Concerns with the proposed Perception change

[Zekram Bogg]

After looking at both Luckmann's post here and his original huge post of balance change stuff from April, I have to say I agree with pretty much everything this man's saying, and if this were a council meeting, would second his motions and suggest the rest of the board adopt them in toto.

 

However this is not such a council meeting, alas. I'll have to settle for signing up for the Luckmann Newsletter of PoE Griping.

 

I think the idea that constitution could add an armor penalty modifier is great actually. Since Might is now relegated to the idea of "nebulous power of the will" rather than sheer burst muscle capability as Strength would be in most RPGs, it really leaves Constitution as the only attribute that conceptually governs physical strength. Considering CON kind of sucks at representing this idea currently, it really does need more of an oomph, and having high Con characters being able to move more freely in armor fits rather well.

 

That said, if we're looking at alternate modifiers to the game that could be added to various Attribute scores (not just Con), here are some things that aren't covered by the current attribute system to consider (some were covered in earlier versions of the game).

 

- Range of attack modifiers - I for one am down with increasing or decreasing the range due to base stats. Perception makes the most sense, then Intelligence, but both of these Attributes are already pretty beefy and useful, especially if PER is going to start affecting Accuracy again.

 

- Movement Speed Modifiers - I'm surprised that there are no modifiers on in battle movement speed to be honest. It would make a lot of sense if a high Dex character could just move faster than a low Dex character. Even if it were a +/- 0.2 Move Speed Modifier per point, that'd mean your 15 Dex character would have a noticeable speed increase over an average Dex character at 10, and could outpace them in a chase scenario. Which will be very important I'd think once Individual stealth is in the game. 

 

- Disengagement Defense Modifiers - Another possibility for CON buffing instead of Armor Penalty, what if a high CON character just got a much higher Disengagement Defense bonus? Like a +/- 1 point per CON. So you could gain an advantage for CON by being able to move around a battle a bit more freely? This would probably also need to be buffed by another Attribute, like Resolve too, so a High CON, HIGH Resolve character could potentially move around a battlefield a lot more freely than the average, and dumping both stats would mean even picking Talents or Gear that raise Disengagement defense would only bring them up to par.

 

- Range or actual Fog of War clearing - I'm a fan of games where each Stat/Attribute has an OBVIOUS in-game effect so you can know that altering it works in some concrete way. I'm rather surprised that no one is considering the size of the player's Fog of War Clear bubble as an effect that could be altered by PER or INT as well.

 

All that said, the big question I'm having right now is this . . . if PER is going to affect Accuracy again in 2.0, potentially at a 1 point per PER rate, then is MGT going to affect Interrupt as proposed in Luckmann's Attribute System of Solidarity (heretofore referred to as LASS)?

 

It seems like it's going to kind of have to, because if PER is going to get the Accuracy buff, then that Interrupt buff has to go somewhere else lest PER become way too strong a stat (as many in this thread are theorizing it will be).

[evilcat]

Adding more ideas:

Maybe instead of Accuracy bonus just +1% crit chance (if that alone is not enought then +0,01 crit multiplier mod as well)

 

I think it will be less significant. With Accuracy, everyone wants accuraccy. Casters or CC to. And it will change the current balance of defense.

With crit chance this will just increase dmg in game which could be balanced with constitucion giving 5% HP per point.

 

For autoattackers Perception is anyway more useful than resolve. For tanks having opportunity to applly on crit effects may be interesting.

From the other side +1Accuracy also makes sense. Since this will allow classes with inherit low tohit have better chances for some autoattacing. In that case flat +1 seems to be better. That would make Perception like must have stat for casters... but that would be more tradeoff betwean Might/Inteligence/Perception/Resolve/Dex. (Still dump con thou)

Maybe that would lead to distincion: For Blaster Caster Might + Dex, for CC Percepction+Inteligence... Flat Acc bonus is not that bad, at worse it will give +10 Acc but... for the cost of something else.

 

But it seems to be more important to buff Constitucion since this is a dump stat.

It could be either changing +5%. INT have similar mod and nobody complains.

Or make it +0,5 Endurance per level per point. So it will benefit/hurt squishes more.

From the other side % is PoE way of class bonuses. And 5% is always more than 0,5. High endurance is a advantage of some classes so making this inherit bonus less important is nerfing them. I dont like situacion "why to tank with Paladin, when i can do just fine with wizzard?" So overall +5% Endurance is better.

 

Maybe there is a space for talent: "Toughtness: Get +2 endurance per leve" (or flat 10 + 1/level). So squishes could be less squishy, while big guys will not need it and can spend talnet on something else.

