Infinitron

FFS, just wait for them to make an announcement.

Holy crap. nipsen, I can tell you in all honesty that Sensuki does not have any special privileges with Josh. On Something Awful, Josh almost makes it a point to ignore him - and god knows there's lots to ignore (no offense, Sensuki!)

Strange indeed. I wonder if you can get a situation where two characters locked in engagement both hit each other with disengagement attacks, if both try to walk away simultaneously.

The changes to the binary nature are related to combat outcomes, rather than engagement. The engagement is a similar process to selecting a response in a conversation; once you've selected it the event is triggered. I don't see a benefit to changing that. Selecting a response in a functionally "turn-based" conversation precisely is trivial. The same cannot be said about maneuvering near and around other characters in a chaotic real-time battlefield.

I assume the countdown would run simultaneously for all characters in the vicinity. How many characters are you expecting to engage? For most classes it can only be one. But if we're talking about a fighter, then that's interesting, maybe there's some kind of gradual time-based mechanism already? Or maybe it's just buggy.

Sure, the bar is just an example. Seeing bars pop up on screen constantly and fill up during combat would definitely be a headache. Doing something with the selection circle could be an interesting option. Make the engagement bar a color that fills the selection circle in a clockwise direction, or shows up alongside it (on its inner or outer rim). I think that done correctly this could be fairly unobtrusive. You could also make it optional, for players who'd prefer to rely on "feel".

It's really not that actiony. Imagine this: When your character is walking towards an enemy capable of engaging him and enters a certain minimum engagement radius around that enemy, some kind of "engagement bar" begins to fill up. The closer he gets, the faster the bar fills up. When the bar fills up, your character becomes engaged. If you get really close to an enemy, or attack him with a melee attack, you won't be able to get away before the bar fills up. If you try to walk around him while inside the engagement radius, you won't be able to pass before the bar fills up. But, if you enter the edge of the engagement radius and immediately about-face, 180 degrees, you can make it out before the bar fills up. This will make the engagement mechanic so much more clear and controllable. It's all about having feedback during engagement, not just after the fact.

It's not really "avoiding engagement". It's giving you a window of opportunity to cancel the movement that would have engaged you in the first place. It's really no different from stepping back from an enemy juuuuuust before the point where he engages you under the current system - just more forgiving.

An animation helps because it isn't instantaneous - it takes time to play and can (should be able to) be cancelled. It gives the player both more visual feedback on what is going on and more control.

1) Don't completely screw up your non-English localization. 2) Don't release right when Steam completely changes its user interface and takes down your ad banner. 3) Don't release at the same time as a bunch of insanely popular Early Access hobo simulators. 4) And yeah, graphics.

Or rather, it is about visual feedback, but it's not enough to know from that feedback that you're engaged after you've already been engaged. In my opinion, it's just as important - maybe even more important - to be clearly aware of the moment in which melee engagement occurs as it occurs, and to be able to control it and cancel it if desired. The thing about being locked into melee engagement is that temporally, it's a binary action. One moment, you're not in melee engagement and free to run around, and the next moment an enemy comes juuuuuust close enough, and boom, now you're locked in place. I believe that in a game that has already taken steps towards acknowledging the necessity for less "swingy" behavior in RPG combat, this is something that is worth addressing. The process of melee engagement shouldn't be a flip of a switch - it should be a visually distinct, gradual and player-controllable process, just like any other in-game action. Thus, as your character closes in on an enemy (or as enemy closes in on him), some animation should begin. This can be a change in the physical stance and/or movement speed of the character model itself, or it can be something that's reflected in the character's selection circle and/or some other symbolic user interface representation. Or both! While this animated process of entering the locked melee engagement state is incomplete, the player can attempt to retreat without suffering a disengagement attack - although he might fail, perhaps if the enemy realizes what he's doing and sticks close to him, or if he's too late to get far enough away before it completes. Obviously, the system would be tuned so that trying to go around a melee character before becoming engaged would be impossible - your only way out would be to retreat backwards, 180 degrees or close to it. And even that should be a bit difficult to pull off, if your timing is imprecise or if the enemy chases you. My inspiration for this system is none other than the much-criticized "Rogue shuffle" animation from Dragon Age: Origins. DA:O was very good about visualizing the transition into and out of melee. You always felt very much in control of the process, as your rogue seamlessly shifted into and out of that slower-moving, hunched, "shuffling" melee stance. It was more-or-less purely cosmetic there, but it would have been very easy to attach an actual game mechanic to it. (Who knows, maybe they even considered that at one point.)

mah ragdoll physics http://forums.obsidian.net/topic/66120-is-there-gonna-be-rag-doll-physics/ http://forums.obsidian.net/topic/66108-will-pillars-of-eternity-have-better-combat-system-than-dragon-age-origins/ What an amusingly successful troll.

