Ok, I will fix your stat system for you

[b0rsuk]

I don't like this. It makes it so each class has a different damage stat. I'd rather have universal stats.

 

If anything, I'd do: Might for up close attacks (including spells), Perception for ranged attacks (iuncluding spells). But I'm not so sure about this anyway. I like having rangers who are not "DPS", and can deal hurt in melee when approached.

[Matt516]

Yeeeaaahhhhh.... That doesn't fix it. Good effort, though! ;P

[scrotiemcb]

On second thought, shouldn't have necro'd this. I don't think anyone is responding to what I wrote.

[hyperlord1337]

Hah, such arrogance. This made me happy! ^_^

 

Thanks, first thing I thought

[LuccA]

I'm hijacking OP's thread to post my own stat suggestions:

 

Fitness

• + % melee damage
• + % Health
• + Fortitude
• + Deflection

Coordination

• +% Action Speed
• + Reflex
• + Deflection
• - Chance to be interrupted

Perception

• Notice objects
• +% Ranged damage
• + Reflex
• + Chance to interrupt

Intelligence

• +% AOE size
• +% Duration
• + Will

Resolve

• +% Spell power
• +% Endurance
• + Fortitude
• + Will

I combined Strength and Constitution into Fitness. Thematically, I don't think its worth having them as separate stats.

Also, notice that Fitness gives Health, but Resolve gives Endurance. This gives more control and forces you to trade off between long-term tankiness and short-term tankiness.

 

I've not been following the forums for a while, so I might have lost the earlier discussion about it, but FINALLY I've seen someone suggest a smaller amount of stats. If there's always a dumping stat, why not shorten the amount to 5 or even 4 stats? It would probably solve a lot of balancing issues. 

 

 

"But god forbids cutting stats of! The IE games had 6, this HAS to have 6 as well!" ¬¬

[Zwiebelchen]

I've not been following the forums for a while, so I might have lost the earlier discussion about it, but FINALLY I've seen someone suggest a smaller amount of stats. If there's always a dumping stat, why not shorten the amount to 5 or even 4 stats? It would probably solve a lot of balancing issues. 

 

 

"But god forbids cutting stats of! The IE games had 6, this HAS to have 6 as well!" ¬¬

 

 

This is actually a pretty damn good idea.

[Luckmann]

Cutting down on stats serves no purpose in itself. There is no inherent reason that a 4-Attribute system would be more balanced than a 6-Attribute system just because the current Attribute Bonuses are lopsided.

 

For my sake, there might as well be 8 like in Arcanum.

 

The core of the Attributes is fine - the bonuses needs to be (heavily) adjusted.

[Amentep]

I fixed the stat system too.

Mine involved 256 individual stats, the ability to assign 1,024 points between the stats, and a 70 minute power point presentation entitled "How to determine your new PoE character with this new stat system in 120 easy flow charts" with accompanying 120 flow charts as hand-outs.

[illathid]

Cutting down on stats serves no purpose in itself. There is no inherent reason that a 4-Attribute system would be more balanced than a 6-Attribute system just because the current Attribute Bonuses are lopsided.

 

For my sake, there might as well be 8 like in Arcanum.

 

The core of the Attributes is fine - the bonuses needs to be (heavily) adjusted.

Moreover, it would actually be hard to balance the defense bonuses if you brought it down to 5 stats.

[perilisk]

My crack at it:

 

Mig: Damage, Heal
Con: Fortitude, Recovery
Per: Accuracy, Sight radius
Dex: Reflex, Move/Attack speed
Int: Deflection, Area
Res: Will, Duration

 

Balance the numerical bonuses as needed. Endurance/Health bonuses are gone, Interrupt/Concentration is gone as a special thing. Aside from perception (but accuracy is boss) each has a defensive/support and offensive application.

[scrotiemcb]

On number of stats: as I've said before, the purpose of balance is not to promote sameness, but to promote variety by presenting prayers with a series of separate viable options. Generally speaking, if the choice is between 4 attributes, 6 attributes, or 8 attributes, the most balanced option will probably be a tie between 6 and 8. It would be a tie because it's very unlikely a 7th or 8th attribute around add viable choices for the player, but instead the illusion of choice... but this doesn't make it less variety than 6 choices, merely less elegant. (Elegance is still a great thing in game design, but conceptually separate.) Meanwhile, a 5th attribute is quite likely to give meaningful choice to a player, so 6 is more balanced than 4.

