My thoughts on the issues with combat systems

[Lephys]

First of all... excellent post, as usual,

 

Others have said a ton in here, so I'm just going to pitch in my two cents, rather than go through the whole thing saying "NO" this and "YES" this. I'm just attempting to contribute to the brainstorming on the matter, as many of the things you've suggested make me thing of even more options that could or could not accomplish the same goals, or to a different extent, etc. (Essentially, this is not a re-write of everything you've suggested. If I don't address something, then I probably had no issue with it, and/or only had an idea of something else to consider not solely because I decided it was better already.)

 

 

Attack Resolution

 

I definitely agree about the Attack Resolution system. I think the way in which it works is fantastic, but the numbers were sort of arbitrarily made symmetrical, which was a great idea initially, I think, but they should be adjusted now, in some fashion.

 

Tying crits to the roll itself would accomplish the goal of toning them down a bit, but I don't know that it wouldn't be somehow better to simply tweak the Attack Resolution math, instead. Maybe every 2 points your ACC-DEF increases grants you a point of Crit chance, for example, instead of every 1 point. Each additional positive point of difference would always shift the bottom half of the spectrum down, but would alternate between adding in a point of Hit or Crit, depending. It could be every other point, or every 3rd point, etc. "Hit, Hit, Crit... Hit, Hit, Crit."

 

You still get a greater chance of criticals, and it's still nice and intuitively packed into Attack Resolution, but it doesn't make +15 differences start being ridiculously absurd, etc. In fact, it might even be a good idea to apply the same principle to full misses. Every 2 or 3 points of shift downward, knock off a point of miss. With every 2 points, you'd have to get to +10 ACC-DEF bonus in order to completely eliminate the chance of missing.

 

And, separately, I'd say that the hit-to-graze range ratio shouldn't be 45-45. It should be more like 35(graze)-55(Hit) or something, at ACC-DEF = 0. I fully agree with the many that say we're seeing too many Grazes, but, I think the answer is to simply tone down the Grazes, not change the mechanic somehow. The initial idea was sound -- to get rid of that "more than 10% of the time I'm just going to whiff and accomplish nothing" feeling, and replace it with actual accomplishment much more of the time, with only a relatively small chance to actually do nothing. And 45 Graze with 5 Miss does that. BUT, I think what was neglected was the fact that, half of the time, you still feel like you're severely underperforming. I think the default entire sub-par (be it partial or zero damage) range of the spectrum needs to be smaller than the positive one. Because, in just-hit-or-miss games, you typically don't deal with a ~50% chance to hit. That's already pretty bad, really. So, even though it's been change from full misses, it's still basically a 50% chance to suck, to a player's brain.

 

Armor System

 

While I agree that switching to DR and abandoning integers would be easy to math, I agree with others in that it doesn't really contribute to the tactical/dynamic nature of combat. It makes damage effectiveness more of a static thing, and not at-all situational. Does the foe have more armor? Then you'll do less damage, proportionately. Your Might modifier, being a percentage, might as well just be called DR reduction. They take 50% less damage, and you deal 25% more damage? Then they're basically just taking 25% less damage from you.

 

I realize that relationship remains regardless of the specific system used (any damage reduction and any damage bonus cancel out, to some degree), but DT does it in a much more interesting fashion. If you boil it down, I think that's one of the core tenants of what we look for in an RPG: situational significance, for lack of a better term. Instead of "oh no... this weapon kinda sucks against armor," you get "Oh, this weapon kinda sucks against this particular enemy." That, and there's a lot more of a dynamic at play. If you can't breach a DT regularly, maybe you can augment your Accuracy/critical chance and start dealing significantly more damage, instead of just some proportional increase ("Oh, I was only doing 2 damage, but now that it's doubled, I do 4! YEAH!").

 

Also, I like the idea of armor JUST contributing to DT, and not affecting Deflection. That, and I think, if that were the case, then it might be awesome for only shields to affect Deflection. The thing is, you could still have Deflection be able to be tied to things like armor value, with Talents and the like. "Every 3 points of DT gives you 1 more point of Deflection," or something similar. On the other hand, you could have low-DT characters with high Deflection, which, I think, is much more interesting than "For every additional point of damage protection, I also get better at not-getting-hit as badly."

