Sorry Sensuki but all those videos only highlight those things

 

1) disengagement attack need cooldown

2) disengagement attack should occure AFTER npc LEAVE engagement zone

3) disengagement attack need animation

4) disengagement attack should not interupt

5) add minimal amount of time spend in zone to attack occure

6) AI is stupid as hell

 

But I generaly agree with your conclusions. Nice research

 

Yup. The system needs fixing, not removing. 

If we go back to the engagement radius,  one fix would be to have a smaller radius to create the *engaged*  status  and a larger radius to trigger the attack.

 

For example, make engaged radius 30% smaller then reach and then have leaving reach be the trigger.   This will stop insta-disengagement attacks.

 

When attacking,  you could designate click-> attack at reach.   shift-click -> engage and attack.

 

 

And obviously,  AI should respond to being engaged rather then moving and procing.  This should be a hard clause in the AI.

Edited December 3, 201411 yr by tdphys

I think one way to create the sticky-ness and lack of kiting that Josh wants without making it completely broken might be to eliminate damage entirely from disengagement and have it ONLY interrupt. Make it a essentially stun-lock feature to by used by you and against you. It would still be abusable, but it wouldn't be so broken in player favor.

 

I am not nearly methodical enough to test this though, so maybe I'm talking out my ass.

 

Edit: Elaborating.

This setup would enable tactical positioning that battles would need with the entrenchment that the dev team was looking for. Think of it as tripping the enemy, instead of getting a full invisible swing of a giant axe for free. If it is interrupt only, those videos sensuki posted play out how they would in a real scenario: as long as the wimpy shooter guy doesn't get caught by the melee, the fighter brutes can crowd their path and control the field like they have been so often described during development.

 

Also, even with the only interrupt system, a limit should be in place to avoid abuse tactics.

Edited December 3, 201411 yr by Kiel29

"It just needs to get fixed.  It just needs tweaking.  It'll be good when it gets patched up.  It's going to be good when it's working."

 

Hey people that love defending this mess, guess what?  This game was supposed to ship a while ago.  It didn't because it was felt by the team that it wasn't ready and needed more work.

 

The reason Sensuki is against this mechanic isn't just because of his personal opinion but because he recognizes that time is money, and money is limited with projects like like these.  Every second being spent on implementing, fixing, and changing this mechanic could be spent elsewhere.  Everyone has their opinion on what's good or not in a game, but the fact of the matter is that a core game play mechanic like this should have been in and working in the game a long time ago at this point in the development.

imo, the engagement system is a good evolution from IE ridiculous combat motion. I believe it can be easely balanced to prevent the exploits you show in your videos. Razsius proposed robusts fixes in this post.

Anyway, since PoE is a solo game, the balance is not as relevant as in competitive multiplayer RTS.
@Sensuki : if you dislike the exploits of engagement attacks ... just don't do it ?

Personally, I like the idea of the engagement system. I am not sure how I like the implementation, as I haven't played enough lately to really assess it. My objection to the suggestion Sensuki has is that it goes against what I prefer to see in a combat system: parity. I want the AI and player to have access to the same exact abilities and skills as possible. This should include intelligence as well/ As much as the combat can be Chess, the better. I don't approve of a system that teaches the AI to do stupid things, unless that can be done to my players as well.

 

So, for example, if I wouldn't be stupid enough to ignore a wizard/ranger/whoever just because a warrior walked in front of me, I wouldn't want the AI to do so either. If I can "taunt" an AI into attacking my warrior against better judgment, then the game should be able to "taunt" my players against my better judgment as well. I understand it is supposed to be a game, but I find it boring/easy to trick the AI into doing stupid things when it can't do the same for me. And yes, this involves kiting. Would you ever allow yourself to be kited? If not, why do we allow AI to let itself be kited? I just don't understand the appeal.

 

In a perfect world, they can fix the system so that it produces natural incentives to avoid running past a tough frontliner instead of just "program the AI to be stupid and attack whoever steps in the way/whatever." That said, if they can't find a way to make it successful, then it should be cut. I'll just enjoy combat that much less.

 

The reason Sensuki is against this mechanic isn't just because of his personal opinion but because he recognizes that time is money, and money is limited with projects like these.  Every second being spent on implementing, fixing, and changing this mechanic could be spent elsewhere.