 

People need to stop looking at balance in terms of monotony or simplicity. Balance shouldn't be about either of those things. It is about validating player choice in a sea of differences.

[AndreaColombo]

As some of you probably already know, Sensuki and Matt516 worked on a paper on attribute design during that beta. The system they proposed was mathematically proven to make all six attributes equally viable, and I still advocate that system should be used in the game.

 

In fact, I keep hoping the IE mod will eventually offer it as an option...

[Captain Shrek]

Well, the entire point of attributes is that they should make : a) RP sense and b) have a significant impact on gameplay. PoE attributes do not. It is much harder to tell difference between Mig 10 and 16, while that difference is HUGE in say, D&D. Not saying D&D is perfect. But rather what is right about it when compared to here. 

[scrotiemcb]

As some of you probably already know, Sensuki and Matt516 worked on a paper on attribute design during that beta. The system they proposed was mathematically proven to make all six attributes equally viable, and I still advocate that system should be used in the game.

 

In fact, I keep hoping the IE mod will eventually offer it as an option...

I can't find this paper (the link yields a dropbox error 404, resource not found). Anyone know where I could find it?

 

Mathematically proven is a bold statement.

[illathid]

My crack at it:

 

Mig: Damage, Heal

Con: Fortitude, Recovery

Per: Accuracy, Sight radius

Dex: Reflex, Move/Attack speed

Int: Deflection, Area

Res: Will, Duration

 

Balance the numerical bonuses as needed. Endurance/Health bonuses are gone, Interrupt/Concentration is gone as a special thing. Aside from perception (but accuracy is boss) each has a defensive/support and offensive application.

Adding accuracy back into the system is a mistake. As I've said more than few times, it's too important for any build you can think of. It immediately becomes a must have stat, regardless of what your build is.

[AndreaColombo]

I can't find this paper (the link yields a dropbox error 404, resource not found). Anyone know where I could find it?

 

You may find it here (not sure which one is the latest and greatest.)

[Luckmann]

 

My crack at it:

 

Mig: Damage, Heal

Con: Fortitude, Recovery

Per: Accuracy, Sight radius

Dex: Reflex, Move/Attack speed

Int: Deflection, Area

Res: Will, Duration

 

Balance the numerical bonuses as needed. Endurance/Health bonuses are gone, Interrupt/Concentration is gone as a special thing. Aside from perception (but accuracy is boss) each has a defensive/support and offensive application.

Adding accuracy back into the system is a mistake. As I've said more than few times, it's too important for any build you can think of. It immediately becomes a must have stat, regardless of what your build is.

 

 

And I continue to disagree. Accuracy is valuable, and I'd like to see some graphs on exactly how much Accuracy is worth in relation to, for example, Might. Is +1 Accuracy worth more or less than +3% to Dmg and Healing? How is it quantifiable compared to +AoE% (which I want on Resolve) and +Duration% (which I want to keep on Intellect)? And is the difference enough to matter in a meaningful way?

 

I'm still not convinced.

[scrotiemcb]

 

I can't find this paper (the link yields a dropbox error 404, resource not found). Anyone know where I could find it?

 

You may find it here (not sure which one is the latest and greatest.)

Thanks. Interestingly, my reaction to the paper ties in to Luckman's post...

 

Adding accuracy back into the system is a mistake. As I've said more than few times, it's too important for any build you can think of. It immediately becomes a must have stat, regardless of what your build is.

 

And I continue to disagree. Accuracy is valuable, and I'd like to see some graphs on exactly how much Accuracy is worth in relation to, for example, Might. Is +1 Accuracy worth more or less than +3% to Dmg and Healing? How is it quantifiable compared to +AoE% (which I want on Resolve) and +Duration% (which I want to keep on Intellect)? And is the difference enough to matter in a meaningful way?

