]

that is great, but can you at least fix the damn NA problem ASAP, since it most likely is a very very easy fix. I think we deserve at least that to be rushed or patch soon after SEpt patch, really dont want to have to wait until sometime in jan/feb

[Alhanelem]

It might be a little early to mention this since the September version update hasn’t even taken place yet, but the development team has already begun preparing for level 99.

To get a glimpse of the direction we are headed with automaton adjustments, the below are the main concepts:

1. Revamp all attachment effects
2. Revamp automaton AI
3. Addition of job abilities that will cure automaton status ailments

We will be bringing you all the detailed info some time after the September version update on the forum and test server.
Seriously some of the best news I've ever seen revealed here. :p

[Nala]

dude camate your my hero!~

Chamaan: lets have some faith in the current dev team... i think they can do this right.

[HimuraKenshyn]

Very very sweet!!!!

[Darkvalkyr]

Get this done, and I'll be a happy PUP again. @.@

[Kaych]

WOHO!!!

..oh wait, I thought he was talking about the useless Spirits for summoners >_>

[xbobx]

Wrong forum Kaych, and no you didn't.

[Kristal]

1. Revamp all attachment effects
2. Revamp automaton AI
3. Addition of job abilities that will cure automaton status ailments

Hopefully bringing an end to the Artificial Stupidity (http://tvtropes.org/pmwiki/pmwiki.php/Main/ArtificialStupidity) :D

[Pyrobunny]

even if they fix the AI.... just means my auto will kill me in my MH insted of some sort of NM fight o_o

THE AUTO REVOLUTION HAS BEGUN RUUUN FOR YOUR LIIIIVES!!!!11!!!!

[xbobx]

Viva la revolucion!!!

Wouldn't that be a great new Puppet head, A hooded head with just the big Hal 9000 red eye. http://loopingsheep.files.wordpress.com/2011/06/the-intellegent-robot-hal-9000.jpg

You think you have control over the puppet, but really, he controls YOU........

[Glamdring]

please, for the love of all that you hold holy (and my continued mental health [such as it is]) let my poor little rdm pup head learn a spell past level 72...

[Glamdring]

Hopefully bringing an end to the Artificial Stupidity (http://tvtropes.org/pmwiki/pmwiki.php/Main/ArtificialStupidity) :D

excellent link Kris, and good info

[Niyariko]

gambit plz, thank you

[Jar]

please, for the love of all that you hold holy (and my continued mental health [such as it is]) let my poor little rdm pup head learn a spell past level 72...
this times about 12890472830914

[Kaych]

Wrong forum Kaych, and no you didn't.

Ohyes I did, Maurahahaha ;D

[Miera]

It might be a little early to mention this since the September version update hasn’t even taken place yet, but the development team has already begun preparing for level 99.

To get a glimpse of the direction we are headed with automaton adjustments, the below are the main concepts:

1. Revamp all attachment effects
2. Revamp automaton AI
3. Addition of job abilities that will cure automaton status ailments

We will be bringing you all the detailed info some time after the September version update on the forum and test server.

I..... I love you.

[Theytak]

2. Revamp automaton AI
All is forgiven. Please proceed with teary-eyed hugs and celebratory drunken dance parties/karaoke contests

[Kajikuro]

I am a newer puppetmaster so i dont have much experience to speak from, but in my experience there is some adjustments to the AI i would like to see.

Valoredge: Works fine as far as ive seen, just run up and stab slash and beat away at the mob as it should.

Sharpshot: I know that the range from the Target that the automation is deployed dictates its actions, but when the situation changes, the auto does not continue as he was told in that if the mob move within his melee range he begins to melee even when he is not the mobs target. there are situations which i wish i could avoid that before it begins mearly with a system to set it up to give him prime directives as simple as "Keep your distance" so that when the mob moves towards him he will intern move away and continue long range bombardment unless he moves to the top of the enmity list, at which time he will hold his position in an attempt to give a chance for his enmity control may be take by somone else if they decide too (as him continuing his directive of distance would lead to an unbalanced abillity to kite mobs as a result of its directive) or "move in and attack" which as it sounds mean move into range to melee for tp and such as normal. This is the Ranger frame so therefore it should in my opinion be able to be given a directive to keep itself away from the enemys AoE range or away from melee range as i, the puppet master, prescribe. This is an attempt to keep my automation being effective in a way i divine as most beneficial to the situation so i dont have him fumbling around by having to recall him and try and determine where to place him to have him be of use to the party.