This.

 

Personally, I'd really love to see some hand-sketched Item descriptions (just like in the Baldur's Gate series), but if i remember correctly -among the devs- Kaz is busy, trying to fix the visual cluttery of the UI part of the Melee engagement system.

 

Well, one can only hope

"It just needs to get fixed.  It just needs tweaking.  It'll be good when it gets patched up.  It's going to be good when it's working."

Hey people that love defending this mess, guess what?  This game was supposed to ship a while ago.  It didn't because it was felt by the team that it wasn't ready and needed more work.

The reason Sensuki is against this mechanic isn't just because of his personal opinion but because he recognizes that time is money, and money is limited with projects like like these.  Every second being spent on implementing, fixing, and changing this mechanic could be spent elsewhere.  Everyone has their opinion on what's good or not in a game, but the fact of the matter is that a core game play mechanic like this should have been in and working in the game a long time ago at this point in the development.

That's terrible logic.  It supposes that completely removing what is obviously a core design mechanic has no opportunity cost of its own. You could turn it on its head and say, every second spent making a new rough mechanic to appease a verbose but limited set of backers could be better spent fine-tuning the existing one.  And in reality, radical changes midway through a game development project almost never signal an uptick in quality.

 

And I doubt the engagement mechanic is the single thing holding back the game like you imply.  Christ, they just added limited VO that's been planned the entire time.  The Linux build isn't even working at all yet.

Edited December 3, 201411 yr by anameforobsidian

The point is not to appease a limited set of backers, but to suggest an efficient solution that gets everyone what they want - with less time spent on (re)designing stuff. Sensuki is saying that a fairly simple AI could achive the goals of engagement with less work. The argument is less about preference than it is about pragmatic implementation.

It's not a core design mechanic. Obviously I need to make more videos against mobs thare are not Medreth's encounter, but if you watched the one I already provided - the sky didn't fall. It looked the same.

 

The Melee Engagement system is abusable by the player, promotes bad AI and removes tactical gameplay from the game. Removing it is only a good thing.

The reason Sensuki is against this mechanic isn't just because of his personal opinion but because he recognizes that time is money, and money is limited with projects like like these.  Every second being spent on implementing, fixing, and changing this mechanic could be spent elsewhere.  Everyone has their opinion on what's good or not in a game, but the fact of the matter is that a core game play mechanic like this should have been in and working in the game a long time ago at this point in the development.

 

Let's not dilute the argument too much here. Removal of ME stands as a worthy goal on its own - because takes away some decision making and reduces the interactivity in combat. In terms of the impact of fixing / developing it on the project schedule we can only second guess at how much time, money and resource OE have left to complete the game. Yes there's a finite amount of all of these, but considering budget alone for example, it would probably be naive to believe that PE's sole source of funding is $3m from Kickstarter.

 

 

eR

 

Great work Sensuki. You've got at least one person who believes you're on the right track - me.

 

 

Having had to chance to finally play this game, a lot of the crazy swinginess from combat I believe actually happens due to engagement. Very few players are aware of the mechanic and even fewer know how to play with it in this game. That makes for a lot of disengagement attacks (hard-hitting too!) that makes combat ridiculously easy (when the enemy gets hit with them) or ridiculously hard (when you've suffering from them).

It's not a core design mechanic. Obviously I need to make more videos against mobs thare are not Medreth's encounter, but if you watched the one I already provided - the sky didn't fall. It looked the same.

 

The Melee Engagement system is abusable by the player, promotes bad AI and removes tactical gameplay from the game. Removing it is only a good thing.

I really am wondering where the people are getting this idea that it's core game mechanic from. It's a minor feature that has been haphazardly thrown in. There's a reason why even this late into development it hasn't been fully implemented; because it wasn't considered important.

 

It's the same reason why it can be removed easily by Sensuki and all the other mechanics still work fine. Because they aren't based on engagement.

Edited December 4, 201411 yr by Namutree

The only thing that is certain is that its abusable. All other opinions on the mechanics have been hashed and rehashed a 100 different ways at this point. What matters now is whether or not the devs can address the abuse.

The only thing that is certain is that its abusable. All other opinions on the mechanics have been hashed and rehashed a 100 different ways at this point. What matters now is whether or not the devs can address the abuse.

