Pet Powers Change:
Recharge times on pet attack powers will no longer be affected by any outside source. This includes buffs and debuffs. What this means is that pets can no longer have the recharge time on their powers increased by player buffs (like Speed Boost) or their recharge time decreased by player or NPC debuffs. This change was made to allow pets to correctly cycle through their attacks instead of getting locked on using the same attack over and over and neglecting to use other available powers.
My only question to the devs is: couldn't just the AI be worked on/fixed/redone instead of a tweak that's going to make many players pissed? If AI is a vital component of the system (after all, that's what makes the majority of NPCs "work"), ensuring that it works flawlessly should be a priority. Why the avoidance?
We've spent *months* exploring alternatives. This change is a) necessary, b) the only way to fix all the issues we're addressing without essentially rebuilding the way the server handles power calculations, c) the least overall damaging to the existing powersets.
Trust me, no one here was happy about this decision, but in the end, it was the only way to go.
So, instead of fixing a few problematic MM sets (e.g. Thugs; namely the bruiser), they've hauled off and nerfed almost virtually all pets and many pseudopets, and obviated the ability of some characters to buff their pets.
Lame, lame, lame, lame, lame.
This is the sort of stupidity I expected with Statesman at the helm. This is not what I expected from the NCsoft team.
They've had this broken-ass pet AI since beta. I can't fucking believe that five years on and multiple AI changes later, we're still enduring nerfs in their half-ass attempts to fix it.
/rage
It's not just the Bruiser. It's basically every pet with more than one attack, including the animated stone, singularity, jack frost, all the MM pets, etc.
As I understand it, the problem is that there is a clause in the pet AI that will, if the wait is short enough, cause them to wait for a power to recharge instead of using another power, as that is the best pick in that case. Once they start getting a certain amount of +rech, that wait becomes short enough that waiting for attack A to recharge is all they will do. This obviously isn't optimal.
From Arcanaville, who spends more time futzing with this stuff than anyone who isn't a dev:
I'm counting everyone that has ever attempted to guess at the internals of the game without first hand knowledge, going back to the people who thought "powers" were "objects" back in '04, who are currently batting 0.00 out of forever.
Out of all the times I could verify their accuracy, none of them has ever been right. Exactly, precisely zero. Not on powers, not on mechanics, not on animations, not on networking, infrastructure, architecture, algorithms. Nada. There are a lot of people out there that have a pretty good handle on different parts of the game, but all of them have hands-on testing and investigative experience with the game, in the *specific* areas of their knowledge. The people who just know, know wrong. The people who think their resume makes them more likely to be right, are really wrong. The people who have logically deduced how the game works from first principles are so wrong their wrongness can be used to amuse small children and animals.
I literally discovered how critters make up their minds how to attack just weeks ago, when I set up a critter in a particular way and then asked pohsyb "so why does it do that?" ANd then he told me. And there is no way I was likely to have guessed that basic algorithm because its not algorithmicly intuitive *or* combat-intuitive. Its simply not intrinsicly guessable. And without knowing what the critters do now, there's no way to guess as to how tweakable it could be without making wholesale changes.
And that means anyone that thinks they do know how it could be tweaked to function is mistaken.
Narbus, why the Numina +/+ and Miracle +rec on your fire/psi? You're getting a full bar of endurance every 64 seconds as well as capped recovery if you hit a few dudes with Drain Psyche, wouldn't they be more useful on another character?
Partly because I'm rich as shit, partly because sometimes Drain Psyche doesn't hit a lot of guys, and partly because I like fighting AVs and hitting ONE guy doesn't give me enough +recovery for much of anything.
EDIT: Oh, hey, WeX. You can't sk/lk anyone at all until you're at least level 10. That's why you had folks stuck at 4 in your mission.
Narbus, why the Numina +/+ and Miracle +rec on your fire/psi? You're getting a full bar of endurance every 64 seconds as well as capped recovery if you hit a few dudes with Drain Psyche, wouldn't they be more useful on another character?
Partly because I'm rich as shit, partly because sometimes Drain Psyche doesn't hit a lot of guys, and partly because I like fighting AVs and hitting ONE guy doesn't give me enough +recovery for much of anything.
EDIT: Oh, hey, WeX. You can't sk/lk anyone at all until you're at least level 10. That's why you had folks stuck at 4 in your mission.
Really? I've found a single Drain Psyche can give me enough +rec for my attack spamming on an AV. Do you have the +hp and +end accolades?
It's not just the Bruiser. It's basically every pet with more than one attack, including the animated stone, singularity, jack frost, all the MM pets, etc.
As I understand it, the problem is that there is a clause in the pet AI that will, if the wait is short enough, cause them to wait for a power to recharge instead of using another power, as that is the best pick in that case. Once they start getting a certain amount of +rech, that wait becomes short enough that waiting for attack A to recharge is all they will do. This obviously isn't optimal.
From Arcanaville, who spends more time futzing with this stuff than anyone who isn't a dev:
I'm counting everyone that has ever attempted to guess at the internals of the game without first hand knowledge, going back to the people who thought "powers" were "objects" back in '04, who are currently batting 0.00 out of forever.
Out of all the times I could verify their accuracy, none of them has ever been right. Exactly, precisely zero. Not on powers, not on mechanics, not on animations, not on networking, infrastructure, architecture, algorithms. Nada. There are a lot of people out there that have a pretty good handle on different parts of the game, but all of them have hands-on testing and investigative experience with the game, in the *specific* areas of their knowledge. The people who just know, know wrong. The people who think their resume makes them more likely to be right, are really wrong. The people who have logically deduced how the game works from first principles are so wrong their wrongness can be used to amuse small children and animals.