Stormwalker: As far as the deployment method goes, i believe the same as the one above would service fine. The only problem i have here is that i would like to decide which school of magic he focuses on more that just by controlling my Maneuvers . For example, in this abyssean age there are many mobs that are effected by few enfeebling magics and some that are basicly unfazed by them, if i could only tell him "Focus on healing us" or "Elemental magic should be your focus" and "Disrupt its ability to fight us" and "Keep me at fighting strengh (to remove debuffs and use enhancing magic)" or things of the sort

Soulsoother/Spritreever: These head units are already given a prime directive and follow it relatively well apart from the choice to give SS a prime directive to heal over all else which i dont think would cause as many unbalanced problems as it seems, on the grounds that it cant be gained untill lvl 40+ and if it is focusing on throwing out cures when it can then it will quickly expend it MP and the master will be forced to find a solution, to the to easy solution of its HP being full therefore a master having the option to deactivate and activate to restore MP for unlimated healing power then, on the SS head piece specifically, it could be implemented that unless both HP and MP of the SS auto build are at 100% and/or only the HP of the auto is at 100% then a 45sec-1 minute recast be put on activate to compensate for the MP that could be abused. Similar things could be used on the SR head. The last thing about these frames is a Directive choice between "Use your MP wisely" saving his MP till he decides it is nessacry to act, or "Go all out with your MP" to get him to do as one would expect.

Again these are not intended to make maneuvers useless but mainly specify the automation to a situation better so that they are more effective as i really enjoy PUP and it is the pet job that i want to pursue but i want it to be able to hold up to the bst which throw wave after wave of mindless jug pets that only attack and use tp moves when dirrected and have no other real purpose but tactical cannon fodder, or smn which have a near endless waves themselves that they can throw at the enemy or they can keep them back to heal enhance and protect them as they see fit, while my automation (the pet job with the longest recast timer to bring out there pet and should have the strongest connection to there pet since it is something that they control, setup and must relies on to survive and help them for the longest amount of time before another can be summoned) stand there any watches me be mauled to death and just stands there, or knows that he or i have low hp and that he has the means by which to fix that but will not do so untill he is sent after an enemy instead of him jumping into the air and using "Healing ruby maneuver" or something of the sort without him deciding "I know you used light maneuvers in an attempt to get me to heal you/i buuuuuuuuuuut..... Ima dia that dragon that you deployed me after in the hope that you could retrieve me before i did something like dia it just because i want to" or even just throwing themselves mindlessly into harms why despite the intentions of how and or where it was deployed and which maneuvers have been used to dirrect its actions/job traits/job abillity even further and disregarding how i intended to use him best to just run blindly at the enemy even if i just wanted him to heal me or stand back and shoot/nuke so that he wouldnt run head first into the the Blizzaga IV that just killed him and all the other pets while the bst and smn have new and fully ready pets near instantly and i have to bring back my auto in an unsound state with low hp and mp and rest or do other things to get him back to fighting strengh (waist time when he should have a better or = survival rate and/or redeployment rate as the other pet jobs)

These are the things that i have noticed about the AI of the automation that i think could be improved in the ways described above or in some other way that could better improve the usefullness and survival rate of his job and bring it into a new age of enjoyment.

Reminding that this is all my opinion and understanding of things and not me saying that i am correct or am better than any other person in anyway ^.^

This is all that i could pull off the top of my head so i hope that this or some other measures may be implemented so that i may continue and expand my enjoyment now and in the future as a puppetmaster

[Alhanelem]

the auto does not continue as he was told in that if the mob move within his melee range he begins to melee even when he is not the mobs target. I don't know why so many people insist on keeping the sharpshot at range. Unless you can't keep it alive otherwise with the tools available to you and your party, you should have it in melee range and it should be meleeing so it gets more TP. The automaton ranged attacks way too slow to not have it melee for TP.

If you absolutely can't have it up close because of too many AoEs, then really the mage frame is your best bet. If you have to restrain sharpshot to keep it alive, then you're not doing as much damage as you can with it.

HOWEVER, if you DO want to keep it at range, and the mob moves out of range, the auto should only close to ranged attack range instead of rushing in (I've noticed the behavior in this situation seems inconsistent though...)

[Kajikuro]

The reason i say that the PUP should be given a choice is because there are situations wear the sharpshot frame should not close to melee range. One example being that in places like WoE were 1 shoting AoE happen often and id rather the auto build tp from a range with proper Fire Attachments and land quick accurate shots with Wind to WS often enough so that i can avoid using a mage auto against high lvl magic resistant mobs like that and have the ability to chose what is best and have him act appropriately . A second example is that during a Turul fight with a Smn friend, i didnt have time amid AoE's to properly distance the my auto so i deployed at max range, to which he (automaton) ran up and started whaling on a shock spiked Turul taking a large amount of damage before i realized what he was doing and called him off and being forced to switch to Spiritreaver, which despite using a scanner, decided to often cast Thunder V on Turul. So im not saying that it is something that should always be used but against mobs that magic does little too and a melee puppet will kill itself attacking a mob with strong spikes or aura the option would provide a reasonable way to make the auto act appropriately to the situation.

