a simple heuristic to compare the relative value of ressources

[clouseau64]

a 100 hours into the beta, and even after the patch, I remain pretty annoyed by the gross misbalancing of ability and ressource. I know it's only closed beta, but this is the famed sequel of a game which managed to reach great after its many patches  and feedback from community, so our level of expectation is much higher.

 

i feel that high priority shall be given in the next patch in fixing the gross misbalancing of some activty equipment and weapon modals so that we can actually test them  ( spells with a decent cast time wouldn't hurt either as it looks like we all shun them).

 

Here is quick heuristic focused on DPS ( and it's reverse survivability), comparing the impact of various ressource and their trade off. This is only approximative as the time element is not taken into account at this stage, but it shall already help in spotting the main incoherences and imbalances. Feel free to complete / comment. i put in bracket if the impact is multiplcative or additive

 

 

 1 penetration  (x)   ~  25% dps   ~  8 might (x)   ~  12 perception  (+/x)  ~  12 accuracy (+/x)  ~  8 dexterity (x) 

 

                               ~  single weapon wielding (+/x)   ~  31 % of dualweapon wielding dps (x) ~ 4 power level for spells(x) ~ 8 class level if single class

 

                               ~ 12 class level (x)  if multiclass ~ 4 levels (+/x) ~ 0.5 x backstabb (+) ~ 0.62 x brutal whip (+)

 

                               ~ 0.5 x 1 weapon quality level (+/x) ( acc,damage +pen) ~ 2 druid class level for claws ( 1 quality level / 4lvl)

 

                              ~ 1 power level for monk fist (+/x) so 2 or 3 monk levels ~ 80% of potd  difficulty ( +/x)

 

                              ~ -20% dps through penetration modals for longbows, sabres, warhammers, stilettos (x)

 

                             ~ 0.5 x marked for the hunt/ sworn ennemy (+/x) ~ 0.3 x lightning fist (x) ~ 0.25 x FOD (+/x)

 

 

 

 1 armor (x)            ~  25% survivability   ~  5 constitution (x)   ~ 12 deflection (+/x)  ~ 12 resolve (+/x)  ~  3 x weapon and shield ( +/x)

 

                              ~  -7% dps (x)  ( through common armor)  ~ - 4 % dps (x) if dual wielding  ~  - 22% dps (x) through medium shield modal

 

                              ~ -60% dps (x) dagger modal ~ 33% of paladin defense with deep faith (+/x) ~ -22% dps (+/x) through quarterstaff modal

 

                              ~ 80% of potd difficulty  (+/x)  ~ 1 level (x, +/x) between level 1-5 ( hitpoint +def) ~ 2 levels  (x, +/x) between level 6-20

 

                              ~ 4 shapeshifting druid level (x) ~ 12 x 1 shield quality level (+/x)   ~ 1 armor quality level (x)

 

                              ~ 2 x weapon and shield (+/x) warrior  ability  ~ 0.4 x mirror image spell  (+/x) ~ 0.3 berserker rage (x)

 

 

 

it becomes pretty clear, that armor shall be stacked to the max ( plate for everyone) rather than go for any of the other mitigant and that everybody shall dual wield. wearing a shield is particlarly bad unless you are an unbroken but why bother, be a paladin you get 9 times more delection and the same armor with the aura and can dual wield.

 

it makes  very little sense to try to limit the stacking of attribute buff, accuracy or deflection when the smallest unit of armor or deflection represent such a major gap in survivability even after the patch. we can see that one unit of armor/pen almost offsets potd difficulty and it is going to be extremely difficult to balance the concept with levels increasing.  5 pens is equivalent to the gap between level 1 and 20. It's  even worse when one considers that pen doesn't go below 5 nor armor below 6 ( wearing no armor is a death warrant). 

 

Armor and pen need at least to be discretized further with increments that are of a magnitude 2  levels or 12.5 dps at the very least, so armor and penetration need to be doubled and on a much larger scale. it is also necessary to allow player to equip weapons that are in the bag during the fight otherwise it's forcing save scumming when armor have finally been uncovered.

 

Personally, i find that the damage system is over determined with this new pen/armor system and that poe1 system was much better. for high armor/low pen the system worked just fine, it was for late game high damage that it somewhat broke. The reason it didn't scale late game was notablity because weapon quality increased at faster rate. if armor had scaled with level of the character by a multiplcative factor or that constition or resolve had applied a multiplcative factor that armor would have scaled just fine.

 

If dev absolutely want to stick to that new system, then i suggest that armor qualiy level shall have a stackable bonus to deflection on top of the armor system, and that shields shall at the very least add to armor. higher quality of shields shall also have increments in armor and deflection ( +4 not +1 per quality level. 