I literally discovered how critters make up their minds how to attack just weeks ago, when I set up a critter in a particular way and then asked pohsyb "so why does it do that?" ANd then he told me. And there is no way I was likely to have guessed that basic algorithm because its not algorithmicly intuitive *or* combat-intuitive. Its simply not intrinsicly guessable. And without knowing what the critters do now, there's no way to guess as to how tweakable it could be without making wholesale changes.
And that means anyone that thinks they do know how it could be tweaked to function is mistaken.
So a wide-ranging and significant nerf is preferable to re-writing the way pet AI cycles powers?
I'd also be interested to see if there's anything significant to Arcanaville's comments, other than vague snark and condescension. Did he post something more informative, or is that it?
And don't get me wrong, I wouldn't be flat-out against this if it was coupled with a corresponding buff of pets, but as it is, for a lot of people who have spent a lot of time and money on their characters, this translates to nothing more than a reduction in their damage output. E.g. if Animate Stone had a functional attack chain on his own (and moved at a reasonable speed), I wouldn't care if I could Speed Boost him anymore or not. But he spends half his time just sitting there without it.
Well, "you don't know the code. Guessing at how shit works and how hard it is to rewrite is silly" is just plain true. On my last project we had to skip many, many practical improvements simply because once you got down into the design deep enough they would have required a substantial rewrite to achieve, months of work for something that seems trivial from a user point of view. Such is the reality of working with mountains of old code where the problems in question were never considered. And from the sound of it, these guys are working with the mother of all spaghetti code here.
One could argue that the problem recharge causes is less harmful than the one the fix creates, which is fair enough, but arguing that fixing something is easy after the developers themselves have stated is not is <insert comparison. Possibly with cars>.
Yeah, but who said the fix was easy? Was anyone even arguing that?
We've had five years of the developers intermittently ignoring the pet AI problem, then offering some half-assed work-around. Pet AI has never worked correctly. And now, all of the sudden, when pet AI is better than it ever has been, they're going to almost universally nerf pets over something we've all learned to work around.
Controller pets (and similar, like Voltaic Sentinel, Lightning Cloud, and Dark Servant) don't have attack chains; they benefit from virtually all of the recharge you can throw at them. It's only MM pets that have enough attacks for the cycling to cause a problem, and even then, only with certain upgrades.
I'm not going to sit here, ignorant of how their code works, and say that I have the grand solution. What I do thin, however, is that Castle is patently wrong by saying that this is the only way to fix the problem (without re-writing recharge handling, etc). They have a small team, and this has been a small problem that has received very little attention, and one they've not deigned to interact with the community over. I doubt they've even scratched the service of feasible alternatives, but they're swatting the fly with the howitzer, now.
Besides, someone in that thread had a good point in his sig: this is the entertainment industry. When you have a change that no one is happy with (by Castle's own admission), it would be wise to not make it.
Besides, none of that is even what Arcanaville was talking about. That "first principles" crap is a philosophical reference; it's another term for an axiom. And he's being a first-order, disingenuous douchebag in that post by conflating first principles with what is, essentially, the scientific method, and using that as a basis for some absurd chest-thumping and bashing everyone else who had some theory about how the game worked that didn't pan out.
I don't know about you, but my days of testing are over. I appreciate the people who still do it. And given the set of circumstances people have to work with (e.g. no access to the underlying data), there are bound to be people who come up with models that seem accurate at first glance, but don't quite pan out. It takes a pretty big tool to bitch at them for it.
I don't know the other MM sets well, but I stopped playing my Bots/Traps MM at 32 because he was so boring. I soloed on max difficulty and never ever died.
So are the other MM sets so much weaker that they can't stand a nerf?
I don't know the other MM sets well, but I stopped playing my Bots/Traps MM at 32 because he was so boring. I soloed on max difficulty and never ever died.
So are the other MM sets so much weaker that they can't stand a nerf?
The thing is that it's a "fix" to some MM pets that have attack chain collision problems (e.g. they get stuck waiting for a crappy attack instead of using their good ones).
Unfortunately, Jack Frost, Animate Stone, Fire Imps, Fly Trap, Singularity, Lightning Storm, Voltaic Sentinel, Dark Servant, and a host of other pets - who are technically prone to this problem but realistically don't experience it - are getting nerfed hard just because they happen to share the distinction of being "pets" with MM minions.
tl;dr: they're nerfing everyone with a pet in order to fix some MM pets.
I would think they could at least roll in some buffs to those pets. I mean at least lower their default recharge for attacks or something. Some of them like Voltaic Sentinel could have used a buff even before this change.
Yeah, but who said the fix was easy? Was anyone even arguing that?
We've had five years of the developers intermittently ignoring the pet AI problem, then offering some half-assed work-around. Pet AI has never worked correctly. And now, all of the sudden, when pet AI is better than it ever has been, they're going to almost universally nerf pets over something we've all learned to work around.
Controller pets (and similar, like Voltaic Sentinel, Lightning Cloud, and Dark Servant) don't have attack chains; they benefit from virtually all of the recharge you can throw at them. It's only MM pets that have enough attacks for the cycling to cause a problem, and even then, only with certain upgrades.