 

There's just a redundant portion of overlap, there. "This armor is protecting me from damage, but is also deflecting potentially damaging attacks, despite the fact that they're striking 'me'." I mean, if your armor grants Deflection that leads to a Graze instead of a Hit, should it provide half DT against that hit? How can it simultaneously mitigate the damage of the hit, AND contribute to the hit being a less-damaging hit? It's a little weird, to be honest. And, like I said, it opens up a lot more interesting possibilities for, as an example, characters who don't get hit often, but still take massive damage when they do (Because of low DT, but high Deflection).

 

Shields

 

Guess I went kind of the opposite on shields, heh. As above, armor shouldn't enhance Deflection (not defaultly), but should provide DT. And shields should enhance Deflection (defaultly... doesn't mean you couldn't have exceptions that provide DT, for example).

 

Attributes

 

I really think this is still just a big matter of a bunch of little factors contributing to a bad result. Like a meal in which your chicken was overcooked, AND there was too much pepper, AND it was cold, etc. Doesn't mean the ingredients are wrong. It just means several things should've been prepared differently. Cook the chicken less, use less pepper, and actually bring it to the table in a timely fashion. Problem solved. You don't have to cook something else to deliver a tasty dish.

 

Anywho, I think the attributes still have some room for moving effects around, but, to be honest, I couldn't really say, at the moment, what specifically should be done.

 

I still think that, ideally, each attribute should have one primary function, and one (or two, even) secondary functions that are measurably less significant, per attribute point, than their primary functions, in general. But, that's a lot easier said than done.

 

But, replaying New Vegas (I say "replaying" but I never actually finished the game story, just played the crap out of it like a slowpoke who explores everything), I notice very interesting effects like Agility affecting your Action points (primary), but also affecting your weapon reload speed (secondary). The reload speed increase is nice if you focus on guns, so you COULD take Agility just for that, if you really wanted to (depending on your particular build/focus), but action points are highly useful to everyone, too. Someone like myself, who's specialized in Unarmed combat, has some guns for backup, but doesn't really care too much about reload speed.

 

I guess the point is, even if you don't really want to put points into a stat just for some Action Points (maybe you're one of those anti-VATS people, heh), you still have other reasons not to just dump that stat.

 

Anywho, I realize that's kind of what they're trying to do with PoE stats, and again, I realize it's just a bit tricky to get them all feeling like they're on par with one another. But, I think as long as you don't stack primaries and secondaries (like pairing damage bonus with defense bonus or action speed or something), the secondaries don't necessarily need to be "balanced" as much. Looking at the New Vegas example, I'm not as worried about reload speed being balanced against Strength's carry weight bonus, for example. Because, neither of those is really the type of thing that makes a stat a no-brainer or insta-wins combat/the game for you.

 

Interrupts

 

Interrupt is one of those things that might work best if divorced from stats. Or, not necessarily fully removed, but at least treated as something that need not be solely supported by stats. This is where Talents could really shine, once again. And I definitely think it would be great if you could have a variety of options for effects on interrupt, as you've suggested.

 

Basically, I just don't think we need to look at any one stat's effect on Interrupt, then say "Oh no, Interrupt's too insignificant now, but if we increase it, then that stat becomes a no-brainer! *tear out own hair* !", since you can always have Talents bolster someone's interruption capabilities, etc. That and/or traits at character creation. That sort of thing.

 

I think that's the strength of talents, right now. And the stat system really needs to rely more on what Talents could do. I think it tries to cover too great an extent of factors on its own.

Edited November 5, 2014 by Lephys

[constantine]

It inspires me how all of you people, the backers, are doing the developers work. This shows just so much you (and I) desire a game like the one we were promised back on Kickstarter days.

 

Obsidian should be fully aware that your work & feedback will save their reputation & pay more attention to topics like this one.