I really hope they can. I sure wish they'd give us an idea if this will be getting fixed.

 

 

 I want the AI and player to have access to the same exact abilities and skills as possible. This should include intelligence as well/ As much as the combat can be Chess, the better. I don't approve of a system that teaches the AI to do stupid things, unless that can be done to my players as well.

 

Well then you better hope that engagement gets cut. The existence of the mechanic actively disincentives OE from making the AI good. Engagement is so tactically restrictive that it makes the difference between good AI and bad AI basically moot.

 

 

 

In a perfect world, they can fix the system so that it produces natural incentives to avoid running past a tough frontliner instead of just "program the AI to be stupid and attack whoever steps in the way/whatever." That said, if they can't find a way to make it successful, then it should be cut. I'll just enjoy combat that much less.

In many cases it would be stupid to go after the back row. 

 

Well then you better hope that engagement gets cut. The existence of the mechanic actively disincentives OE from making the AI good. Engagement is so tactically restrictive that it makes the difference between good AI and bad AI basically moot.

What constitutes good or bad AI is largely subjective really. In actuality, systems with engagement mechanics can exist with good AI. Systems without it can have bad AI. Essentially, the issue of AI quality is a separate issue.

 

For example, I think the IE games had essentially bad AI since their targeting clauses often had them stick to the first thing that ran into them despite that not being the most tactically advantageous target. Some would call this good AI as it serves as an abstraction that leads to combat stickiness. Opinions on this come down almost entirely to perspective. This is supremely debatable and we would get nowhere. The only thing everyone agrees on is that the abuse sucks and it should be fixed. Frankly, if it cant, then engagement should go.

 

Why dont we stick with that for now since most agree on that? If we start getting decrying systems again and claiming one side sucks or that the other sucks, this will just unravel into another flame war.

Edited December 4, 201411 yr by Shevek

"I have been meaning to do an official topic on it for some time"

You're not a mod and you're not a dev (at least to my knowledge).  There's nothing official about this, just self-aggrandizing.

"This pledge, combined with the serious problems under the hood makes me really concerned about the amount development time and resources that will be spent on *trying* to fix a mechanic isn't a core mechanic of the game and isn't essential to invoking a modern Infinity Engine experience."

Two things.  One is that as an outsider you're in a precarious danger of presumption trying to define what the "core" mechanics of the game.  The other is replacing it has a similar opportunity cost.  It will take money and resources to change the game either way.  Removing engagement would significantly alter the tactical utility of weapons with reach, and that entire system / art would have to be reexamined.

"Obsidian has stated that they think the problem with Melee Engagement is visual feedback, well as these videos demonstrate, there are far more serious problems than that. Here are some of the components of either the Melee Engagement system or the game itself that make these abuses possible."

The attack resolution system, disproportionate two-handed weapon damage, save-scumming, and pathfinding are not components of the engagement system at all, whatsoever.  If these are problems then they need to be approached separately, and will likely have to be fixed regardless of any change to the engagement system.  It's not constructive to lump every problem the game has on one system.

Furthermore, if Obsidian was dedicated to making the truest recreation of the Infinity Engine, then they would need to put in save-scumming and exploitable AI, and pathfinding.  Where would we be without the games that encouraged reloading on stat-rolls, had demon knights murdering hordes of rabbits, and introduced Drizzt Do'urden to his greatest enemy: a lake.

Yes, that would make it less bad, but it would not fix the abuse. I would still be able to get an initial round of disengagement attacks off on groups of enemies and it would still be the absolute best tactic to use at the start of combat.

It's a non-problem if the game encourages you to start combat by tricking the enemy into an ambush.  That is in and of itself encouraging more tactical thinking than send the fighter in, have everyone hit him, and run away when he's low on health.

 "This is one of the problems with the Melee Engagement system – it promotes simple AI targeting because of the fact that the “first enemy engaged by” targeting clause that is part of the Melee Engagement system would override most ‘smart’ AI targeting clauses for melee units in the first place and melee enemies always re-evaluating for the closest enemy would open up kiting abuse and also disliked by many of the players."