I'm not going to sit here, ignorant of how their code works, and say that I have the grand solution. What I do thin, however, is that Castle is patently wrong by saying that this is the only way to fix the problem (without re-writing recharge handling, etc). They have a small team, and this has been a small problem that has received very little attention, and one they've not deigned to interact with the community over. I doubt they've even scratched the service of feasible alternatives, but they're swatting the fly with the howitzer, now.
Besides, someone in that thread had a good point in his sig: this is the entertainment industry. When you have a change that no one is happy with (by Castle's own admission), it would be wise to not make it.
Besides, none of that is even what Arcanaville was talking about. That "first principles" crap is a philosophical reference; it's another term for an axiom. And he's being a first-order, disingenuous douchebag in that post by conflating first principles with what is, essentially, the scientific method, and using that as a basis for some absurd chest-thumping and bashing everyone else who had some theory about how the game worked that didn't pan out.
I don't know about you, but my days of testing are over. I appreciate the people who still do it. And given the set of circumstances people have to work with (e.g. no access to the underlying data), there are bound to be people who come up with models that seem accurate at first glance, but don't quite pan out. It takes a pretty big tool to bitch at them for it.
You start off by saying that no one says the fix is easy, and then go on to say it's just a small problem that they should be able to fix. Honestly, anyone who has developed any certain size application realizes a couple of things:
No matter how well you try to write it, complex code will always be complex code, and the interactions between parts will become harder to change over time.
At some point in time a "small" feature request/bugfix will come around that unintuitively requires rewriting gigantic chunks of the existing codebase due to that.
At that time you have to make an analysis of what is more useful. Do you rewrite the sucker from the ground up for one feature, or do you work on higher priority fixes and features? Generally you provide a work around, and then make a list of other things that will be fixed/possible through the rewrite and make it a new version.
This is, of course, taking them on their word that this is tied to the way damage works on the server, and that it's a complex fix.
jonxp on
Every time you write parallel fifths, Bach kills a kitten.
3DS Friend Code: 2707-1614-5576 PAX Prime 2014 Buttoneering!
What I do thin, however, is that Castle is patently wrong by saying that this is the only way to fix the problem (without re-writing recharge handling, etc).
You start off by saying that no one says the fix is easy, and then go on to say it's just a small problem that they should be able to fix.
Reading comprehension for the lose?
I never said it was easy. Not once. What I said is that this is a small problem; it's something that doesn't pose a significant problem to the player base because we've been dealing with it for years, now, and virtually everyone who has to deal with it has worked around with it. Furthermore, it has been a well-known problem that has received very little attention from the developers in the past; a one-off change here and there that made things better, but never really fixed them for good.
Small, in the scope of the problem it represents, doesn't equate to easy to fix. I never said it was easy, although leaving it alone, and maybe changing out the powers on the problematic pets seems pretty easy to me.
What I do thin, however, is that Castle is patently wrong by saying that this is the only way to fix the problem (without re-writing recharge handling, etc).
Based on what?
Largely on the above: it's a small development team, dealing with an issue that hasn't warranted much attention at all, and has never been discussed with the community, or even beta testers.
It's a resources and creativity thing; NCNC doesn't have many of the former, and they can't even come close to competing with the player base on the latter.
And further because of reading between the lines: the developers are loathe to change powers, even if they're irredeemably bad, for some reason. This is why we still have sets with crap like Dimension Shift, Black Hole, and Torrent. They're tying their own hands by saying that they won't take away the bruiser's Handclap or change Jab to Haymaker before they even start looking at the problem.
I don't see myself noticing this change at all myself. Nobody ever speed boosts my pets, so other than, what, hasten passing over to ice storm or something (which I can't say I noticed a difference with in effect) there's no +recharge on my pets to start with.
If recharge enhancements affected pets attack speeds then it'd be noticable, but they don't.
So, when he said they were looking at alternate solutions for it for months, he was in fact lying?
I'm saying that the set of all possible solutions hasn't been explored via a tiny amount of resources when they've eliminated a huge number of potential solutions - and likely the most poignant ones - before they even started looking.
How does "We've spent *months* exploring alternatives" equate them not even starting to look? And what does the team size matter when they're not likely to get bigger?
So, when he said they were looking at alternate solutions for it for months, he was in fact lying?
I'm saying that the set of all possible solutions hasn't been explored via a tiny amount of resources when they've eliminated a huge number of potential solutions - and likely the most poignant ones - before they even started looking.
But you admit you don't have any idea if this statement is actually true?
The thing is that it's a "fix" to some MM pets that have attack chain collision problems (e.g. they get stuck waiting for a crappy attack instead of using their good ones).
Unfortunately, Jack Frost, Animate Stone, Fire Imps, Fly Trap, Singularity, Lightning Storm, Voltaic Sentinel, Dark Servant, and a host of other pets - who are technically prone to this problem but realistically don't experience it - are getting nerfed hard just because they happen to share the distinction of being "pets" with MM minions.
tl;dr: they're nerfing everyone with a pet in order to fix some MM pets.
(QR)
One of the problems with the Bruiser (and Animate Stone, as well) stems from a bug that exists in the attribute inheritance code. In a nutshell, the more recharge bonus the caster has, the more often these two pets will get stuck in the "Hurl" mode. It's something we're working on.
It's basically every pet with more than one attack, including the animated stone, singularity, jack frost, all the MM pets, etc. Controller pets included, just like I mentioned above. Reading comprehension for the lose?
Also fun to note, Castle made that post over two months ago, so the notion that "they never comment on it," or that "they paid no attention to it," is patently false.