[Malekith]

All of the kickstarters benefited hugely from backer feedback. Wasteland 2 was terrible at the start of the beta, and came out as it did thanks to backer feedback.

Divinity Original Sin the same. While not so bad by itself, backers proposed many cool ideas and features that Larian delayed the game just to include them.

The result? Better game for all.

And now Obsidian is doing the same. Backer feedback is one of the better things about kickstarter

[Cubiq]

I'm not really a fan of interrupt mechanic working on melee :S

Since it can affect nearly everything, i think it works like a short duration stun with diminishing returns that requires no special ability to execute.

I also feel like it tips the balance of power way too quickly in combat. Once one of the enemy is dead the rest quickly get overwhelmed, since your free character can now do aditional damage as well as interrupt others, preventing them to retort.

I don't think it's very noticable right now because the enemy can 1-2 shot you, so the intensity of the combat stays okey near the end of the fight.

But once this gets toned down, i think it will make combat too 1sided near the end.

[Sensuki]

I think there's one thing that I have not considered for Dexterity (Action Speed) and flat attack speed increases in general and that is the fact that it allows you to make more interrupt rolls, that's a hard thing to calculate how effective that is mathematically as well.

 

Anyway thanks for everyone's input. Josh also replied to me via PM (as originally I sent a very similar version of this post to him via PM) and I will consider the input I got from everyone before deciding to investigate any of these issues further.

Hi, Sensuki. Right now I'm trying to focus most of my efforts internally on bugfixing and getting team feedback. That isn't to say that we're not doing any tuning, but we're in a stage right now where I'm trying to temporarily limit outside input. At a high level, my feedback would be that I don't think it would be good to switch to "natural die roll only" for crits for a few reasons. First, it creates a separate type of resolution for attacks, one based on modified die roll and one based on natural die roll. Second, it means that "true" crits and misses are things that the player achieves purely through random chance instead of by pushing the odds in their favor. That's the AD&D style, but I never thought that resolution model was rewarding because it never happened based on me doing something other than rolling the die.

 

That said, I could see a scenario where crits occur on higher rolls, e.g. 105+, meaning you need more significant Accuracy advantage to even have a chance of pulling them off. Even so, I don't personally think that the Crit system in place causes serious problems. While +50% damage or duration is not trivial, it's also not insanely powerful. In some cases, the Crit exacerbates what is an extant balance problem like a Paralyzed effect with a base 20s duration, which is nuts. In that case, the base effect needs to be modified.

 

DT can be harder to balance than DR and DT + DR can be even harder, but DR alone has a lot of problems. First, DR-reliant systems cause a huge inflation of damage values as the DR values start to rise and the characters' HP are also rising. Unless you tightly control DR escalation (a problem in its own right) eventually you wind up with armor absorbing 80%+ of incoming damage on massive hits against characters with enormous amounts of HP. In Fallout and F2, DR was the primary component of high end armor and it led to plinking back and forth until someone was hit with an armor-bypassing x3 crit to completely annihilate the target. With DR, you also eventually "run out of runway" in terms of equipment progression. You have 100 points to play with, that's it. Unless the progression is really tightly constrained (in which case you won't notice much change), you will eventually top out where the equipment can go.

 

Though it's more difficult to balance DT than DR, it's not incredibly hard and in the BB areas I really haven't done much tuning at all. People have moved around creatures and whatnot but I haven't been trying to create the perfectly balanced area or anything. There's another element that can help alleviate some of the pain of low damage vs. a particular DT type, which is to raise Min Dam Through DT to 20%. It's currently 10%, which is half of what it was in F:NV.

 

I've talked to QA about Dexterity and they completely disagree with it being a dump stat. Its effects are harder to observe but speed is always valuable.

Armored Grace is not the same as an 18 Dexterity because Armored Grace affects Recovery Time, not base speed. Dexterity actually directly affects base speed, which means in turn it affects Recovery Time because RT is derived from the time to perform an action. Weapon Focus grants the same Accuracy bonus as 20 Per to one class of weapons and has no effect on Reflexes. It also stacks, so I don't really see how that creates a problem.