Making the optimal solution of a problem more difficult to achieve should if anything, encourage smarter AI targeting because it more readily exposes weaknesses.  An engagement system forcs devs and users to consider the cost of movement and enemy positions, so theoretically the perfect solution found with an engagement system is better than one without it. 

"Another conclusion you might have come to is that enemies might need to be able to path around Engagement circles or be able to determine when to cross engagement circles of player units. This raises a bunch of concerns such as how that would interact in different environments such as open spaces, and corridors ? How would it interact with dynamic environments (such as the player moving a unit left and right over and over again) ? How would it interact with engagement circle overlapping ? How would it interact with general pathfinding with multiple units ? It seems like a solution that would need very thorough research that is best undertaken during pre-production and with three to four months left on the project, this is not a solution that is even remotely viable to look into."

And here I completely disagree with you.  You raise a potential powerful solution (one that was my immediate go to), and then discard it as impractical without giving it serious thought.  Interacting in dynamic environments means that the game is more reactive to your tactics, and thus a better game.  Engagement circles could be represented as a cost rather than impassibility, so circle overlapping would naturally encourage mobs to avoid ambushes, producing smarter behavior.  And since each unit would be doing its own pathfinding, they should interact in an emergent fashion.

 

And here I completely disagree with you.  You raise a potential powerful solution (one that was my immediate go to), and then discard it as impractical without giving it serious thought.  Interacting in dynamic environments means that the game is more reactive to your tactics, and thus a better game.  Engagement circles could be represented as a cost rather than impassibility, so circle overlapping would naturally encourage mobs to avoid ambushes, producing smarter behavior.  And since each unit would be doing its own pathfinding, they should interact in an emergent fashion.

I don't think the solution is particularly difficult either.

Edited December 4, 201411 yr by Shevek

The engagement system makes it so you can't retreat your injured character from the front-line. Why would anyone want to defend this ****ty mechanic?

False. It just makes it so you can't retreat your injured character from the front-line for free. You can still stun the foe and flee, or use Escape or a similar ability and flee, etc.

 

I don't understand why people think the actual design goal of engagement is "As long as someone with a melee weapon is relatively close, YOU CAN NEVER DO ANYTHING! MUAHAHAHA!" Like melee combatants are supposed to be black holes. No. The problem Josh specifically cited (as Indira quoted above) was the avoidance of melee folk "with impunity."

 

Now, I will say that Sensuki's level of objectivity and research/data presentation is exquisite, and, if engagement were to be abandoned, I would definitely consult with Sensuki's thoughts on the matter. However, I simply still don't believe that the very idea of engagement is somehow inherently bad or wrong, or that there's no way to fix the current system to bring it more in line with its goal, and less in line with the not-its-goal things it's currently accomplishing. Also, I very much agree that it's odd that Team Obsidian have not acknowledged any problems with the current implementation of engagement, compared to what it should be doing (and not-doing).

 

I'm sorry but "look how great playing this exact build of the game is without a flawed implementation of engagement" in no way means "see, it should just be gone from the final game."

 

A) That's great that a lack of engagement works fantastically with the currently pretty-dumb AI, but I'd rather have a game with good AI. If the AI actually handled engagement like a person would, and took advantage of opportunities to disengage and switch targets and such, would the removal of engagement still be so splendid?

 

B) "You can just stop people with active abilities" isn't a very good argument, because that's kind of the point of the "problem" initially referenced by Josh. SINCE people could simply jog past your melee folk with impunity, the only means you had of stopping them, or otherwise hindering them in any fashion, was the use of active abilities. "Crap... my melee guy is useless in preventing that orc from rushing my mage. Better burn a Web or something so that I can specifically deal with that one guy before he gets to me."

 

Basically, with a non-melee character, you can simply direct your attention to some incoming combatant and deal with them rather easily. A spell here. Some arrows there. But, with a melee person, you have to play Benny Hill if you want to try and handle them. So, the whole purpose of engagement is simply to give melee folk an advantage when it comes to handling foes on the battlefield. Not to prevent any and all movement or significant decision-making because the AI's bad and you can disengagement-attack-exploit people to death, etc.

 

So, I dunno. I can look at the current system and say "that's not right," and simultaneously support the conceptual goal of engagement. I can also recognize the merits of the "here's what I would do instead of engagement" arguments, without deciding "yes, that's the obvious thing to do," or "No, that would be the worst thing in the universe and if we don't just keep and fix the mechanic, we're all dead."