I'm still not convinced.

Emphasis mine.

 

This line of thinking, which was so prevalent in the Sesuki/Matt paper that it spawned massive charts, is unfortunately completely wrong. By which I mean it is a bad question.

 

Consider the context of allocating attributes at character creation. 60 points to spread around, if you include race, culture, and reduce everything to 3 first. So under that context, asking "which of these two attributes is best for this single task?" is an utterly worthless question. If that task is the only thing which you care about, you are maxing both. Because you only need about 33 points to do that, and you have 60.

 

Let's imagine, however, that you only had 17 points to allocate towards your single goal. How would you decide to allocate them? The answer is you'd either do a calculation, like Sesuki's fancy charts, or you'd be doing it wrong. Either you'd find some equilibrium point, or if not you'd max one attribute and put the rest in the other. But a calculation isn't a meaningful choice, it's just a math test.

 

The only meaningful choice is what came before: what tasks do you want your character to excel at? Weapon DPS? Tanking (and if so, tanking what?). Healing other party members? Applying debuffs?

 

The thing to understand here is that these different tasks are INCOMPARABLES. You can't put DPS and healing on the same line chart.

 

When players decide how many points to allocate to an attribute, they are doing one of two things: a calculation, or a choice between incomparables. Only the latter represents meaningful choice!

 

So let's go back to Accuracy and Might. Accuracy doesn't change speeds, so at their core both attributes modify the same subcontinent of DPS, damage per action. They're the same thing! So there is no meaningful choice added, just a calculation.

 

Worse, the same thing is true regarding debuffs. Both Intelligence and Accuracy determine the length of the debuff, so once again, two attributes doing the same thing. Except the Accuracy attribute is doing two things. So it probably gets its rate turned down such that, if the two tasks you want to do are damage and debuffs, you can take the Accuracy attribute for a discount because it acts as a kind of hybrid Might/Intelligence. Except you can already go hybrid Might/Intelligence, by putting points in each. Why the discount? Why allow these attributes to feel "overcapped?"

 

Now I feel I should mention the current Dexterity here. Yes, it acts as a second Might for weapon DPS. But it is not "more damage per Rest" for spells or "more damage per Encounter" for per-Encounter abilities (unless encounter length is universally short). Dexterity gives better Interrupts per second, which Might doesn't. Most important, Dexterity has an impact on kiting which no other attribute has, as well as a unique defense against enemy Interrupts. It does have a fair bit of sameness to it - otherwise I wouldn't have brought it up - but (barely) maintains enough differences to make it feel truly different.

 

What we need here are true differences in kind, not differences in scale. We need to put an absolute end to identical traits being on multiple stats, such as Fortitude, Reflex, Will, and most importantly Deflection. (Probably should add Interrupt to that list; if you want to Interrupt more, get gear with bonuses or pump Dexterity!) We need to throw out the charts, because they mean nothing. We need to stop with players comparing apples to apples, and have them compare apples to oranges.

 

And, of course, we should never have an Accuracy- increasing attribute. One to increase damage and one to increase duration. This splits accuracy into two halves which follow the differences-in-kind principle.

Edited April 27, 2015 by scrotiemcb

[illathid]

 

 

 

My crack at it:

 

Mig: Damage, Heal

Con: Fortitude, Recovery

Per: Accuracy, Sight radius

Dex: Reflex, Move/Attack speed

Int: Deflection, Area

Res: Will, Duration

 

Balance the numerical bonuses as needed. Endurance/Health bonuses are gone, Interrupt/Concentration is gone as a special thing. Aside from perception (but accuracy is boss) each has a defensive/support and offensive application.

Adding accuracy back into the system is a mistake. As I've said more than few times, it's too important for any build you can think of. It immediately becomes a must have stat, regardless of what your build is.

And I continue to disagree. Accuracy is valuable, and I'd like to see some graphs on exactly how much Accuracy is worth in relation to, for example, Might. Is +1 Accuracy worth more or less than +3% to Dmg and Healing? How is it quantifiable compared to +AoE% (which I want on Resolve) and +Duration% (which I want to keep on Intellect)? And is the difference enough to matter in a meaningful way?