[Alhanelem]

The reason i say that the PUP should be given a choice is because there are situations wear the sharpshot frame should not close to melee range.In virtually all of those situaitons, you should be switching to the BLM head and frame.

Sharpshot gains TP like a lethargic sloth in a drunken stupor when left out of melee range.

[esoR]

in the situation he gave, he tried blm frame and it didnt work. which supports the argument for changing the placement behavior for the sharpshot. in theory, SS should stay at range shooting, in practice, this doesnt work as well (yet). in my experience, sharpshot has never been my goto frame, even though it's my favorite. either it's too hard to keep alive, to in-accurate, or too slow to deal dmg versus VE or SR.

not sure if there's a better playstyle to keep SS more effective, but in high aoe situations, it just doesnt have the sustainability through all our healing methods that even SR (blm) has. it puts itself in danger by running into aoe range constantly or meleeing for tp, it caps hate way too fast, and requires so much attention that it lowers your overall dps worrying about it. all these factors lead me to doing one of 3 things. either dont melee and keep repositioning it, actively run around moving mobs while meleeing, or simply switch to a more effective frame that syncs with my play style.

/side note
i think we should take some responsibility to keep our pets safe, and not just leave them in harms way, but i agree they should be more defensive in non melee positioning. i think that's a little harder to code though, starting from point A moving to target C is the current system. if its ranger or mage it stops at point B ( minimum attack range.) the issue being aside from manual correction, there's no way for it to know when to retreat from point C, back to B for some reason. though if there was a way, the puppet would probably just run away constantly until retrieved.

thats atleast an idea for an attachment. seeing as basic functions seem to be implemented that way... (scanner,damage gauge, tactical processor)

radar or sonar would work for the name <.<

[Anza]

In virtually all of those situaitons, you should be switching to the BLM head and frame.

Sharpshot gains TP like a lethargic sloth in a drunken stupor when left out of melee range.

I agree 100% with your point that in the current state, you should use Spiritreaver in these situations where you want the puppet at range.

However, I've always wished Sharpshot acted more RNG-like. My ideal situation would be to have the puppet stay at range, but ALSO adjust the damage potential from staying at range. E.g. increase rate of ranged attacks, adjust some attachments (say, Drum Magazine acting like Snapshot), maybe even give the frame some amount of innate regain. Balance would ideally be about the same DoT as the frame can produce currently with melee strikes plus WS/Ranged attacks.

S-E got RNG away from the 2004 style of meleeing for TP and firing off ranged WS, but our automatons still act like these very old school RNGs. Would be cool to me to see them focusing on shooting stuff instead of clubbing it with their arms and blasting away with point blank WS. And it would differentiate Sharpshot and Valoredge a bit more. I still think it's a bit misguided to keep trying (and not really succeeding) at halfhearted attempts to improve VE's tanking ability. Just let it be the melee DD, and make Sharpshot truly a ranged DD.

[xbobx]

I love VE more then SS. I think part is because his weaponskill plays nicely with my set. We have an emp katana and dagger in set and I dont have VS yet. So was we are all spamming ws you tend to see lots of dark chains one after another. The 2k+ extra dmg adds up fast and we tend to take down mobs very fast. VE's ws tends to do almost as much dmg as SS, tbh, I am not sure in my situation SS does more damage.

[esoR]

SS always does the most (physical) spike damage, but overall, i think VE out damages it in time. in the previous job manifesto, SE stated that they wanted VE to rival pld in tanking abilities, or something close to that. it doesn't have the self healing, but it will have pdt, better defense, an ok shield block rate, shield bash, provoke/flash, and weapon skills that increase its enmity. overall VE's tanking abilities are pretty decent, as they need to be.. since each pet job that isnt drg, uses one of its pets to tank things on occasion.

[Dfoley]

I feel that if you get SS to deploy to melee range 100% of the time, you will see it far out damages VE, but thats not practical without slightly gimping SS by using VE head.

[Dfoley]

Check the feedback thread on test server.

They updated cure AI to prioritizes with just 1 light maneuver. Didnt want to make a new thread since last time it got moved to feed back ><

[Castrato]

I like this concept. It has lots of potential. Maybe those useless Cardian cards could have a use after all.