 

That's just my 2 cents. If it's useful, great. If not, then drat.

Edited December 4, 201411 yr by Lephys

You're not a mod and you're not a dev (at least to my knowledge).

Don't worry sir, I'm from the internet!

 

There's nothing official about this, just self-aggrandizing.

Yeah there is, I haven't created a thread about the problems with engagement before. I've talked about them but not presented them for others to see. Your disapproval is noted, but no farks have been given.

 

Two things.  One is that as an outsider you're in a precarious danger of presumption trying to define what the "core" mechanics of the game.

It's not. You may have a different definition of what a core mechanic is, and it's obvious that you have an attachment to the mechanic for some reason (but that's okay, some people do) but I removed it from the game by deleting a single method in the code and guess what? The game plays pretty much the same!

 

The other is replacing it has a similar opportunity cost.  It will take money and resources to change the game either way.  Removing engagement would significantly alter the tactical utility of weapons with reach, and that entire system / art would have to be reexamined.

Removing engagement will take less resources to make combat functional than attempting to fix it will. Engagement has nothing to do with reach weapons.

 

The attack resolution system, disproportionate two-handed weapon damage, save-scumming, and pathfinding are not components of the engagement system at all, whatsoever.  If these are problems then they need to be approached separately, and will likely have to be fixed regardless of any change to the engagement system.  It's not constructive to lump every problem the game has on one system.

Based on this statement you didn't read that section of my post you quoted properly. Read it again, multiple times if necessary.

 

Furthermore, if Obsidian was dedicated to making the truest recreation of the Infinity Engine, then they would need to put in save-scumming and exploitable AI, and pathfinding.  Where would we be without the games that encouraged reloading on stat-rolls, had demon knights murdering hordes of rabbits, and introduced Drizzt Do'urden to his greatest enemy: a lake.

This is an argument tactic that many people who don't really care much about an Infinity Engine experience often fall back on. The Infinity Engine games did many things very well and Pillars of Eternity does many things worse currently. Where it does worse I am only asking for at the bare minimum what the Infinity Engine games did, if not something better.

 

Making the optimal solution of a problem more difficult to achieve should if anything, encourage smarter AI targeting because it more readily exposes weaknesses.  An engagement system forces devs and users to consider the cost of movement and enemy positions, so theoretically the perfect solution found with an engagement system is better than one without it.

Nope. I have made several arguments against the opinion that real-time games need automatic systems derived from turn-based games that break the rules of real-time to penalize movement. Movement in itself is already an opportunity cost much of the time, and the Engagement system in Pillars of Eternity promotes simply not moving at all and playing around that. I am a player that plays as optimally as possible most of the time, and moving while engaged is never optimal.

 

And here I completely disagree with you.  You raise a potential powerful solution (one that was my immediate go to), and then discard it as impractical without giving it serious thought.  Interacting in dynamic environments means that the game is more reactive to your tactics, and thus a better game.  Engagement circles could be represented as a cost rather than impassibility, so circle overlapping would naturally encourage mobs to avoid ambushes, producing smarter behavior.  And since each unit would be doing its own pathfinding, they should interact in an emergent fashion.

You can feel free to disagree, but I can think of many ways to abuse such a system and while it is theoretically possible that such a system could be developed, I don't even think Blizzard's pathfinding programmers would be able to implement it in four months.

 

False. It just makes it so you can't retreat your injured character from the front-line for free. You can still stun the foe and flee, or use Escape or a similar ability and flee, etc.

Tactical retreating is never free. Both melee and ranged enemies that are targeting the character of yours can still make attacks against that character in real time. A system that gives an invisible attack that breaks the rules of real-time is completely unnecessary. AoOs are a turn-based mechanic, and they should stay there.

 

A) That's great that a lack of engagement works fantastically with the currently pretty-dumb AI, but I'd rather have a game with [i]good[/i] AI. If the AI actually handled engagement like a person would, and took advantage of opportunities to disengage and switch targets and such, would the removal of engagement [i]still[/i] be so splendid?

It would be even better with good AI, actually. But for some reason people around here seem to think that good AI = enemies that target your backline. Enemies did this in the very earliest Pillars of Eternity builds if you opened combat with your Wizard etc ... and it was laughably easy to exploit.