How does "We've spent *months* exploring alternatives" equate them not even starting to look? And what does the team size matter when they're not likely to get bigger?
I swear you just make shit up to respond to. Where did I say that?
Team size - as in the development team size - matters when they're talking about "We've spent *months* exploring alternatives." When that sort of statement comes from a huge development team (e.g. Blizzard), there's a good chance it was a major effort. With NCNC, there's a good chance it means that they had one, maybe two guys with it as one low-priority bullet point a to-do list with a hundred other things.
So, when he said they were looking at alternate solutions for it for months, he was in fact lying?
I'm saying that the set of all possible solutions hasn't been explored via a tiny amount of resources when they've eliminated a huge number of potential solutions - and likely the most poignant ones - before they even started looking.
But you admit you don't have any idea if this statement is actually true?
I'm not sure what sort of resources they've committed to it. They haven't told us. What I am rather sure of is that changing problematic pets' powers wasn't a consideration.
It's basically every pet with more than one attack, including the animated stone, singularity, jack frost, all the MM pets, etc. Controller pets included, just like I mentioned above. Reading comprehension for the lose?
Also fun to note, Castle made that post over two months ago, so the notion that "they never comment on it," or that "they paid no attention to it," is patently false.
But is it a problem for them? No, not really. In fact, the problem for the vast majority of pets being effected by this is purely academic; something so minuscule and/or difficult to quantify that even the most hardcore number crunchers don't care if the pet is throwing three Stone Fists and a Seismic Smash instead of four Stone Fists.
The problem is that some pets - all of them MM pets - have problems with attack chain collision. Do controller pets experience that? Yes. As you said, anyone with more than one attack will. So do players. Pets have had it forever. Why is it all of the sudden worth nerfing everyone over?
And the post from Castle? Where are you coming up with two months ago? It was made two days ago.
(QR)
One of the problems with the Bruiser (and Animate Stone, as well) stems from a bug that exists in the attribute inheritance code. In a nutshell, the more recharge bonus the caster has, the more often these two pets will get stuck in the "Hurl" mode. It's something we're working on.
Seriously. if you post something like "reading comprehension for the lose?" then you need to actually read other people's posts. This is just dumb.
(QR)
One of the problems with the Bruiser (and Animate Stone, as well) stems from a bug that exists in the attribute inheritance code. In a nutshell, the more recharge bonus the caster has, the more often these two pets will get stuck in the "Hurl" mode. It's something we're working on.
Seriously. if you post something like "reading comprehension for the lose?" then you need to actually read other people's posts. This is just dumb.
...I'm talking about the post at issue, here, not your straw man.
And the post at issue was made over two months ago and directly involves the challenges they were facing in making an ENTIRE AT PRIMARY work. Which completely invalidates your "they haven't discussed it at all!" "they've paid NO attention to this problem before!" and "this is a MINOR ISSUE" posts.
EDIT: Hurl is just the most obvious power that doesn't work. I've had my grave knights run right up to an enemy, swords drawn, and cycle gloom. Due to the nature of the AI hurl is a "better" attack than gloom so it shows up more often, but to suggest that an entire archetype isn't affected is just silly.
So, when he said they were looking at alternate solutions for it for months, he was in fact lying?
I'm saying that the set of all possible solutions hasn't been explored via a tiny amount of resources when they've eliminated a huge number of potential solutions - and likely the most poignant ones - before they even started looking.
But you admit you don't have any idea if this statement is actually true?
I'm not sure what sort of resources they've committed to it. They haven't told us. What I am rather sure of is that changing problematic pets' powers wasn't a consideration.
...What the?
The problem is in the AI, not the powers themselves, and if you were referring to the AI...I'm pretty sure that's explicitly what they were looking at. Do you really think they spent the time just playing with their ass?
It really comes off like you just want to bitch about NCsoft for some reason.
Some people like playing a game because they enjoy it. Other people like playing a game because it gives them something to get really mad about. I'm convinced of this. I've seen the same thing over in TF2.
jonxp on
Every time you write parallel fifths, Bach kills a kitten.
3DS Friend Code: 2707-1614-5576 PAX Prime 2014 Buttoneering!
How does "We've spent *months* exploring alternatives" equate them not even starting to look? And what does the team size matter when they're not likely to get bigger?
I swear you just make shit up to respond to. Where did I say that?
Team size - as in the development team size - matters when they're talking about "We've spent *months* exploring alternatives." When that sort of statement comes from a huge development team (e.g. Blizzard), there's a good chance it was a major effort. With NCNC, there's a good chance it means that they had one, maybe two guys with it as one low-priority bullet point a to-do list with a hundred other things.
It doesn't matter when the team size is not likely to increase. What does it matter to them that Blizzard could have 400 people spend a month finding a better solution when it takes them a year to do so with their team size? Where are they supposed to pull the resources from to find this solution of yours?
The only thing I really, really want them to fix is being to put all my toggles on quickly again. It makes things like self res powers a lot more viable. On Mime it's entirely possibly that he'll get mes'd and killed before the toggle comes back up.
I mean if I get knocked out/run out of endurance, I need to retoggle everything and it takes a LONG time, easily enough to get hit with a stun or just KTFO again. It's at the point where if my toggles go down rather than popping an insp I just put on super speed and leg it. It's a relatively small issue in that the character is still a fucking wrecking ball but it's still annoying.
