Matt516

Well, here's how the math for any individual interrupt roll works: http://forums.somethingawful.com/showthread.php?threadid=3593502&userid=17931&perpage=40&pagenumber=10#post434357315 Since the constants in that post were from memory and subject to change, I was (and am) hesitant to do the maths on how interrupt affects expected DPS. That and the fact that it's dependent on a number of random factors like attack speed ratio, who attacked first, what weapon they're using, etc. To complicate things further, how much you get interrupted also depends on if you're attacking them and what your interrupt (and their concentration) is since when you interrupt them you reduce the number of times they attack (and interrupt) you. It's very difficult to boil the value of a point in concentration to a single graph without doing some kind of simulation of a bunch of different battles with a bunch of different values for parameters and averaging out the results or something. Even then, my results would largely be dependent on my arbitrary choices of which conditions to test, how to "average" it, etc. It's kind of a mess. Here's a quick summary table I put together from page 22, Gromnir. That should catch you up on the major attribute systems suggested up to that point, though it won't catch you up on the other random discussion re: AoEs and the like (much of which was very good). http://forums.obsidian.net/topic/68526-how-to-fix-the-attribute-design-in-pillars-of-eternity/?p=1509826 Cheers!

Stahp hijacking our thread with bickering. Fight elsewhere pls.

Yarr.

I mean, it doesn't make a difference as long as they balance the base values correctly. So if you know that, you can dump freely without worry.

Yeah, this was an idea I had a while back, but ultimately I was convinced that it would be too hard to balance to make it viable.

Well anyway.... Regarding the "10 is average" thing. While I recognize that there are a few problems with it, my overall feeling is that it will be a change for the better. Maybe they'll want to revise it down to 9 at some point (I dunno), but I think that since the math behind things can be the same either way, and since most of the PoE audience (not all) is used to the D&D way of doing things, it doesn't hurt to frame very low scores as losses. I dunno - I don't really care all that much either way myself, but I got the feeling that a lot of people were clamoring for this change and I understand why. As for how this affects min-maxing? The people who are going to min-max will figure out what they want their stats to be and pick based on the math anyway. The people who want to role-play will be aided, as if you want to play a "dumb" character (for example) it helps to see that combat modifer as a penalty - reinforces the idea that the character really is deficient in some way and not just "smart, but not as smart as that 18 INT guy over there". So ultimately - I don't think it will affect the mechanical balance of the game much, other than encouraging slightly more "middle-of-the-road" builds. And (from my point of view) it aids the role-playing aspects by reinforcing with mechanics the RP "feel" one will be going for if they intentionally dump a stat.

Nipsen, you mentioned that you don't like people playing designers without describing their approach to embarrassing detail. Here's the thing - we did describe our approach in an extremely detailed way. We wrote a 20 page paper describing exactly what our approach was and why we think it results in a better system. So I'll ask you again - did you read the pdf linked on page 1 of this thread? If not, there's not much point in any continued discussion as you will remain uninformed of our actual arguments and the math supporting them. If so, your accusations that we didn't go into enough detail are strange to say the least. You claim that our makeup will never lead to perception and resolve being maxed (and that the current system somehow does? Wat.) You're claiming that our approach somehow lessens the penalties of dumping certain stats compared to in the current system. Both of these assertions are simply... wrong, and I'm really not sure how you arrived at them other than completely misunderstanding either our suggested system or the game mechanics themselves... Other forum-goers - am I missing something? Help me out here.

While you're reading this, Adam (if you still are), how are the attribute changes Josh was talking about (if you're allowed to talk about that, I don't know how it works ) going?

Well, from playing Badminton, I know that if you're in a 1-on-1 game and starting to get exhausted, and the opponent gets the upper hand and delivers powerful shot after shot that you can just barely defend against & recover from, it can take some determination to keep blocking them no matter what it takes until he makes a mistake that allows you to turn things around. (Rather than, having a moment of weakness and deciding to "give up & let him have the point".) Now I've never been in a sword fight (let alone a sword fight to the death ), but I can imagine conceptually similar situations arising there... So, I don't think Resolve increasing Deflection is problematic. I know this rationalization does not realistically explain why Resolve is the only attribute that increases Deflection (i.e. failure of the attribute system to represent things like physical strength or dexterity contributing to evading blows), but note that: It's no worse than tying weapons + spell damage to the single attribute Might (i.e., failure to represent things like physical strength affecting one type of damage but not the other). If we're honest, it's also no worse than some of the D&D rules in the Infinity Engine games -- we're just more used to those. E.g. how melee Thac0 was only affected by STR (i.e. failure to represent things like dexterity/speed/perception affecting to-hit chances). The attribute system is only part of defining your character's abilities -- even with +1 Deflection bonus per point of Resolve, the biggest factor for determining your character's Deflection will still be your choice of class (e.g. "25 + bonus" as fighter vs "5 + bonus" as barbarian). I actually complained about the lack of "realism/intuitiveness" of parts of the attribute system myself in earlier threads, but I've since made peace with it. It's meant to be that way. (In line with Josh's "gamism over simulationism" design principle, which Sensuki and Matt are also supporters of.)As long as they make sure the end result is fun and allows interesting choices, I can live with some impure rationalizations... By the way, I really like this explanation of how Resolve can fit thematically with Deflection. Good on ya.