 

I'm still not convinced.

That's fair, but I think you and I will probably never be able to see eye to eye on this.

[AndreaColombo]

scrotiemcb - I disagree with you on multiple levels
 

I believe this boils down to my approaching attributes from a system design perspective, whereas you're approaching them from a player/character-building perspective.

 

From a system design perspective, attributes are indeed a matter of calculation. You have design goals that you want to achieve (in PoE's case, no dump stats and no pigeonholing) and six attributes to fill with effects that achieve said goals. And no matter how many fancy names you come up with for things attributes do, they all boil down to two things: increasing your DPS, or increasing your durability (by either making you harder to hit, or increasing your Endurance/Health.)

 

In this context, if you don't do your math, or if you do it wrong, you end up with dump stats, must-pump stats, and a system that overall doesn't allow much room for meaningful player choice. The current system in PoE, for example, offers 3 stats that are universally good for all characters (Might, Dexterity, Intellect), 2 are must-pump for tanks and dump for everyone else (Perception, Resolve), and 1 that squishes can dump and others can ignore (Constitution). You can still make meaningful choices with regards to character concept and development (as well as overall roleplaying), but there are only so many ways in which you can make your character optimal/efficient. It's hard to make a completely hopeless character, but very easy to make a sub par one.

 

Matt's and Sensuki's paper propose a system where all stats are equally viable, so if you make a jack-of-all-trades character it will be about as good as a polarized one. That's because investing in one stat gives you about the same amount of benefit you'd get from investing in another. To make an overly simplistic example—if investing in Might makes me kill enemies 10 seconds faster, investing in Resolve will make me last 10 seconds longer while taking punishment in combat (numbers are random and this is just an example to get the point across.) This, imho, is how a good attribute system should be. Players would be free to create their character however they like, safe in the knowledge it will be viable in combat and not particularly worse than any archetypical, optimal build.

[Zwiebelchen]

The current system in PoE, for example, offers 3 stats that are universally good for all characters (Might, Dexterity, Intellect), 2 are must-pump for tanks and dump for everyone else (Perception, Resolve), and 1 that squishes can dump and others can ignore (Constitution).

 

This is summarizes the problem very well... and immediately delivers one question: why, instead of advocating a whole (unrealistic) makeover of stats, don't we try to make the obvious dumps more meaningful?

 

It has often be suggested that Constitution should give a recovery penalty reduction. I'm still all in for that. Heck, this mechanic already exists in the game (as Fighters have a class specific talent that reduces recovery penalty on armors), we only need to friggin' add it to the attribute and be done with it!

This basicly makes the stat generally interesting for many builds and adds a fourth "must have for everyone" attribute.

 

And if you now argue that Dex and recovery are basicly the same, then the game developers already added a nice twist to mix it up: dexterity doesn't change weapon loading speed. This basicly creates a non-linearity in both attributes, which allows both attributes to shine in different situations.

 

 

Which leaves only Per and Res to be "fixed". Basicly, the tools to do that are already there: concentration and interrupt. The problem is: beta days have shown that concentration and interrupt are extremely hard to balance. We had both phases of hilariously overpowered interrupts and meaningless interrupts. Currently, we have a tendency for the latter. Maybe interrupt/concentration is a dead end we should reconsider here; maybe if we add something else to these stats, like:

- extra spellcasting range for Perception (my biggest gripe with PoE spells so far is how extremely short their casting range is; especially for priest buffs). This would automaticly benefit all those classes that don't benefit from resolve: Mages, Druids, Ciphers, whereas those that currently pump resolve would have almost no additional benefits from this (Fighters, monks, Paladins).

- a retaliate mechanic for Resolve (every melee attack against the character has a chance to trigger a free attack; kind of like the fighter talent) ... this also adds an advantage to carrying a least one melee weapon as a caster... currently, there is no real point in having a mage equip for example a rapier or dagger. It's a wasted weapon slot that could have been an Arbequs or Arbalest.