Oh, that I'll agree with. I'd much rather they supressed the toggle until there's enough endurance for it to function than have them all drop because I fucked up and had a big attack queued up. It'd still be equally dangerous, but at least you wouldn't have to spend 10 seconds putting them all on again.
It's also horrible if you have a self rez that doesn't stun/knockback/make you untouchable for a bit. Resurgence is pretty crappy compared to like, Rise of the Phoenix.
The problem is in the AI, not the powers themselves, and if you were referring to the AI...I'm pretty sure that's explicitly what they were looking at. Do you really think they spent the time just playing with their ass?
It really comes off like you just want to bitch about NCsoft for some reason.
Scooter, the problem is the AI. That is true. But they're not fixing the AI. The AI is basically beyond fixing. At least that's the impression I'm getting from the posts I've read over the years. It would require too much work to fix; addressing the AI itself isn't an option.
Here's the problem:
Say you have a theoretical pet. His name is Fred. Fred has three attacks:
Originally - basically from beta through I8 or I9 or whenever it was that they first tweaked pet AI, what would happen is that the pet would look to see what attack he could execute soonest from where he was. Remember how Poo Man would never melee? It's because he would start combat at range, and stay there, because Hurl Boulder was the only attack he could execute at range. Unless a mob ran into melee range, he'd just sit there, chucking boulders and ignoring his other attacks.
So, the original AI would basically go:
A3! okay, can't use A1, can't use A2, I'll just stay here and wait for A3 to recharge. A3! Okay, can't use A1, can't use A2, I'll just stay here and wait for A3 to recharge. Etc.
They changed this by - to the best of my understanding - by getting rid of the "right here" part of the AI. So, in effect, he'd hurl a boulder, then when he looked for the next attack to use, it would be a melee attack, so he'd run into melee range and start using those.
So the revised AI would go:
A3! Okay, A1's available, but I need to get into melee range. Charge! A1! A2! Okay, A3's available, but it's a ranged attack, and I don't like to use those in melee, so I'll just wait for A1. A1! Etc.
However, because of the way the pet AI selects attacks, it something has a sufficiently low recharge time, it won't cycle its attacks. It will basically sit there and say:
A3! Okay, A3's almost recharged. I'll wait for it. A3! Okay, A3's almost recharged, I'll wait for it. A3! Etc.
The thing to keep in mind is that pets inherit your global recharge bonus, and can also proc things like the +recharge IO on their attacks, giving them even more speed. Storm makes really heavy use of this. Lightning Storm, by itself, is a pretty terrible power. But with enough global recharge, it can be a significant damage source, because it inherits the player's recharge bonuses when it's summoned, which makes it spam its one attack much faster.
This is a problem with pets like the Bruiser, who is supposed to be a melee pet, but for lazy players who also have tons of global recharge and can't be arsed to to give a Go To command (and I have a hard time imagining that such a group of players constitutes more than a couple people on any given server), the stupid Bruiser will just sit there and use Hurl, instead of meleeing.
This also causes a problem in melee range, where the bruiser will tend to prefer to whatever attack is in that recharge sweet spot, and reduce how often he uses his other attacks.
Now, Narbus keeps insisting that this effects Animate Stone, too, but my Earth/Kin is sporting anywhere from 130% to 190% recharge at any given time, plus has a Speed Boosted pet, and I've never once seen him get stuck in Hurl Boulder mode since the AI change a while back. And this is why I find the whole thing so distasteful.
The "problem" the devs are correcting is largely overcome by Go To. And the problem - although it might technically apply to controller pets - seems to require a nearly impossible-to-get amount of recharge; so much recharge that you could only really get it via dumping a billion influence into purple sets.
And so, what they're doing is changing pets so that they no longer inherit any recharge. This means that they get a tighter reign on some pets' attack chains, and prevents the pets from being fixated on one attack. This benefits MMs (in some cases), but substantially reduces the effectiveness of non-MM pets, because those pets don't have complete attack chains.
To put this another way: the problem they're trying to combat is to prevent pets from sitting there, not doing anything while they wait for their attacks to recycle. And by making this change, what they're doing is taking pets that don't experience that problem, and making it so that their incomplete attack chain causes them to stand there, not doing anything while they wait for their attacks to cycle.
Or, in yet another way of putting it, because some MM pets have attack chains that don't gel with the AI, their problems get passed on to my pet, which currently doesn't have them.
This prevents buffing your pet as well; so if you (or someone on your team) has Speed Boost, AM, Adrenaline Boost, etc, it will no longer make your pet attack faster. Sure, it might be funny to point and laugh at Fire/Kins getting nerfed, but this hits any other combination of Ice/Fire/Earth/Grav/Plant control and Kinetics/Radiation/Empathy even harder.
Another side effect, which isn't discussed much, is MA. Note that pets aren't subject to -recharge anymore. Think of what that means to ice control and psy attacks when they're fighting NPCs who summon pets.
Sorry for the long post, but I don't know of a way to illustrate the problem any more concisely without cutting out important bits.
How does "We've spent *months* exploring alternatives" equate them not even starting to look? And what does the team size matter when they're not likely to get bigger?
I swear you just make shit up to respond to. Where did I say that?
Team size - as in the development team size - matters when they're talking about "We've spent *months* exploring alternatives." When that sort of statement comes from a huge development team (e.g. Blizzard), there's a good chance it was a major effort. With NCNC, there's a good chance it means that they had one, maybe two guys with it as one low-priority bullet point a to-do list with a hundred other things.