I still don't see what you mean at all - neither Sensuki and I's suggested system nor Josh's second suggested system (both of which I think would work equally well) encourages a max-min build. They let the player build a character however they want and not be punished for it, provided they then play to that character's strengths. I don't understand your assertion that the current system is somehow less encouraging of max-min stats... This bit here, I'm right there with you. That's true. Sort of. We haven't really changed the impact of attributes, just shifted the bonussen around such that all the attributes are worth roughly the same per-point. Here's where you lose me. Comparing the new suggestions to the old system, nothing's really changed with respect to the viability of running around with a few stats maxed out and a few stats not maxed out. All we've done is ensure that you won't be screwed over in combat if you choose to max the "wrong" stats. If you're referring to maxing a few stats out for conversation checks, well maybe you'll want to do that - but our changes haven't affected that in any way other than allowing you more freedom in which stats to max out while remaining viable in combat. So if you don't like that aspect of the system, I apologize - but our suggested changes have nothing to do with it. Additionally, Josh's news that they're planning on setting the "zero point" for stats at 10 with penalties below 10 and bonussen above 10 discourages only maxing a few stats, so if that's your complaint it will be somewhat addressed. So to summarize..... I'm still really confused as to what your complaint is (or if it is what I think it is i.e. the incentive to max a few stats for conversation options, how our suggested changes have anything to do with it). I apologize for not being able to better address your concerns. EDIT: Ok, done editing.

That's exactly the idea. I dunno, "Boni" sounds like something I'd do to my girlfriend. In the interest of propriety, then, "bonussen" it is.

If this is implemented, I will have at least one character who fights exclusively with single-shot weapons and abilities. Maybe a chanter. Hopefully the weapon switch system will allow you to bypass the reload time of a gun (but not recovery time obviously) by switching to another until they run out (as happened in real fights in this "era" of technology). That'd be fun.

I keep trying to make it happen, but it's just not catching on. EDIT: Bonussen would also be an acceptable alternative. "Bonuses" is just so... boring.

So this design is more about combat focus than roleplaying? And balancing the combat aspect of it? And roleplaying has taken a back seat? Oy. Let me start over. RP concerns such as whether certain stats fit on certain attributes are subjective, and as such are harder to get "right" for everyone. As long as an explanation can be provided that is reasonable, I'm not concerned about having an absolutely perfect correspondence between the attribute names and which stats fit them "perfectly". Deflection makes the most sense from an RP perspective on Dexterity or Perception, but it can be finagled to make sense on Intellect or Resolve as well. It's not perfect, but it works. The combat balance, on the other hand, needs to be correct in order to meet the two primary design goals (because they are combat-related). If the attributes are not roughly equal in power from a combat perspective, dump stats will be created. If the boni of the attributes are not generally applicable to a wide range of builds, then some attributes will not be useful for all classes in some way. To continue with my example from before, it would make the most RP sense to have Deflection and Accuracy both on Perception, or Deflection and Speed both on Dexterity. Problem is, that makes those two stats overpowered in combat and leaves Resolve high and dry. Since Deflection can also make sense on Resolve (albiet with a bit of explanation - Sensuki had one, as does Ineth above), it should go there to balance out the attributes since Resolve on its own with only Concentration is boring. I hope that clears things up - the combat effects of the attributes should be balanced with combat in mind primarily, with RP "fittingness" as a consideration but a secondary consideration. As long as it makes sense in the end, I'm ok with maybe one attribute having a slightly odd stat attached to it (such as Deflection on Resolve). This doesn't affect the roleplaying aspects of the stats in the least. We don't propose any changes to the dialog checks or anything, and in fact Sensuki has actively opposed renaming (for example) Intellect to Cunning for the very reason that it would screw up the dialog checks. TL;DR - RP concerns are important, but our paper is focused on how to balance the combat effects of the stats so combat balance is indeed the primary concern in the paper. We like RP too though.