 

I could see bonus Accuracy from Per going from +1 to +2 and Dam/Heal from +2 to +3, but from +1 to +3 is enormously influential. An 18 would be +21, greater than what you'd get from an 18 on an attack stat in A/D&D (already too high, IMO).

 

QA also disagrees with you on Intellect, though we're doubling the AoE effect that it has because it was never actually supposed to be 3%, but 5%. After fooling around a bit, 6% seems a bit better. It has a large influence on AoE. While you could argue that classes don't necessarily need huge AoEs for their attacks/spells, watching them dramatically shrink is actually a real problem that they have to deal with.

 

Interrupts and Concentration need to be communicated better on equipment/attacks and the character sheets since people still don't really make the connection that the basic offensive component comes from the attack.

 

Edited November 5, 2014 by Sensuki

[Cubiq]

There's another element that can help alleviate some of the pain of low damage vs. a particular DT type, which is to raise Min Dam Through DT to 20%. It's currently 10%, which is half of what it was in F:NV.

I hope they do something like this, it should normalize the damage a bit.

 

And looking at the Intellect section of the reply, it looks like they still have no plans to implement a mechanic that allows you to lower the size of the AoE circle.

As long as that's the case, the +AoE% increase will be a very poor stat for any class that isn't a Priest.

[archangel979]

Second, it means that "true" crits and misses are things that the player achieves purely through random chance instead of by pushing the odds in their favor. That's the AD&D style, but I never thought that resolution model was rewarding because it never happened based on me doing something other than rolling the die.

And their system does not roll dice?! And in IE games you could take weapon group proficiencies and improve your critical chance just like in PoE. I am sorry but this explanation does not work. He would be right if there were abilities that granted you 100% critical hit in exchange for something else (like defense penalty for a while or move speed penalty).

 

And I don't like it how they don't care about making attributes more powerful (which again makes no sense because he does not want to improve accuracy but wants to improve area of effect bonuses). Just lower base accuracy for all classes and give more bonus with Perception. Set it up so characters with about 15 perception have same total accuracy as before but those that will put less into it, they will feel it. Everyone is maxing perceptions anyways, no class is better with lower accuracy because if I understood right all physical and magical attacks use this common accuracy value (which is a big difference from older D&D rules and even in 5e each class uses a different main attribute to generate accuracy for different attack and spells).

 

I really don't see how is there a choice currently in the game when accuracy beats all other stats and it is on one attribute.

And by keeping it also affecting critical chances it just makes it worse and more important to max.

Edited November 5, 2014 by archangel979

[Lephys]

And their system does not roll dice?! And in IE games you could take weapon group proficiencies and improve your critical chance just like in PoE. I am sorry but this explanation does not work. He would be right if there were abilities that granted you 100% critical hit in exchange for something else (like defense penalty for a while or move speed penalty).

No no no... he said "purely" through chance. If you get a crit on rolls of 96-100, for example, you can't affect what the dice land on. You could roll 95 or below 50 times in a row, and always get hits, but never crits. OR, you could roll 20 criticals in that same set of rolls. Doesn't matter what else you do or decide, as long as the dice get rolled that many times.

 

What you CAN actively effective on a tactical, in-the-moment level is the difference between your ACC and the DEF of your target, by choosing a different target, buffing, debuffing, etc. That's what the whole Attack Resolution system is based on: the actual situational influence of characters' Accuracy and defense values. He doesn't want to have one thing (criticals) be completely divorced from that system and tied purely to the actual dice roll, while everything else is a part of that system.

 

In other words, he doesn't want two separate systems. You can already modify any number of things within the given system. There are already abilities that convert a portion of grazes to hits, etc., so you can already directly affect the critical range if you wish to. You can have talents, as it stands, that passively increase your critical chance. That doesn't mean that divorcing only critical hits from the rest of the Attack Resolution system somehow doesn't make the game unnecessarily more complex.