Edited April 28, 2015 by Zwiebelchen

[Luckmann]

 

 

 

My crack at it:

 

Mig: Damage, Heal

Con: Fortitude, Recovery

Per: Accuracy, Sight radius

Dex: Reflex, Move/Attack speed

Int: Deflection, Area

Res: Will, Duration

 

Balance the numerical bonuses as needed. Endurance/Health bonuses are gone, Interrupt/Concentration is gone as a special thing. Aside from perception (but accuracy is boss) each has a defensive/support and offensive application.

Adding accuracy back into the system is a mistake. As I've said more than few times, it's too important for any build you can think of. It immediately becomes a must have stat, regardless of what your build is.

 

And I continue to disagree. Accuracy is valuable, and I'd like to see some graphs on exactly how much Accuracy is worth in relation to, for example, Might. Is +1 Accuracy worth more or less than +3% to Dmg and Healing? How is it quantifiable compared to +AoE% (which I want on Resolve) and +Duration% (which I want to keep on Intellect)? And is the difference enough to matter in a meaningful way?

 

I'm still not convinced.

 

That's fair, but I think you and I will probably never be able to see eye to eye on this.

 

I don't know. I think we just need to crunch some numbers. Accuracy is valuable, but it's not the end-all be-all. Going by game logic in relation to damage alone, Savage Attack adds *1.20 Dmg at the cost of 5 Accuracy. Worth it? Arguably.

 

scrotiemcb - I disagree with you on multiple levels

 

I believe this boils down to my approaching attributes from a system design perspective, whereas you're approaching them from a player/character-building perspective.

 

From a system design perspective, attributes are indeed a matter of calculation. You have design goals that you want to achieve (in PoE's case, no dump stats and no pigeonholing) and six attributes to fill with effects that achieve said goals. And no matter how many fancy names you come up with for things attributes do, they all boil down to two things: increasing your DPS, or increasing your durability (by either making you harder to hit, or increasing your Endurance/Health.)

 

In this context, if you don't do your math, or if you do it wrong, you end up with dump stats, must-pump stats, and a system that overall doesn't allow much room for meaningful player choice. The current system in PoE, for example, offers 3 stats that are universally good for all characters (Might, Dexterity, Intellect), 2 are must-pump for tanks and dump for everyone else (Perception, Resolve), and 1 that squishes can dump and others can ignore (Constitution). You can still make meaningful choices with regards to character concept and development (as well as overall roleplaying), but there are only so many ways in which you can make your character optimal/efficient. It's hard to make a completely hopeless character, but very easy to make a sub par one.

 

Matt's and Sensuki's paper propose a system where all stats are equally viable, so if you make a jack-of-all-trades character it will be about as good as a polarized one. That's because investing in one stat gives you about the same amount of benefit you'd get from investing in another. To make an overly simplistic example—if investing in Might makes me kill enemies 10 seconds faster, investing in Resolve will make me last 10 seconds longer while taking punishment in combat (numbers are random and this is just an example to get the point across.) This, imho, is how a good attribute system should be. Players would be free to create their character however they like, safe in the knowledge it will be viable in combat and not particularly worse than any archetypical, optimal build.

This. A good system is one that allows you to mix and match at will, yet get a viable and interesting character. The mechanical value lies in everything being about as good, and the roleplay value lies in how you decide to play.

 

Practically the opposite of how it is today. You want to tank? You *will* be a charming understudy with the eyes of a falcon. You're a caster that wants to do damage? You *will* be a meathead with a bachelor's degree in advanced astrophysics.

[Lightzy]

I still think my original fix is best

 

 

QFT though, the guy who said that POEs stat system fails on everything that a stat system should do in a game. In POE levels are king, stats are practically nothing.

It's so difficult to judge any difference between might 10 and might 16 in POE, whereas in D&D the difference is ginormous. That's how you can tell it's a good stat system. Because it freakin matters.

[illathid]

I don't know. I think we just need to crunch some numbers. Accuracy is valuable, but it's not the end-all be-all. Going by game logic in relation to damage alone, Savage Attack adds *1.20 Dmg at the cost of 5 Accuracy. Worth it? Arguably.

 

If accuracy only played a role in DPS I'd agree with you, but it doesn't. Crowd control and debuffing also rely on accuracy as well.