It doesn't matter when the team size is not likely to increase. What does it matter to them that Blizzard could have 400 people spend a month finding a better solution when it takes them a year to do so with their team size? Where are they supposed to pull the resources from to find this solution of yours?
...I didn't say what you're claiming I did in that link. Not even close. How did you manage to not understand that? I didn't say that they didn't even start to look. I said they eliminated a lot of potential solutions before they even started to look.
And the team size is an issue because Castle made the statement trying to placate the angry mob, and it's a crappy metric. It doesn't matter how long they considered it without some indication of how many people and resources were thrown at it. Better to just forget that whole line of reasoning and tell us, matter-of-factly, A) what's wrong, why it's wrong, C) why other solutions don't work.
This is pretty drastic for a pretty insubstantial problem.
How does "We've spent *months* exploring alternatives" equate them not even starting to look? And what does the team size matter when they're not likely to get bigger?
I swear you just make shit up to respond to. Where did I say that?
Team size - as in the development team size - matters when they're talking about "We've spent *months* exploring alternatives." When that sort of statement comes from a huge development team (e.g. Blizzard), there's a good chance it was a major effort. With NCNC, there's a good chance it means that they had one, maybe two guys with it as one low-priority bullet point a to-do list with a hundred other things.
It doesn't matter when the team size is not likely to increase. What does it matter to them that Blizzard could have 400 people spend a month finding a better solution when it takes them a year to do so with their team size? Where are they supposed to pull the resources from to find this solution of yours?
...I didn't say what you're claiming I did in that link. Not even close. How did you manage to not understand that?
Because I assumed you were making that statement taking Castle's post into account, rather than just making stuff about how the developers worked and what they'd done up on the spot. My bad.
And the team size is an issue because Castle made the statement trying to placate the angry mob, and it's a crappy metric. It doesn't matter how long they considered it without some indication of how many people and resources were thrown at it.
All the time and resources they felt they could afford. It doesn't really matter beyond that- they feel it's a necessary change and demanding they show graphs of effort logs to prove they did enough won't change that. Frankly, he could have said anything and you'd still complain it wasn't enough, because you don't like the solution.
Frankly, he could have said anything and you'd still complain it wasn't enough, because you don't like the solution.
This is absolutely untrue. What I don't like is not knowing what's going on; I don't understand why I have to get nerfed so that an AT that is already wildly overpowered can be made even moreso.
Had they said "Oh, we're removing global recharge inheritance on pets, but we'll be taking a look at buffing pets who are adversely effected by this change," I'd be fine with it. As I've already said. But this is just a nerf to a pet that already doesn't have a whole lot going for him, in the case of my earth controller, and the only reasoning offered is, basically, "trust us, it's for the best" by a dev team whose balance decisions are borderline schizophrenic.
Posts
It's not just the Bruiser. It's basically every pet with more than one attack, including the animated stone, singularity, jack frost, all the MM pets, etc.
As I understand it, the problem is that there is a clause in the pet AI that will, if the wait is short enough, cause them to wait for a power to recharge instead of using another power, as that is the best pick in that case. Once they start getting a certain amount of +rech, that wait becomes short enough that waiting for attack A to recharge is all they will do. This obviously isn't optimal.
From Arcanaville, who spends more time futzing with this stuff than anyone who isn't a dev:
Partly because I'm rich as shit, partly because sometimes Drain Psyche doesn't hit a lot of guys, and partly because I like fighting AVs and hitting ONE guy doesn't give me enough +recovery for much of anything.
EDIT: Oh, hey, WeX. You can't sk/lk anyone at all until you're at least level 10. That's why you had folks stuck at 4 in your mission.
I...
FFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFF
*NERDRAGE*
Really? I've found a single Drain Psyche can give me enough +rec for my attack spamming on an AV. Do you have the +hp and +end accolades?
So a wide-ranging and significant nerf is preferable to re-writing the way pet AI cycles powers?
I'd also be interested to see if there's anything significant to Arcanaville's comments, other than vague snark and condescension. Did he post something more informative, or is that it?
And don't get me wrong, I wouldn't be flat-out against this if it was coupled with a corresponding buff of pets, but as it is, for a lot of people who have spent a lot of time and money on their characters, this translates to nothing more than a reduction in their damage output. E.g. if Animate Stone had a functional attack chain on his own (and moved at a reasonable speed), I wouldn't care if I could Speed Boost him anymore or not. But he spends half his time just sitting there without it.
One could argue that the problem recharge causes is less harmful than the one the fix creates, which is fair enough, but arguing that fixing something is easy after the developers themselves have stated is not is <insert comparison. Possibly with cars>.
We've had five years of the developers intermittently ignoring the pet AI problem, then offering some half-assed work-around. Pet AI has never worked correctly. And now, all of the sudden, when pet AI is better than it ever has been, they're going to almost universally nerf pets over something we've all learned to work around.
Controller pets (and similar, like Voltaic Sentinel, Lightning Cloud, and Dark Servant) don't have attack chains; they benefit from virtually all of the recharge you can throw at them. It's only MM pets that have enough attacks for the cycling to cause a problem, and even then, only with certain upgrades.
I'm not going to sit here, ignorant of how their code works, and say that I have the grand solution. What I do thin, however, is that Castle is patently wrong by saying that this is the only way to fix the problem (without re-writing recharge handling, etc). They have a small team, and this has been a small problem that has received very little attention, and one they've not deigned to interact with the community over. I doubt they've even scratched the service of feasible alternatives, but they're swatting the fly with the howitzer, now.