Edited August 5, 2015 by evilcat

[AndreaColombo]

Con is already going to give 5% per point.

[Nobear]

Con is already going to give 5% per point.

 

Link please.

[MunoValente]

- Movement Speed Modifiers - I'm surprised that there are no modifiers on in battle movement speed to be honest. It would make a lot of sense if a high Dex character could just move faster than a low Dex character. Even if it were a +/- 0.2 Move Speed Modifier per point, that'd mean your 15 Dex character would have a noticeable speed increase over an average Dex character at 10, and could outpace them in a chase scenario. Which will be very important I'd think once Individual stealth is in the game. 

 

- Disengagement Defense Modifiers - Another possibility for CON buffing instead of Armor Penalty, what if a high CON character just got a much higher Disengagement Defense bonus? Like a +/- 1 point per CON. So you could gain an advantage for CON by being able to move around a battle a bit more freely? This would probably also need to be buffed by another Attribute, like Resolve too, so a High CON, HIGH Resolve character could potentially move around a battlefield a lot more freely than the average, and dumping both stats would mean even picking Talents or Gear that raise Disengagement defense would only bring them up to par.

 

- Range or actual Fog of War clearing - I'm a fan of games where each Stat/Attribute has an OBVIOUS in-game effect so you can know that altering it works in some concrete way. I'm rather surprised that no one is considering the size of the player's Fog of War Clear bubble as an effect that could be altered by PER or INT as well.

 

 

These are things that I think could be linked to skills.  Athletics could be used for movement speed or disengagement.  Stealth could be used make engagement more difficult, Survival could provide bonuses related to fog of war or spotting enemies.

[Nobear]

I would certainly not be opposed to something interesting to make Survival more useful.

 

I think Con granting 5% (I'd still like to see that link ) is really all it needs. It's already somewhat useful for tanking (especially soloing), moreso for the Fortitude than the endurance boost, but 5% would bring the endurance effect to meaningful enough values without going too far.

[AndreaColombo]

I think Con granting 5% (I'd still like to see that link )

 

Link

[Nobear]

 

I think Con granting 5% (I'd still like to see that link )

 

Link

 

 

Thanks, and wow, retraining is officially coming! Thank you Obsidian!! Looking forward to seeing how it's implemented. My top choice would be to have it be very costly so it's not something you could do willy-nilly, but serve as a system that makes a moderate degree of RP sense and will allow for (still costly) respecs in future expansions too.

[evilcat]

Thanks, and wow, retraining is officially coming! Thank you Obsidian!! Looking forward to seeing how it's implemented. My top choice would be to have it be very costly so it's not something you could do willy-nilly, but serve as a system that makes a moderate degree of RP sense and will allow for (still costly) respecs in future expansions too.

 

Or maybe not. Sure some retrain cost is ok. Everyting costs. But why make it prohibiting?

It is one feature which is very easy to ignore. And in SP game we can;t blame people how to play and shame them for having fun.

[limaxophobiacq]

I can't see how getting 20 perception for +10 accuracy on CC is going to break to game.

 

+10 accuracy only matters one out of 5 castings anyway, you could very well cast ALL your 6th level spells in a row and have the 10 per vs 20 per not matter.

 

Sacrificing everything for that last little bit of accuracy on a spell just isn't worth it the way it is for stacking deflection, stacking deflection matters because that last bit is what brings your chance of getting hit down from 10% to 0%. For Accuracy the equivalent would be the bit of accuracy that get's you just high enough that you have a chance to get a hit in at all but for CC hit vs. graze isn't all or nothing the way it is for damaging attacks vs. Deflection & DR as you still get half the duration which is oftentimes enough.

 

Only way it would really matter is if you have a hard time hitting even a graze with your CC, and with Wood Elf bonus + Eldritch Aim that is very unlikely and if you're going up against defenses that far over your accuracy you're probably screwed even with the perception bonus (as again it only matters one out of 5 castings) and you are just hoping to get lucky.

 

Generally you are probably better of getting 20 dex to just cast more CC faster.

Edited August 6, 2015 by limaxophobiacq

[jsaving]

 

It is one feature which is very easy to ignore. And in SP game we can;t blame people how to play and shame them for having fun.

 

It's one thing to bar people from respeccing every 30 seconds in dungeons so they'll always be optimized against the foes in the next room.  But it's quite another to say, we're going to periodically alter the fundamental rules under which you developed your character build, but because it would break immersion to let you respec, you need to either stick with your build or else start over.