[scrotiemcb]

I believe this boils down to my approaching attributes from a system design perspective, whereas you're approaching them from a player/character-building perspective.

 

From a system design perspective, attributes are indeed a matter of calculation.

If you would research systems design, systems psychology would arise as one of the subdisciplines. I cannot possibly imagine a serious argument that psychology should be abandoned in a system for a GAME, in favor of raw math. Players are not (usually) robots.

And no matter how many fancy names you come up with for things attributes do, they all boil down to two things: increasing your DPS, or increasing your durability (by either making you harder to hit, or increasing your Endurance/Health.)

This argument is a lot like saying an object is nothing more than protons, neutrons and electons. I'd consider it a false statement, not because I deny these three things to be the essential building blocks of matter, but because the arrangement of those blocks is key to the identity of a thing. A paperweight on a desk isn't *just* particles, but particles arranged in a particular manner.

 

In the same way, saying everything is DPS or survivability is a gross oversimplification. By this view, an AoE with a duration which debuffs accuracy of anyone in it is just survivability.

In this context, if you don't do your math, or if you do it wrong, you end up with dump stats, must-pump stats, and a system that overall doesn't allow much room for meaningful player choice.

This isn't wholly untrue.

 

I'll admit, I got a little overzealous in my previous post. Throwing out Sesuki's work was going a bit too far. I was just trying to be dramatic.

 

However, this is the important distinction: those charts were player activities, not (primarily) designer activities. They are still relevant to the designer, because those charts express a player psychology. But the core goal is to bring balance to attributes so none is a dump stat. The core of doing this will always be what you call the "character building perspective," because you need to get into the minds of gamers approaching the attribute system from various INFORMED angles and successfully encouraging them not to dump certain attributes.

The current system in PoE, for example, offers 3 stats that are universally good for all characters (Might, Dexterity, Intellect), 2 are must-pump for tanks and dump for everyone else (Perception, Resolve), and 1 that squishes can dump and others can ignore (Constitution). You can still make meaningful choices with regards to character concept and development (as well as overall roleplaying), but there are only so many ways in which you can make your character optimal/efficient. It's hard to make a completely hopeless character, but very easy to make a sub par one.

What I interesting here is how you easily identify psychological profiles early in this quote, but do not seem to identify "min-maxer" as a psychological profile itself.

Matt's and Sensuki's paper propose a system where all stats are equally viable, so if you make a jack-of-all-trades character it will be about as good as a polarized one. That's because investing in one stat gives you about the same amount of benefit you'd get from investing in another. To make an overly simplistic example—if investing in Might makes me kill enemies 10 seconds faster, investing in Resolve will make me last 10 seconds longer while taking punishment in combat (numbers are random and this is just an example to get the point across.) This, imho, is how a good attribute system should be. Players would be free to create their character however they like, safe in the knowledge it will be viable in combat and not particularly worse than any archetypical, optimal build.

Emphasis mine.

 

Here's an example of that idea taken to its ultimate extreme:

Might: +3% damage

Con: +3% damage

Dex: +3% damage

Per: +3% damage

Int: +3% damage

Res: +3% damage

 

Is this "perfectly balanced?" I would argue it has zero balance, because it isn't balancing anything, really; it is just xeroxing. Copypasta isn't balance.

 

And no attribute should copypasta anything another attribute does. Fortitude shouldn't be on both Might and Con. Reflex shouldn't be on both Ref and Per. Will shouldn't be on both Int and Res. Deflection shouldn't be on both Per and Res. There should be only one.

 

Among the not-copypasta stuff, players will *try* to go all Grand Unification Theory on a game, making charts and such. As a designer, the objective is to prevent such theories from being true. If someone is trying to reduce a combination of attributes to just DPS and survivability, you must stop them with situational superiority/inferiority scenarios.

 

I do not mean balance is just a matter of the designers ignoring things. That's silly. I mean they should make things appear a near tie, with pros and cons for each, but with an unknowable victor. It's anything but easy.

 

I'll try to reply to others later

Edited April 28, 2015 by scrotiemcb