 

B) "You can just stop people with active abilities" isn't a very good argument, because that's kind of the point of the "problem" initially referenced by Josh. SINCE people could simply jog past your melee folk with impunity, the [i]only[/i] means you had of stopping them, or otherwise hindering them in any fashion, was the use of active abilities. "Crap... my melee guy is useless in preventing that orc from rushing my mage. Better burn a Web or something so that I can specifically deal with [i]that[/i] one guy before he gets to me."

This is completely incorrect. In the Infinity Engine games you could employ positioning, movement, tactical blocking, tactical retreating, manipulation of the AI targeting clauses AND crowd control abilities. Pillars of Eternity's Melee Engagement solution is the polar opposite of this and gets it all backwards.

Edited December 4, 201411 yr by Sensuki

 

The engagement system makes it so you can't retreat your injured character from the front-line. Why would anyone want to defend this ****ty mechanic?

B) "You can just stop people with active abilities" isn't a very good argument, because that's kind of the point of the "problem" initially referenced by Josh. SINCE people could simply jog past your melee folk with impunity, the only means you had of stopping them, or otherwise hindering them in any fashion, was the use of active abilities. "Crap... my melee guy is useless in preventing that orc from rushing my mage. Better burn a Web or something so that I can specifically deal with that one guy before he gets to me."

 

Or you know, give melee classes CC abilities i.e. Knock Down.

Edited December 4, 201411 yr by Seari

 You may have a different definition of what a core mechanic is, and it's obvious that you have an attachment to the mechanic for some reason (but that's okay, some people do) but I removed it from the game by deleting a single method in the code and guess what? The game plays pretty much the same!

 

It blatantly doesn't or else you wouldn't be campaigning for it's removal.

 

This was the OFFICIAL Kjaamor post of the 3rd page. Accept no substitutes.

.... I suspect it is perhaps to allow them to make mobility a part if the character development system and encounter design in general. ...

 

 This could be the reason - though building a character (other than a tank) without engagement breaking abilities seems like a trap choice.

 

 ...They may also simply prefer mobility to have an opportunity cost. ...

 

 I think it has an opportunity cost without the engagement mechanic. Moving to a new target means not finishing off the current target. That is a significant cost because a half dead (or even a 99% dead) enemy continues to inflict damage so switching targets is never free even if you don't need to move to attack the new target.  The engagement mechanic certainly makes it more expensive to switch (or to retreat).

 

 In any case, Sensuki's videos clearly show that engagement is currently very broken. Maybe it can be fixed with some targeting clauses (e.g. if  enemy is attacking but not engaged, change targets upon becoming engaged to avoid the (multiple) disengagement attacks) or in some other way.

 

  I'm leaning towards removing it being the better choice.

 

 Does removing it break anything? It looks like the answer is no, in that it seems to remove exploits without introducing new ones and doesn't seem to adversely change game play with the targeting AI working the way that it does (of course, it may be that future improvements to targeting AI would change this). 

 Does removing it break anything? It looks like the answer is no, in that it seems to remove exploits without introducing new ones and doesn't seem to adversely change game play with the targeting AI working the way that it does (of course, it may be that future improvements to targeting AI would change this). 

There are some issues with various weapon attack animation interaction with moving targets, and the run animation doesn't blend correctly with the combat stance sometimes when chasing. Some programming and animation time will need to be spent on making it so that if your recovery time = 0 and you are in range of your target, your attack animation instantly plays.

 

I'll be making a new thread showing where it doesn't work, and how to easily fix it.

 

 You may have a different definition of what a core mechanic is, and it's obvious that you have an attachment to the mechanic for some reason (but that's okay, some people do) but I removed it from the game by deleting a single method in the code and guess what? The game plays pretty much the same!

 

It blatantly doesn't or else you wouldn't be campaigning for it's removal.

 

It basically does. Sensuki's already released one video showing how it plays the same. He already said he'll release more. 

 

The crux of his argument isn't that engagement will ruin the game. It can be removed via modding after all, but rather getting it to work decent will require way too much work and resources. Especially since it doesn't really contribute much of anything to the game.

Edited December 4, 201411 yr by Namutree