 

The folks at Wizards of the Coast had a couple of articles about how to handle periodic rules updates that might adversely impact people's builds.  The gist of what they had to say was that every time we make a rules change, we are altering your character sheet without your permission, possibly in a way that takes your character away from whatever vision you might have had for that character.  And when you decide that a respec best enables you to realize that vision -- or that you need to try a fundamentally new vision because your old one just can't be done under the new rules -- then it is incumbent upon DMs to let those respecs happen rather than making players stick with the same-on-paper-but-drastically-different-in-spirit character sheet they were using before the rules changes were made. 

 

I think they're right and am happy respeccing will soon be available for a modest fee.  But I do hope respecs will only be available in certain game locations, so you have to leave your current dungeon in order to change your build.  Is this how it works in the beta patch, or can you respec on-the-fly wherever you happen to be?

Edited August 6, 2015 by jsaving

[View619]

It looks like you can only respec at inns.

Edited August 6, 2015 by View619

[gkathellar]

+10 accuracy only matters one out of 5 castings anyway, you could very well cast ALL your 6th level spells in a row and have the 10 per vs 20 per not matter.

I'm no expert, but I'm pretty sure that's not how probabilities work.

 

Perhaps more importantly, the endgame is not when these adjustments have the most impact. A +10 boost is never anything to sneeze at, but by level 11-12, it's dwarfed by other bonuses from level and special effects. Throughout the early and mid game, however, there are two things to consider. First, the boost is relatively more important. You may not consider this important, but the notion that only the endgame matters is a troubling one.

 

Second, the outcomes accuracy affects are arguably more important at low levels. At level 11, grazing with Slicken during a boss fight is an inconvenience; at level 4, it costs you precious resources and can even cost you the fight. And IMO, the fact that grazing with CC is less of a pain later on is itself the result of the game's dramatic difficulty drop at high levels - tougher fights, like the Adra Dragon and some of the bounties, make you feel your Grazes again.

 

(Third: In solo, at least some Perception will be mandatory. Accuracy matters too much in the solo game.)

 

I also feel it's important to remember that it's not only high Perception scores that are improved, but also low ones that are greatly worsened. You can get through the game without an extra 5-10 accuracy, but you can't get through it with a -7 accuracy penalty. Dumping Perception is basically just a non-option at this point, no matter the build. Maybe that doesn't matter in your evaluation, but it's worth remembering.

 

I don't think the accuracy bonus to perception is a strictly bad idea. But let's not pretend that casters - whose limited resources make Accuracy crucially important - don't get way more mileage out of it the change than warriors, while also losing much less.

[limaxophobiacq]

I'm not saying perception giving accuracy isn't good, I'm just saying it's not gamebreaking. Compared to the short time it takes to cast Eldritch Aim or the 'sacrifice' of playing a wood elf rather than another race the price you have to pay for increasing accuracy by raising perception (dumping another attribute) is much bigger.

 

Personally I feel Eldritch Aim could probably take being lowered to a +10 and between it and Alacrity (which could probably also take a slight nerf)  elven wizards would still be the late-game CC kings.

 

 

 

+10 accuracy only matters one out of 5 castings anyway, you could very well cast ALL your 6th level spells in a row and have the 10 per vs 20 per not matter.

I'm no expert, but I'm pretty sure that's not how probabilities work.

 

 

I was mistaken about attack resolution, I thought it always was 50 points between graze to hit to crit but turns out graze to hit is 35 points and hit to crit is 50 so its at best a 30% chance per casting to matter rather than 20%. At worst if you are so low that you can only miss or graze with no chance of hitting its a 10% chance to matter per casting, same with if you are so high you can only hit or crit.

 

(There seems to also be a rule that a raw roll of 96-100 is always at worst a hit and 0-5 always at best a graze but when that rule comes up accuracy doesn't really matter at all.)

Edited August 7, 2015 by limaxophobiacq

[Fardragon]

The thing with the idea of "all stats are important to all classes" is it really needs random rolled stats in order to work. Or at the least weighted point buy. Otherwise, even if perfect balance was achieved, it would simply mean that there would be no point in changing your stats around at all.

 

Without randomisation, it should be possible to allocate your stats to either (a) reflect the way you want to play your character. Want to play a crowd control specialist? Do your stats this way; want to focus on single target ranged damage: go this way; want to be a healer: this is your best option; and so on.

 

Or (b) reflect role playing choices. Want to play a big strong person: focus on this stat; want to be smart: put your points into this; etc. Which puts us back into the DnD territory of stat preference being tied to your class choice.

[limaxophobiacq]

 

Without randomisation, it should be possible to allocate your stats to either (a) reflect the way you want to play your character. Want to play a crowd control specialist? Do your stats this way; want to focus on single target ranged damage: go this way; want to be a healer: this is your best option; and so on.