 

 

 

Power is a new currency that looks totally necessary given the introduction of multiclassing. It look pretty obvious that penetration of spell shall at the very least increase twice as rapidly by power level ( having to wait 12 level if multiclassed to do one more pen is ludicrous), i would suggest 1 penetration per power level if penetration is more discretized or 0.5 pen per level in the current system). i also find that classes bonus are very front loaded with very large bonuses at inception or early tiers. it would make more sense to have those toned down and then scale with power or with level.

 

 

Modals are just catastrophically balanced whether the defensive one which are all more punitive than armor in dps, or the penetrative one which in the best case scenario ( ennemy armor is 2 points higher than current pen) increase dps by 10%!!!!. one is much better off having accuracy and wait for a crit.

 

 

 

The heuristics doesn't take into account time so probably overvalues constitution  ( as healing is not in this picture) , defense and dps through recovery.

it looks to me that either hitpoints need to be massively boosted or base damage toned down as combat is hardly lasting  more than 5 to 6 rounds ( assume a round is 3s). In poe1 armor always reduced based damage and armor piercing modals only added back 5 damage, not a whopping 25 to 75% of multiplied damage, so damage look much bigger in poe2. We need combat to last much longer and not by overinflating recovery.

 

 

 

 

 

[theBalthazar]

This explain the nerf of devoted and berserker and even with that ! there is  the "suppressed effect" with weapons. (case of berserker, read lower)

 

So, at the end, the system can break because Obsidian want to limit the gain of penetration, but nerf again and again all the subclasses with a boost in penetration.

 

So eventually, at the end, you have the penalty and not real gain.

 

Like now. Berserker have a bigger penalty than devoted. But, with the nerf all the two classes have Penetration + 2.

 

BUT Half-sword, a modal of sword, give a +2 in penetration. If the logic is respected : stack. Obligation, because if not, you broke the interrest of berserker (bigger malus and base fighter have 50 % hit to crit, without penalty). Yet this is not currently the case, and it is not normal. The non-stack is a solution of ease, illogical.

 

So Obsidian MUST create a violent and logical metagame. If not, the game will be less fun, or the bonuses still have few nerfs (at the risk of breaking the logic)

 

Like you said, one of the solution is to increase HP of ennemies.

 

But Obsidian will not do that. Why ? Because Obsidian want to create dynamic fights AND not that novices get lost in endless battles where 5 resources would not be enough. So it's limited as an option. (When you are novice, you don't know the best options like optimal way of play).

 

Indeed : if you have "no pen" "no pen", more HP is worse ! This moves the problem. Even, make it worse.

 

Obsidian must balance his game, but this is limited : the penetration, as you have shown, is a fairly restricted space from level 1 to 20. So there is too few solutions, except accept a meta-game quite unbalanced, they will try to balance at best.

 

In PoTD, I'm afraid we'll see more Ranger / Berserker / Fighter. Penetration is a key for do all your damage and +30 % damage. It is not nothing...

Edited January 7, 2018 by theBalthazar

[AndreaColombo]

Subbing thread.

[Dr. Hieronymous Alloy]

I really like the idea of opening up the penetration scale with more increments. 

 

I'm a little sceptical of a precise point for point breakdown because these values are all curved; going from 14 to 15 pen is very rarely going to do anything, going from 5 to 6 is usually going to do a lot.

 

Dual Wielding currently has a 50% modifier from the first game; it was increased to that from an 20% or 30% original modifier due to the vagaries of the AR system; needs to be downtuned back and then we can see where the numbers boil out.

[IndiraLightfoot]

What I love about this thread, is the overall perspective on the combat system and its ambition to compare it with the mature iterations of PoE1.

 

We know this:

-The PoE1 system worked well, it had its own CRPG signature, including some of its flaws

-With Deadfire, Josh & Co seem to wish to make the game sleeker and more in line with other similar CRPGs. Thus, 50% of the combat system is currently altered.

-The system is not at all balanced right now.

 

It's like that saying: "If we only could be strangers again." Obsidian is trying to make a new and fresh system, but we have trouble forgetting our long relationship with the old and tried one. I applaud their bravery, but it may turn out to be a mistake, especially as I fear there are lots of nerfs incoming, when I'd rather want to see a stern focus on Deadfire offering as fun and engaging and varied combat experiences as possible.

Edited January 7, 2018 by IndiraLightfoot

[Madscientist]

I really aggree to the things you two said.

And I like the damage system of PoE1 much more than the one of PoE2.

 

PoE2 is all or nothing.

If you understand the system, you max out penetration, armor and strengh and obliterate everything with dual wielding.