Ah, but what if you want to make all attributes useful for all classes, and you also make all of them useful to both "line of fire" characters and non-"line of fire" characters (regardless of class)? Because it sounds to me like Josh is toying with that idea. Well, I was wrong: http://jesawyer.tumblr.com/post/97494143726/re-attributes-i-think-you-need-to-clarify-your-design But in that case, why the need for the INT/RES swap?

Ah, but what if you want to make all attributes useful for all classes, and you also make all of them useful to both "line of fire" characters and non-"line of fire" characters (regardless of class)? Because it sounds to me like Josh is toying with that idea.

I don't see how splitting Health and Stamina would help make CON a more attractive choice for non-"line of fire" classes. It would only make it worse. Both Josh and you seem to be operating under an assumption that more Health/Stamina is an obvious universal benefit to all classes. That's why of all the attributes in Pillars of Eternity, Constitution is the one that's most unaltered from its D&D incarnation. I think that assumption needs to be examined. Especially since Josh now seems to believe that Deflection is also not an obvious universal benefit to all classes, which is why it needs to be beefed up by being placed in the same attribute together with AoEs.

Obviously, but then you can say the same thing about Deflection, which makes Josh's suggestion to swap that effect from RES to INT unnecessary as well.

I think that if they want to keep the current system, the best thing to do in terms of good UI design might be to give up on the traditional IE "filling up with blood" effect and just put two equally sized bars next to or above each portrait, one for health and one for stamina/endurance. But, fans wouldn't like this, so a number may be the next best thing. Or wait, maybe a horizontal bar at the bottom of the portrait?

OK...but if we're now trying to make sure that every attribute is clearly beneficial both to characters who are on the "line of fire" and to those who aren't, what about CON (that is, Health and Stamina)? Is that useful enough to those who aren't getting hit very often?

Here's an idea: In addition to the number, represent health visually as a gradual "greying out" of the character's portrait as he loses health. This would be appropriate since the portraits of dead characters in the IE games went monochrome.

But wouldn't that result in Resolve not being useful to certain characters/classes/builds for the same reason it isn't right now? They all hinge on the character being frequently hit. Only characters that get hit a lot need more healing, only characters that get hit a lot need more Stamina, and, as before, only characters that get hit a lot need better defense against interrupts. You have two or three weak reasons for the stat instead of one, but they're still weak. I see that is possibility, although AI that also targets back line characters especially when they are armed with ranged attack and AoE will probably rise need to put points in Resolve even for back line characters especially it they don't wield armor. So I would like to see how new AI works and couple additional encounters that have several ranged enemies in them before drawing any defined conclusions. Sure, and I realize my argument also applies to Deflection, which is also something that a character who rarely gets attacked theoretically doesn't really need. But, I think Deflection is more useful in that it can it defend you from "opportunistic" attacks. For example, a mage who is relocating to another area on the battlefield doesn't want to get opportunistically smacked by some nearby enemy...but if he does get smacked, it probably won't be often enough that he needs more stamina/better healing, and since he's just moving interruption doesn't matter either.

But wouldn't that result in Resolve not being useful to certain characters/classes/builds for the same reason it isn't right now? They all hinge on the character being frequently hit. Only characters that get hit a lot need more healing, only characters that get hit a lot need more Stamina, and, as before, only characters that get hit a lot need better defense against interrupts. You have two or three weak reasons for the stat instead of one, but they're still weak.

...describe your computer.

Hmmm, I don't think it's that different in terms of intuitiveness. They've just shuffled things around a bit, and added Deflection and Speed into the mix. So, it's actually more complex now.

Definitely check this out, Obsidian devs and everybody else. It is not just your everyday fan wank.