 

But that is what you can do RIGHT NOW, 'Important to all classes' doesn't mean the same as important for every application. Might f.ex. while good for every class doesn't do anything for non-damage/healing abilities so if you want to focus on that might can be deprioritised in favor of perception if you want to do lets say a CC wizard.

 

Same with perception if you're going to play a priest who only heals and buffs since none of your casts are going to go up against a defence.

 

Want to focus on singe-target damage? Then prioritise might/dex/per over Int.

Edited August 7, 2015 by limaxophobiacq

[Fardragon]

 

 

Without randomisation, it should be possible to allocate your stats to either (a) reflect the way you want to play your character. Want to play a crowd control specialist? Do your stats this way; want to focus on single target ranged damage: go this way; want to be a healer: this is your best option; and so on.

 

But that is what you can do RIGHT NOW, 'Important to all classes' doesn't mean the same as important for every application. Might f.ex. while good for every class doesn't do anything for non-damage/healing abilities so if you want to focus on that might can be deprioritised in favor of perception if you want to do lets say a CC wizard.

 

Same with perception if you're going to play a priest who only heals and buffs since none of your casts are going to go up against a defence.

 

Want to focus on singe-target damage? Then prioritise might/dex/per over Int.

 

Only to a very limited degree. Any DPS (ranged/melee/single target/aoe) needs Might and Perception equally. Healers still need Might; Crowd Control and Tanks still needs Perception. It really hasn't been thought through.

[evilcat]

 

 

Without randomisation, it should be possible to allocate your stats to either (a) reflect the way you want to play your character. Want to play a crowd control specialist? Do your stats this way; want to focus on single target ranged damage: go this way; want to be a healer: this is your best option; and so on.

 

But that is what you can do RIGHT NOW, 'Important to all classes' doesn't mean the same as important for every application. Might f.ex. while good for every class doesn't do anything for non-damage/healing abilities so if you want to focus on that might can be deprioritised in favor of perception if you want to do lets say a CC wizard.

 

Same with perception if you're going to play a priest who only heals and buffs since none of your casts are going to go up against a defence.

 

Want to focus on singe-target damage? Then prioritise might/dex/per over Int.

 

 But why? What for? Is it worth the effort?

 

For me: So far perception is not important outside tanks. Interupts are single target, and are generally balanced by resolve of target. For CC purpose there are better aoe picks. Ok it is fun to have 2WF with stilletos and lock down target but it is not needed since best CC is death.

Increasing Acc will make it more interesting. And opens some synergy with all this abilities "trade Acc for..." If it was from the release i doubt anyone will notice the difference or suspect it is not best options ever. 

 

What i really miss is that 0-10 deflection is taken from the mechanic, so everyone will be easier to hit.

[anameforobsidian]

The stats already strongly influence the way your class is played, and this will be even better because it makes crit builds viable.  I think people may be missing the fact that stats synergize with each other.  D&D has conditioned people to think about maxing one stat and then playing with the rest.  In PE, builds depend on what degree you emphasize two to three stats, not maxing one.

 

A perception / dex / might character rogue is going to specialize in getting crits and could easily favor lighter weapons while being very fragile, or use bows to attack from a distance.  A might / res / per rogue is going to be more of a brawler, using something like an estoc to get nasty hits.  Note that only one stat has changed out, but the way they've been emphasized has completely changed the profile of the character.

[limaxophobiacq]

For clarification, when I said 'right now' I was refering to perception in 2.0 beta with accuracy rather than deflection.

[limaxophobiacq]

 

 

 

Without randomisation, it should be possible to allocate your stats to either (a) reflect the way you want to play your character. Want to play a crowd control specialist? Do your stats this way; want to focus on single target ranged damage: go this way; want to be a healer: this is your best option; and so on.

 

But that is what you can do RIGHT NOW, 'Important to all classes' doesn't mean the same as important for every application. Might f.ex. while good for every class doesn't do anything for non-damage/healing abilities so if you want to focus on that might can be deprioritised in favor of perception if you want to do lets say a CC wizard.

 

Same with perception if you're going to play a priest who only heals and buffs since none of your casts are going to go up against a defence.

 

Want to focus on singe-target damage? Then prioritise might/dex/per over Int.

 

Only to a very limited degree. Any DPS (ranged/melee/single target/aoe) needs Might and Perception equally. Healers still need Might; Crowd Control and Tanks still needs Perception. It really hasn't been thought through.