Until you understand this ( and some things are hard to understand ), enemies might destroy you just because you have 1 point pen or AR too little.

 

I also aggree that the modals are terrible.

They give +100% recovery (almost twice the value of heavy armor) or they give a huge penalty of -20 to acc or deflection.

 

 

by the way: dual wielding was also very powerful in PoE1. Eder (dual wielding sabres) or Hiravias (shapeshifted) do much more damage than my paladin with a 2h weapon.

 

What if we change the system of PoE1 so it is more like tyranny or Dragon Age Origins:

Like PoE1, damage done =  weapon damage - ( armor - penetration ). Each weapon has a different damage and penetration, like a sword has high damage and low pen while a hammer has the opposite. In this system you need to chose your weapon regarding your pen and enemy armor like in PoE2, but it was not all or nothing.

 

Were the different attack types more balanced in NWN2 ( dual, 2h or 1h + shield)

For dual wielding, there you had to spend talents to learn it well, you have only half the strengh bonus and -2 hit chance for the off hand and you need a small weapon in the off hand to avoid another penalty. There were some very powerful dual wielders ( e.g. kaze no kama), but you had to invest a lot for it. Just putting a weapon in each hand without having any talents would make you much weaker while in PoE2 it makes you much stronger.

2H weapons got a 1.5 strengh modifier while a shield gave you the bonus of ( big surprize ) being harder to hit.

Edited January 7, 2018 by Madscientist

[IndiraLightfoot]

 

PoE2 is all or nothing.

If you understand the system, you max out penetration, armor and strengh and obliterate everything with dual wielding.

Until you understand this ( and some things are hard to understand ), enemies might destroy you just because you have 1 point pen or AR too little.

I also agree that the modals are terrible. They give +100% recovery (almost twice the value of heavy armor) or they give a huge penalty of -20 to acc or deflection.

 

 

 

Absolutely true. This is right now the biggest flaw with Deadfire: it's all or nothing. In PoE1, there I could build a thousand varied and surprisingly nuanced builds. It had like a greater number of sliding scales that I could adjust and jog.

Edited January 7, 2018 by IndiraLightfoot

[theBalthazar]

I really like the idea of opening up the penetration scale with more increments. 

 

 

change absolutely all the statistics of the bestiary is generally a bad idea.

 

Because if you change amplitude of penetration. You must adjust armor in return.

 

It could be a solution, if it's done soon enough.

 

But here an other problem : and that why Obsidian make this choice : Obsidian don't want a non-viable system for starters. Because if you go on a system at 5 - 20 penetration rather than 5 - 14, you take a risk for novice players. With a bad build you have 5 of penetration rather than a highter Max (here 20).

 

Same problem with old game with high growth rate.

 

Obsidian create POE2 to decide faster if the current strategy will work. All... or nothing.

 

I also aggree that the modals are terrible.

They give +100% recovery (almost twice the value of heavy armor) or they give a huge penalty of -20 to acc or deflection.

 

 

That is true. + 100% recovery, we don't want to take it.... -20 deflection is terrible but... it's the least worst : p

Edited January 7, 2018 by theBalthazar

[Dr. Hieronymous Alloy]

The only way the modals make sense is if, much like casting times, they decided to start out with them being absolute **** deliberately because it's easier to balance up than balance down.

[dunehunter]

The heuristics doesn't take into account time so probably overvalues constitution  ( as healing is not in this picture)

 

Yep, if you take Second Wind or other heal spell into account, the actual hpmax is probably doubled or more. So the Constituition is overvalued for sure.

 

In poe1 armor always reduced based damage and armor piercing modals only added back 5 damage, not a whopping 25 to 75% of multiplied damage, so damage look much bigger in poe2. We need combat to last much longer and not by overinflating recovery.

 

I like this last summary best, there are much more multipliers in the game than we realized, and this will cause an overinflation eventually.

 

Looking at how DOS 2 performs in late game, everything is inflated too much in late game and I really don't like it.

Edited January 7, 2018 by dunehunter

[Dr. Hieronymous Alloy]

I like the instincts in this approach but the more I think about it the more nervous I am about carrying this analysis too far; you can't just trade AR for deflection or constitution, etc.; these things aren't perfectly fungible.

 

So one approach to balacing these features would be to just set it so there are very few or no opportunities to trade anything else for AR / Penetration. The problem of course is that since this game has "money" there's only so much limitation like that you can do, since gear gives ar/pen and can be bought and sold.

[JFutral]

I like the instincts in this approach but the more I think about it the more nervous I am about carrying this analysis too far; you can't just trade AR for deflection or constitution, etc.; these things aren't perfectly fungible.

Bonus points for using the word 'fungible' correctly!

 

Carry on.