Besides, someone in that thread had a good point in his sig: this is the entertainment industry. When you have a change that no one is happy with (by Castle's own admission), it would be wise to not make it.
Besides, none of that is even what Arcanaville was talking about. That "first principles" crap is a philosophical reference; it's another term for an axiom. And he's being a first-order, disingenuous douchebag in that post by conflating first principles with what is, essentially, the scientific method, and using that as a basis for some absurd chest-thumping and bashing everyone else who had some theory about how the game worked that didn't pan out.
I don't know about you, but my days of testing are over. I appreciate the people who still do it. And given the set of circumstances people have to work with (e.g. no access to the underlying data), there are bound to be people who come up with models that seem accurate at first glance, but don't quite pan out. It takes a pretty big tool to bitch at them for it.
So are the other MM sets so much weaker that they can't stand a nerf?
The thing is that it's a "fix" to some MM pets that have attack chain collision problems (e.g. they get stuck waiting for a crappy attack instead of using their good ones).
Unfortunately, Jack Frost, Animate Stone, Fire Imps, Fly Trap, Singularity, Lightning Storm, Voltaic Sentinel, Dark Servant, and a host of other pets - who are technically prone to this problem but realistically don't experience it - are getting nerfed hard just because they happen to share the distinction of being "pets" with MM minions.
tl;dr: they're nerfing everyone with a pet in order to fix some MM pets.
You start off by saying that no one says the fix is easy, and then go on to say it's just a small problem that they should be able to fix. Honestly, anyone who has developed any certain size application realizes a couple of things:
At that time you have to make an analysis of what is more useful. Do you rewrite the sucker from the ground up for one feature, or do you work on higher priority fixes and features? Generally you provide a work around, and then make a list of other things that will be fixed/possible through the rewrite and make it a new version.
This is, of course, taking them on their word that this is tied to the way damage works on the server, and that it's a complex fix.
3DS Friend Code: 2707-1614-5576
PAX Prime 2014 Buttoneering!
I never said it was easy. Not once. What I said is that this is a small problem; it's something that doesn't pose a significant problem to the player base because we've been dealing with it for years, now, and virtually everyone who has to deal with it has worked around with it. Furthermore, it has been a well-known problem that has received very little attention from the developers in the past; a one-off change here and there that made things better, but never really fixed them for good.
Small, in the scope of the problem it represents, doesn't equate to easy to fix. I never said it was easy, although leaving it alone, and maybe changing out the powers on the problematic pets seems pretty easy to me.
Largely on the above: it's a small development team, dealing with an issue that hasn't warranted much attention at all, and has never been discussed with the community, or even beta testers.
It's a resources and creativity thing; NCNC doesn't have many of the former, and they can't even come close to competing with the player base on the latter.
And further because of reading between the lines: the developers are loathe to change powers, even if they're irredeemably bad, for some reason. This is why we still have sets with crap like Dimension Shift, Black Hole, and Torrent. They're tying their own hands by saying that they won't take away the bruiser's Handclap or change Jab to Haymaker before they even start looking at the problem.
If recharge enhancements affected pets attack speeds then it'd be noticable, but they don't.
But you admit you don't have any idea if this statement is actually true?
It's basically every pet with more than one attack, including the animated stone, singularity, jack frost, all the MM pets, etc. Controller pets included, just like I mentioned above. Reading comprehension for the lose?
Also fun to note, Castle made that post over two months ago, so the notion that "they never comment on it," or that "they paid no attention to it," is patently false.
when it's colorful I click it and when it's all faded and tiny I wait to click it
I swear you just make shit up to respond to. Where did I say that?
Team size - as in the development team size - matters when they're talking about "We've spent *months* exploring alternatives." When that sort of statement comes from a huge development team (e.g. Blizzard), there's a good chance it was a major effort. With NCNC, there's a good chance it means that they had one, maybe two guys with it as one low-priority bullet point a to-do list with a hundred other things.
I'm not sure what sort of resources they've committed to it. They haven't told us. What I am rather sure of is that changing problematic pets' powers wasn't a consideration.
But is it a problem for them? No, not really. In fact, the problem for the vast majority of pets being effected by this is purely academic; something so minuscule and/or difficult to quantify that even the most hardcore number crunchers don't care if the pet is throwing three Stone Fists and a Seismic Smash instead of four Stone Fists.
The problem is that some pets - all of them MM pets - have problems with attack chain collision. Do controller pets experience that? Yes. As you said, anyone with more than one attack will. So do players. Pets have had it forever. Why is it all of the sudden worth nerfing everyone over?
And the post from Castle? Where are you coming up with two months ago? It was made two days ago.
Seriously. if you post something like "reading comprehension for the lose?" then you need to actually read other people's posts. This is just dumb.
...I'm talking about the post at issue, here, not your straw man.
EDIT: Hurl is just the most obvious power that doesn't work. I've had my grave knights run right up to an enemy, swords drawn, and cycle gloom. Due to the nature of the AI hurl is a "better" attack than gloom so it shows up more often, but to suggest that an entire archetype isn't affected is just silly.
...What the?
The problem is in the AI, not the powers themselves, and if you were referring to the AI...I'm pretty sure that's explicitly what they were looking at. Do you really think they spent the time just playing with their ass?
It really comes off like you just want to bitch about NCsoft for some reason.
3DS Friend Code: 2707-1614-5576
PAX Prime 2014 Buttoneering!