 

Also, I didn't see anywhere where he said they flatout don't care about making attributes more powerful. He just mentioned several improvements to attribute modifiers. And if Accuracy is currently too powerful, then maybe suggest how to change Accuracy, not insist that attributes be changed to indirectly sort of maybe make Accuracy a bit better. You can reduce the graze range (making not-putting points into Perception less devastating to your ability to hit stuff well), and/or adjust any of the other ranges as you see fit in various ways.

 

I don't understand why people are so allergic to the idea of tuning systems, and are so ready to jump straight to complete remodeling.

[archangel979]

Change is so it does not effect critical hits because too many good bonuses are dependent on it. It is too useful as it is.

 

Or Change it so different attributes give bonus to accuracy for different kind of attacks. Put melee accuracy with a stat that is less useful for melee characters and spell accuracy with a stat that is less useful for casters.

 

Or accept that attribute balance means little in this system when accuracy is always going to be maxed. Why? Because limited access abilities depend on it too much that one time you use them. 

So maybe remove accuracy from attributes completely and just have it go up by level and when choosing talent/class ability. This is actually how spell saves worked in IE games. They were not influenced by attributes but only by class level and race. Then you could modify them during game with equipment or spells (and lower enemy saves with spells). That sounds like a system Sawyer wants. Better than what he has now.

 

Or accept that the game is better if you take a more simulationist approach (frack balance and give us flavor) and just let us have high Int wizards and high Str warriors and high Dex Rangers. 

Edited November 6, 2014 by archangel979

[Quetzalcoatl]

Change is so it does not effect critical hits because too many good bonuses are dependent on it. It is too useful as it is.

 

Or Change it so different attributes give bonus to accuracy for different kind of attacks. Put melee accuracy with a stat that is less useful for melee characters and spell accuracy with a stat that is less useful for casters.

 

Or accept that attribute balance means little in this system when accuracy is always going to be maxed. Why? Because limited access abilities depend on it too much that one time you use them. 

So maybe remove accuracy from attributes completely and just have it go up by level and when choosing talent/class ability. This is actually how spell saves worked in IE games. They were not influenced by attributes but only by class level and race. Then you could modify them during game with equipment or spells (and lower enemy saves with spells). That sounds like a system Sawyer wants. Better than what he has now.

 

Or accept that the game is better if you take a more simulationist approach (frack balance and give us flavor) and just let us have high Int wizards and high Str warriors and high Dex Rangers. 

It's not how saving throws worked in 3.x ed. I also find it amusing that despite your constant pleas for simulationism, you prefer a system where saving throws are determined essentially randomly over a system where saving throws are derived logically from related stats (Int/Resolve for Will saves, Dex/Perception for Reflex saves and Str/Con for Fortitude saves).

[Hiro Protagonist II]

Josh also replied to me via PM #as originally I sent a very similar version of this post to him via PM# and I will consider the input I got from everyone before deciding to investigate any of these issues further#

 

 

Hi, Sensuki, Right now I'm trying to focus most of my efforts internally on bugfixing and getting team feedback. That isn't to say that we're not doing any tuning, but we're in a stage right now where I'm trying to temporarily limit outside input

 

 

Pretty much what I said in this post. Obsidian are bug fixing. Not really taking suggestions for major changes.

Edited November 6, 2014 by Hiro Protagonist II

[archangel979]

 

Change is so it does not effect critical hits because too many good bonuses are dependent on it. It is too useful as it is.

 

Or Change it so different attributes give bonus to accuracy for different kind of attacks. Put melee accuracy with a stat that is less useful for melee characters and spell accuracy with a stat that is less useful for casters.

 

Or accept that attribute balance means little in this system when accuracy is always going to be maxed. Why? Because limited access abilities depend on it too much that one time you use them. 

So maybe remove accuracy from attributes completely and just have it go up by level and when choosing talent/class ability. This is actually how spell saves worked in IE games. They were not influenced by attributes but only by class level and race. Then you could modify them during game with equipment or spells (and lower enemy saves with spells). That sounds like a system Sawyer wants. Better than what he has now.

 

Or accept that the game is better if you take a more simulationist approach (frack balance and give us flavor) and just let us have high Int wizards and high Str warriors and high Dex Rangers. 