I suppose the problem we're running into is that you don't actually "balance" the game against it's own design context. Then you just changed the design and did something else instead. And I'm worried that suggestions like this turn up and look like they make sense within the existing design, and are argued to be necessary to make sense of it. But the reason it's really wished for is to make a better max-min build. And you somehow fail to explain that along the way. I'm a bit confused by these assertions - the design goals we used, those of "no bad builds" and "all attributes are useful in some way for all classes" were not our own design goals, but goals stated by Josh Sawyer for his attribute system. As for the second remark - that's not true. These suggestions aren't to allow better min-maxing at all, they're to roughly equalize the impact of each attribute so the player isn't punished for favoring some stats over others. If anything, that's anti min-maxing, though I don't really think it's min-maxing related at all. The whole point of the attribute redesign is to allow players to "create characters that seem to make sense, with weaknesses and strengths, etc." - and we believe our redesign does a better job of that than the current system does. I may have misunderstood you, would you mind clarifying? Because as I read your concerns, they seem to be related to a misunderstanding of our goals for this attribute redesign. We laid out those two design goals above as our ideal, looked at the current system, and came up with an alternate system we believe fits those goals better than the current system. That's all there is to it.

Every party in PoE after this is implemented:

The limited range is just kind of a "necessary evil" thing IMO. I don't think I've ever played a Tactical RPG with an actually accurate representation of the range of a bow. BG did the same thing - it's just a "gamey" thing that allows the action to all fit on one screen while still allowing for tactical considerations such as protecting the back line, etc.

I'd be interested in reading it as well (even without having access to the beta). Dunno if you've seen it yet, but it's been posted for a bit. http://forums.obsidian.net/topic/68526-how-to-fix-the-attribute-design-in-pillars-of-eternity/

Oh Josh's second one is fine. To each their own though. Modding is great.

You're welcome to comment! On the previous page (or maybe the one before? Anyway it's close) I posted a table that sums up all the attribute system ideas people have had so far. The pdf linked on the first page is also good if pretty long (though I'm a bit biased since I Co wrote it ).

But they're still a bonus on top of the (class/level determined) base value, right? Whereas for defenses, the attributes are actually the main input for determining the base value itself. (Or maybe that's a distinction without a difference...?) Anyhow I've left them as dashed lines for now, but updated the captions to reflect that they are not percentage increases: BB278 attributes SM attributes Btw Matt516, are you sure that nerfing the Concentration bonus from +3% to +1% is necessary to balance the additional +1 Deflection bonus to Resolve? Wouldn't +2% Concentration still be fair? After all, Concentration fits really well with RES conceptually (and may even be what the attribute was invented for), so it should have a noticeable impact. I do like your proposal overall; Accuracy and Interrupt will go really well together I think, and I agree that MIG/CON/INT are fine as they are, mechanically speaking -- although a little grating from a simulationist perspective (but that's something we'll just have to live with, given Josh's design principles), and in need of some overall boosting compared to other things that influence the same derived attributes. Yeah, it's a distinction without a difference - 10 Resolve (SM system) will give you 10 Deflection no matter if your base Deflection is 5 or 25. It's all additive - any given one of those is calculated as "base value (for deflection)" + "(lvl-1)*3" + "attribute bonus", where attribute bonus is individual to the stat. Your call though ultimately on how you want to represent that - it's your chart and your system. As for nerfing concentration... I honestly don't know. As I've said, actually quantifying the value of a given concentration percentage is pretty difficult. 2% would probably be fine. I'm not hugely concerned with numbers at the moment though - we suggested numbers that would work, but OE will likely be changing the scale anyway (what with 10 soon to be the "default" value for attributes). I'm just trying to get the stat distribution across attributes balanced such that the numerical values won't have to be wildly different to create a balanced system - so you won't have to, for example, have +5 Deflection on one attribute so another attribute with +5% Duration and +1 Accuracy and +2% Speed won't be more powerful than it. That's an extreme example, but hopefully it illustrates the goal of having the stat distribution across attributes be relatively even - hence the categories I made up of "primary" and "secondary" stats.

I agree with Fiebras - until they implement the stuff everyone's talked about, I don't really have much more feedback to give. Reiterating complaints about lack of combat feedback etc seems pointless - they've heard us and they'll get to it when they get to it. As for balance, we can't really give useful feedback on balance until the major changes to various systems (like combat feedback, defenses calculating correctly, interrupt, attributes) are implemented. Sidenote: a list of "things we're still working on" would be great to include with the patch notes. That way we don't bother beating various dead horses the dev team already plans to fix when the next build is released.

The RP concerns are secondary IMO. As long as it makes reasonable sense, the mechanical balance is the most important thing to me. You're welcome to make suggestions and welcome to disagree - just don't expect Sensuki or I to agree with you on a system that places Concentration as the only major stat on an attribute. Neither of us is a fan of that for various reasons already stated. To each their own though.

I read somewhere (I think on SA) that the bear is getting smaller for sure.

Well no of course not - my statement was more "don't leave bugs that make feedback on the new systems impossible". So a very qualified statement. xD