 

 

Actually unless your attacks have very large % bonuses to damage the way rogues do, Might is simply way more effective for damage than Perception (or dexterity) is (perception/dexterity being more effective with large % damage bonuses is a result of how damage calculation is done with % bonuses being addative rather than multiplicative and thus making less of a relative difference the bigger a bonus you already have).

 

If you are going damage over time and AoE, int is more effective than might and FAR more effective than perception.

 

I just don't understand what you want. First you say that this is how you think it should be:

 

 

 

Want to play a crowd control specialist? Do your stats this way; want to focus on single target ranged damage: go this way; want to be a healer: this is your best option; and so on.

 

 

 

But this is literally how the system is right now, but you then complain that that is what it's like, unless what you really mean is that there is no way to bump ranged vs melee damage or accuracy but that would be completely pointless because you'd just bump the one you use and dump the one you dont and you'd not lose anything.

 

Other than that lets go through those examples one by one:

 

 

 

crowd control specialist

 

 

Max int, secondary Perception, third dexterity.

 

 

 

 

want to focus on single target ranged damage

 

 

Forget int (unless your class has some really good singe-target damage over time ability), might first (unless you're a rogue), then perception and dexterity.

 

 

 

 

want to be a healer:

 

 

Forget perception, you're not doing anything that your target is trying to oppose so accuracy is useless, put points in Int first since healing spells have AoE and Duration, followed by might and then third dexterity.

Edited August 8, 2015 by limaxophobiacq

[Fardragon]

How I think it should be? Any of these:

 

a) Random roll, no reallocation or re-rolling, or;

 

b) stats that describe your character in role playing terms. You want to play a big strong guy, max Might and play a Fighter. You want to be clever? Max in and play a mage. I.e. DnD system, or;

 

c) Stats should describe your role. You want to play a tank? Max out Tanking. Want to be ranged DPS? Max out Ranged Combat. Some MMOs do this. PoE ditches any role playing meaning to stats anyway, so they might as well be called by what they do.

[Nobear]

c) Stats should describe your role. You want to play a tank? Max out Tanking. Want to be ranged DPS? Max out Ranged Combat. Some MMOs do this. PoE ditches any role playing meaning to stats anyway, so they might as well be called by what they do.

 

Yeah, no thanks on that for me. I'd feel like my hand was being held and my intelligence was being insulted. It'd seem like a false choice, and possibly be a false choice. It'd be like one of those tutorials for a cel phone game for kids, where it says "You're a bunny, give yourself carrots!" and you press on the carrots and it says "Good job! Now press here, also exactly where I tell you!"

 

Yes, things like Might affecting spell power are totally not obvious to the beginner and mean this game has a fairly steep learning curve, but I'd still rather have that than have cookie-cutter stats where one and only one stat affects each role and is the obvious no-brainer choice for it.

[KDubya]

This and the constitution change together might make build choices tougher by making there be more wants than there are available points.

 

Currently a caster is going to max might, intellect, dexterity, dump con and leave res and per wherever. In 2.0 you'd want to max per, might, intellect and dexterity but for that you'd have to dump both con and resolve. At -5% per con point a 3 con will have much lower hit points than the current -3% per con. Couple this with higher accuracy for everyone and lower deflection for you and combat will be riskier due to being more glass-like.

 

Max perception will be the norm, hitting or missing is more important than anything else. At the early part of the game having +8 accuracy is like being three or four levels higher. It will be interesting to see how things are balanced around this. If it is balanced for 0 additional accuracy then having +8 will tear things up while if it assumes that you will have more accuracy, having base or even less will make it a whiff fest.

 

A two hander fighter can still max might, per and resolve while dumping int but won't dump con into dex anymore or at least not as much. The defender nerf might make it a poor choice for anyone but a punching bag so you'd replace with confident aim for more damage and savage attack for even more.

 

Barbarians will have it rough as they will want might, dex, int, per and res while not losing too much con. They will not have near enough points.

 

Last year I went to Gunsite for a week for pistols and they asked the simple question of "How fast do you need to miss in order to win a gunfight?" Accuracy is king

[limaxophobiacq]

 

Max perception will be the norm, hitting or missing is more important than anything else. 

 

https://forums.obsidian.net/topic/67761-dps-vs-accuracy-deflection-heres-the-maths-enjoy/

 

This was from back in the beta when Might was 2% damage per point and Dex was +1 accuracy per point. As you can see, esp. now at 3% damage per point of Might, Might will always win big (like at least twice as efficient) over Perception for damage abilities unless you are getting massive damage bonuses from non-might sources.

 

Perception should only be prioritised over might if you want the interrupts or if you're focusing on CC.