It's not how saving throws worked in 3.x ed. I also find it amusing that despite your constant pleas for simulationism, you prefer a system where saving throws are determined essentially randomly over a system where saving throws are derived logically from related stats (Int/Resolve for Will saves, Dex/Perception for Reflex saves and Str/Con for Fortitude saves).

 

That was just one of the suggestion based on how Sawyer wants the game to work. And it was a fix for Accuracy for spells, not saves. You didn't need accuracy for spells in IE games. As you said, that is something 3e invented with spellcasters setting up their spells difficulty by combining Int bonus and level of spell. The system was like PoE but rolled differently, the target rolled his save vs spellcasters DC where in PoE spellcaster rolls his attack vs target DC. 

 

It is not a suggestion I would choose out of the ones I suggested in previous post. I would choose the "screw balance, give us flavor through simulationism". Attributes in real life are not balanced, and attributes in games should not be. Seek balance in other design elements. 

D&D also tried to balance stuff (although not very well), which is seen by low HP wizards, more XP needed to level, crappy Thaco and worst saves. Bards and Rogues level really fast to balance their weaker combat roles and lack of spells and they get access to abilities only they can use. 

Druids have weaker spells and level faster than wizards and clerics. 

Edited November 6, 2014 by archangel979

[Lephys]

Attributes in real life are not balanced, and attributes in games should not be. Seek balance in other design elements.

Sure they are. You just don't get to choose how many points to put into which attributes in real life, is all.

[archangel979]

 

Attributes in real life are not balanced, and attributes in games should not be. Seek balance in other design elements.

Sure they are. You just don't get to choose how many points to put into which attributes in real life, is all.

 

No they are not. Intelligent and wise people earn 10x more then fast or strong ones. Compare top IT company owners or businessmen  to top athletes. 

[constantine]

Unfortunately if criticals do not occur on a natural die roll, like Sensuki suggested, and their occurence can be affected by the player, *anything* that can add to Accuracy will be too important to miss...

 

+50% on what you do, be it dmg or durations, is really, really important, despite what Josh & Team may think.

 

So, to begin with, Perception stays King, as do talents that give Acc bonuses.

 

As for Dexterity, c'mon let's be honest, who can find a reasonable cause to prefer it over other attributes, except for role-playing purposes, scripted interractions and such.

 

I can see an important reason to make crts happen on the Natural 20, and that is to keep combat encounters more controlled. The drawback to this is that major lv differences will not make obvious impact, and that may be a bad thing.

[Captain Shrek]

Unfortunately if criticals do not occur on a natural die roll, like Sensuki suggested, and their occurence can be affected by the player, *anything* that can add to Accuracy will be too important to miss...

 

+50% on what you do, be it dmg or durations, is really, really important, despite what Josh & Team may think.

 

So, to begin with, Perception stays King, as do talents that give Acc bonuses.

 

As for Dexterity, c'mon let's be honest, who can find a reasonable cause to prefer it over other attributes, except for role-playing purposes, scripted interractions and such.

 

I can see an important reason to make crts happen on the Natural 20, and that is to keep combat encounters more controlled. The drawback to this is that major lv differences will not make obvious impact, and that may be a bad thing.

 

 

Hmm. A crit chance of 5% constant is a bummer. I liked how D&D handled that based on the weapon. Add to those feats and you can build a char around crit-hit. And that is a fair trade.

 

Right now, crit chance in PoE is a floating number that can quickly dominate the damage dealt. I kinda like it actually, as it makes the game much simpler. 

[Mr. Magniloquent]

Unfortunately if criticals do not occur on a natural die roll, like Sensuki suggested, and their occurence can be affected by the player, *anything* that can add to Accuracy will be too important to miss...

 

+50% on what you do, be it dmg or durations, is really, really important, despite what Josh & Team may think.

 

So, to begin with, Perception stays King, as do talents that give Acc bonuses.

 

As for Dexterity, c'mon let's be honest, who can find a reasonable cause to prefer it over other attributes, except for role-playing purposes, scripted interractions and such.