[Theytak]

I've been thinking about how we've been asking for more control over our automaton's WS, and also about how it would be nice to be able to utilize the puppets to proc yellow if we could, and I came up with a wonky-spitballed idea that reminded me of this thread.

Maneuver Cards

Basically, you take cor's quickdraw elemental cards, and combine them with cardian cards, in some low level alchemy synth. I mean the 8 original QD cards, btw, excluding trump. Basically, the cards go in your ammo slot, and act as maneuvers for the automaton, but ONLY in the decision making process. They have no effect on attachment potency. There would be 3 tiers, based on the cardian card used:

Light Crystal:
Quick Draw Card x3
Cardian Card x1

Tier 1:
Cardian 2, 3, 4, Jack cards: NQ: Jack of (element) x33 HQ1: Jack of (element) x66 HQ2/3: Jack of (element) x99
Cardian 5, 6, 7, Queen cards: NQ: Queen of (element) x33 HQ1: Queen of (element) x66 HQ2/3: Queen of (element) x99
Cardian 8, 9, 10, King cards: NQ: King of (element) x33 HQ1: King of (element) x66 HQ2/3: King of (element) x99

full list:
Jack/Queen/King of Flames (Fire Maneuver)
Jack/Queen/King of Tides (Water Maneuver)
Jack/Queen/King of Storms (Thunder Maneuver)
Jack/Queen/King of Tremors (Earth Maneuver)
Jack/Queen/King of Gales (Wind Maneuver)
Jack/Queen/King of Frost (Ice Maneuver)
Jack/Queen/King of Shadows (Dark Maneuver)
Jack/Queen/King of Radiance (Light Maneuver)

Each tier acts as a different number of maneuvers; Jacks are 1, Queens are 2, Kings are 3, and again, this only applies to decision making. Each tier of card will effectively negate one of the actual maneuvers' ability to influence decisions. For example, with King of Storms cards equipped, you can have 3x fire maneuvers active, but valoredge would still use String Shredder. Each ws would consume 1 card. Additionally, although they wouldn't affect attachments like loudspeaker or tension spring, cards could be used in place of, or in addition to, maneuvers that would be consumed by attachments (still capping at 3). For example, having a Jack of Radiance card would allow you to use eraser regardless of what maneuvers you had up. The automaton would prioritize card tiers over maneuvers, so say for barrage turbine, if you had Jack of Gales equipped and 3 wind maneuvers, it would eat 1 card and 2 maneuvers, not 3 cards; queens would be 1 card and 1 maneuver, kings would be 1 card. However for activating these abilities, you would still need to have at least one of the correct maneuver active (so you can't use King of Frost and 3x water maneuvers to tap in to the extra MAB from mana channeler, you'd have to have at least 1 ice maneuver active to trigger ice maker).

However, the cards would work slightly differently with stormwaker/harlequin frames. Instead of deciding ws priority, they would modify spellcasting, depending on the head equipped.
Spiritreaver: Modifies Elemental Spell Tier
Soulsoother: Modifies Cure/Regen Spell Tier
Stormwaker: Modifies both Elemental and Cure Spell Tiers

Basically, each head would react to cards like so:
All: Cards will be consumed for matching maneuver consuming attachments
Valoredge, Sharpshot, and Harlequin: Cards act as top priority for deciding WS
Spiritreaver:
- No card: as is
- King of: Tier V elemental spell matching card
- Queen of: Tier IV elemental spell matching card
- Jack of: Tier III elemental spell matching card
Soulsoother:
- No card: Cure VI, Cure V with less HP lost (I mean how it currently works on the test server); Regen IV
- King of Radiance: Cure V, Cure IV with with less HP lost
- Queen of Radiance: Cure IV, Cure III with less HP lost
- Jack of Radiance: Cure III, Cure II with less HP lost
- King of Tides: Regen III
- Queen of Tides: Regen II
- Jack of Tides: Regen I
Stormwaker:
- No card: as is
- King of: Tier IV elemental spell; Cure III
- Queen of: Tier III elemental spell; Cure II
- Jack of: Tier II elemental spell; Cure I

Note that cards wouldn't allow a spell to be cast outside of its recast timer, its category recast, or the global recast. If the proper timers aren't all up the puppet would cast the spell that would otherwise be its first choice for the given recast/maneuver situation.


Would this be horrible for our inventory? absolutely. Would it be an annoying, unnecessary system? Yep. Would it solve a lot of problems we have with WS priority and spellcast AI? yea it would. Would it be easier to implement than SE fucking with the AI and not telling us what they did? Definitely. But as I said, this is just a spitball brainstorm idea.