It doesn't matter when the team size is not likely to increase. What does it matter to them that Blizzard could have 400 people spend a month finding a better solution when it takes them a year to do so with their team size? Where are they supposed to pull the resources from to find this solution of yours?
Scooter, the problem is the AI. That is true. But they're not fixing the AI. The AI is basically beyond fixing. At least that's the impression I'm getting from the posts I've read over the years. It would require too much work to fix; addressing the AI itself isn't an option.
Here's the problem:
A2: 1s activate, 10s recharge, 200 damage, melee
A3: 1s activate, 20s recharge, 200 damage, ranged
Originally - basically from beta through I8 or I9 or whenever it was that they first tweaked pet AI, what would happen is that the pet would look to see what attack he could execute soonest from where he was. Remember how Poo Man would never melee? It's because he would start combat at range, and stay there, because Hurl Boulder was the only attack he could execute at range. Unless a mob ran into melee range, he'd just sit there, chucking boulders and ignoring his other attacks.
So, the original AI would basically go:
They changed this by - to the best of my understanding - by getting rid of the "right here" part of the AI. So, in effect, he'd hurl a boulder, then when he looked for the next attack to use, it would be a melee attack, so he'd run into melee range and start using those.
So the revised AI would go:
However, because of the way the pet AI selects attacks, it something has a sufficiently low recharge time, it won't cycle its attacks. It will basically sit there and say:
The thing to keep in mind is that pets inherit your global recharge bonus, and can also proc things like the +recharge IO on their attacks, giving them even more speed. Storm makes really heavy use of this. Lightning Storm, by itself, is a pretty terrible power. But with enough global recharge, it can be a significant damage source, because it inherits the player's recharge bonuses when it's summoned, which makes it spam its one attack much faster.
This is a problem with pets like the Bruiser, who is supposed to be a melee pet, but for lazy players who also have tons of global recharge and can't be arsed to to give a Go To command (and I have a hard time imagining that such a group of players constitutes more than a couple people on any given server), the stupid Bruiser will just sit there and use Hurl, instead of meleeing.
This also causes a problem in melee range, where the bruiser will tend to prefer to whatever attack is in that recharge sweet spot, and reduce how often he uses his other attacks.
Now, Narbus keeps insisting that this effects Animate Stone, too, but my Earth/Kin is sporting anywhere from 130% to 190% recharge at any given time, plus has a Speed Boosted pet, and I've never once seen him get stuck in Hurl Boulder mode since the AI change a while back. And this is why I find the whole thing so distasteful.
The "problem" the devs are correcting is largely overcome by Go To. And the problem - although it might technically apply to controller pets - seems to require a nearly impossible-to-get amount of recharge; so much recharge that you could only really get it via dumping a billion influence into purple sets.
And so, what they're doing is changing pets so that they no longer inherit any recharge. This means that they get a tighter reign on some pets' attack chains, and prevents the pets from being fixated on one attack. This benefits MMs (in some cases), but substantially reduces the effectiveness of non-MM pets, because those pets don't have complete attack chains.
To put this another way: the problem they're trying to combat is to prevent pets from sitting there, not doing anything while they wait for their attacks to recycle. And by making this change, what they're doing is taking pets that don't experience that problem, and making it so that their incomplete attack chain causes them to stand there, not doing anything while they wait for their attacks to cycle.
Or, in yet another way of putting it, because some MM pets have attack chains that don't gel with the AI, their problems get passed on to my pet, which currently doesn't have them.
This prevents buffing your pet as well; so if you (or someone on your team) has Speed Boost, AM, Adrenaline Boost, etc, it will no longer make your pet attack faster. Sure, it might be funny to point and laugh at Fire/Kins getting nerfed, but this hits any other combination of Ice/Fire/Earth/Grav/Plant control and Kinetics/Radiation/Empathy even harder.
Another side effect, which isn't discussed much, is MA. Note that pets aren't subject to -recharge anymore. Think of what that means to ice control and psy attacks when they're fighting NPCs who summon pets.
Sorry for the long post, but I don't know of a way to illustrate the problem any more concisely without cutting out important bits.
...I didn't say what you're claiming I did in that link. Not even close. How did you manage to not understand that? I didn't say that they didn't even start to look. I said they eliminated a lot of potential solutions before they even started to look.
And the team size is an issue because Castle made the statement trying to placate the angry mob, and it's a crappy metric. It doesn't matter how long they considered it without some indication of how many people and resources were thrown at it. Better to just forget that whole line of reasoning and tell us, matter-of-factly, A) what's wrong,
This is pretty drastic for a pretty insubstantial problem.
I've always had that problem with the fire armor toggles
combat jumping pops like, immediately, but he has to stop and do this big ARMS OUT I'M HARDCORE flex thing before his plasma shield comes up
All the time and resources they felt they could afford. It doesn't really matter beyond that- they feel it's a necessary change and demanding they show graphs of effort logs to prove they did enough won't change that. Frankly, he could have said anything and you'd still complain it wasn't enough, because you don't like the solution.
This is absolutely untrue. What I don't like is not knowing what's going on; I don't understand why I have to get nerfed so that an AT that is already wildly overpowered can be made even moreso.
Had they said "Oh, we're removing global recharge inheritance on pets, but we'll be taking a look at buffing pets who are adversely effected by this change," I'd be fine with it. As I've already said. But this is just a nerf to a pet that already doesn't have a whole lot going for him, in the case of my earth controller, and the only reasoning offered is, basically, "trust us, it's for the best" by a dev team whose balance decisions are borderline schizophrenic.