 

I can see an important reason to make crts happen on the Natural 20, and that is to keep combat encounters more controlled. The drawback to this is that major lv differences will not make obvious impact, and that may be a bad thing.

 

I agree that Accuracy is the most important statistic. I don't really look at it from the point of view that I need to gain an extra 50% damage though. To me is critical to surmount the -100% to -50% miss/graze penalty. DR + DT + Miss + Grazing really makes for ineffective damage.

 

Dexterity is a good fit for the Cipher, because they have to divide their actions between two different states. The faster they attack, the more soul energy they gain. The more soul energy they gain, the more spells they can cast. The faster they execute and recover from casting, the faster they may return to leeching soul energy. Considering that any given spell will consume about 25% to 50% of their maximum soul energy, doing all of these actions even 16% faster makes a difference.

 

Dexterity also really helps when using firearms and polearms. I made an Elven Rogue focused in using pikes to flank and interrupt, and it was a terror. It was attacking faster than any enemy, and between its high accuracy & high damage, I would say that greater than 2/3 of her attacks resulted in an interrupt.

[Sensuki]

Here is the actual math if anyone cares:

 

 

Courtesy of Matt516.

 

Accuracy is better all the time when ACC-DEF is between +5 and -5 and below -20. It might be better than Might between -5 and -20 depending on what you have in each score. In BB v333 most of the time, enemy defenses are between the -20 and +5 range by default (without buffs).

Edited November 7, 2014 by Sensuki

[Captain Shrek]

I have a feeling that this has not taken Crit into account. I remember the discussion Mutonator had with Matt. And I now my gut tells me that mutonator was right all along. 

[Sensuki]

You're kidding right? The large spike in the center of the graph represents the portions of ACC-DEF where +1 to ACC removes a miss and adds a crit. The sharp incline of Might on the right side of the graph is because you're critting more often, the might 2% bonus is more effective.

 

Feel free to check the math yourself, his spreadsheet is readily available.

Edited November 7, 2014 by Sensuki

[Captain Shrek]

You're kidding right? The large spike in the center of the graph represents the portions of ACC-DEF where +1 to ACC removes a miss and adds a crit. The sharp incline of Might on the right side of the graph is because you're critting more often, the might 2% bonus is more effective.

 

Feel free to check the math yourself, his spreadsheet is readily available.

 

 

Hmm. Maybe then the DT is the issue? Is that in the calculation? My memory is now hazy as to what was included in the stats. But I had made the argument about Per then, which seems to have justified itself perfectly well empirically. Maybe your theory crafting is incomplete somewhere. 

[Sensuki]

No, DT is not the issue. DT makes attack speed flat out worse than pure damage though.

 

It's just that enemies now have higher defenses than they did in previous versions of the beta. Stone Beetles were given +50 to Defelection (yep) and they now have 63 Deflection instead of 13.

 

So you can see how where you previously might have been at ACC-DEF +30 where Might was better, you're now in the range where Perception is better.

Edited November 7, 2014 by Sensuki

[Captain Shrek]

No, DT is not the issue. DT makes attack speed flat out worse than pure damage though.

 

It's just that enemies now have higher defenses than they did in previous versions of the beta. Stone Beetles were given +50 to Defelection (yep) and they now have 63 Deflection instead of 13.

 

So you can see how where you previously might have been at ACC-DEF +30 where Might was better, you're now in the range where Perception is better.

Well, you must be wrong somewhere as the game is definitely more rewarding towards Per investment than a balance of Per and Might. 

 

Now this may be a perception effect, but as I said, Guts don't agree with you. 

[Sensuki]

I just told you, enemies have higher Deflection so that ACC-DEF falls in the +5 or lower range on the graph. That's why there has been a shift away from Might to Perception.

 

Also in v301 when Attributes changed to the 10 is 0 system, all classes lost 10 Accuracy, whereas Deflection stayed the same, which pushed things further to the left side of the graph.

Edited November 7, 2014 by Sensuki

[archangel979]

His graph has not taken into account 1/day abilities or X/encounter abilities. It is only true for at will attacks. Anything that has limited charges always does better with bigger chance to work.