 

Joe

[clouseau64]

I like the instincts in this approach but the more I think about it the more nervous I am about carrying this analysis too far; you can't just trade AR for deflection or constitution, etc.; these things aren't perfectly fungible.

 

So one approach to balacing these features would be to just set it so there are very few or no opportunities to trade anything else for AR / Penetration. The problem of course is that since this game has "money" there's only so much limitation like that you can do, since gear gives ar/pen and can be bought and sold.

 

Absolutely agree.

My main point was to show that armor/pen in these kind of increments overule the rest and are tough to scale.

My second is that blatant misbalancing can be easily avoided just by using these basic comparisons so that we can focus on testing the fine balance of it.

 

Back to pen/armor,  what i meant by the system is overdetermined looks like it's the same currency as base damage. they have just circonvoluted the system in two but there is no real trade off.

Take the great sword for instance, it does 25% percent more damage than weapon that have +2 pen. so basically we are saying that against high armor we do 25% less, equal on same armor and 25% more the rest of the time. Really all that hassle for that?

System like that can work for wasteland 2, because the linear path controls the tiering of weapons vs armor and you don't have abilities that drastically change the balance. for poe, i don't think so.

 

I also dislike the new health system, because with the healing and lack of endurance/ health  pool there is much less the feeling of a good difference between accuracy/ deflection and damage/health. there is no penalty for being hit often especially now that interupt has been overhauled. I hope that in future patch perception will stop being debased and brought to a pure dps function, it leads to stereotyping the builds, at least when it comes to weapon dps.  unfortunately weapon dps looks like the only thing that we test extensively given that elemental spells or debuff/CC have been brought to oblivion.

[FreeKaner]

It seems to me Obsidian in an attempt to reduce the impact of some minor imbalance of DR bypass completely destroyed all of the balance in the game. What sort of game is this where ideal for everyone is to wear heaviest armour and wield two weapons? Which we can see comes directly from removal of DR and instead switching to a system that's essentially two layers of abstracted percentage conflict (I.E defence rolls then armour rating). It inevitably will devolve into either extremity (I.E either highest armour and highest attack speed or vice versa) due lack of asymmetric reaction to action speed and flat values.

 

Instead of this invalidation of a system that was approaching a great standard in PoE1, why didn't they diversify the layers of defence and offence between rolls, flat reductions, percentage reductions and bypasses? A system similar to and is expanded from both PoE1 and New Vegas? A system that could increase both depth and intuitiveness of the game, instead of creating layers and layers of abstraction of ratings and classes on top of each other. Addition of New Vegas style percentage reduction defence consumables, as well as including penetration that was not a flat value but percentage in bypass is both intuitive and strengthening of each layer individually in conjunction.

 

Thoroughly disappointing how the extremely promising class system is being pushed into munchkin min-maxing of extremities of the system, no balance to speak of.

Edited January 8, 2018 by FreeKaner

[IndiraLightfoot]

You may very well be right, FreeKaner. And yeah, when I backed Deadfire, then seeing multi-class was in, I assumed it would be like PoE1, but with fantastic multi-classing, hence, even more nuanced and varied building of characters. I never asked for Health/Endurance to get removed, the resting and the attributes to get changed, or having a Pen vs Armor as a new pivot, and so forth. I was in general very happy with what PoE1 had as its signature system. This may be the right time to at least consider a toggle under settings called "PoE1 mode", where all the rules and systems from that game can be used for those of us who want to play the game in its classic classy glory.

 

At least, I'd love to try the Deadfire beta with late PoE1 in full effect.

[IndiraLightfoot]

Perhaps the highest praise PoE1 has received is something that's been repeatedly posted in these very beta subforums:

-You could finish PoTD solo with a character with like 10 on each attribute, if you went the sneaky route

-Almost any kind of builds could make a PotD party and finish the game, no problems

 

It would be a shame if Deadfire turned out to be yet another of those games where you have to have meta knowledge of those three munchkin super builds in order to take on the most difficult modes.

 

And even if PoE1 was slightly unintuitive and complex for newbies, the game did adapt beautifully, letting most builds on normal mode pass the game with flying colours. When I returned to the game, with the release of WM pt 2, I accidentally had that huge battle for my keep when my party was lvl 5, and the gear I had sucked, still, after a few tries, I managed to defeat an entire army! I loved that!!

No real need then to remove health/endurance, grazes, and half the system, no? And is Deadfire as it stands now really much less esoteric and intricate to beginners? CRPGs always have some sort of adaption time/threshold to get over (mileage may vary).

Edited January 9, 2018 by IndiraLightfoot

[Boeroer]

[...] And even if PoE1 was slightly unintuitive and complex for newbies [...]

This is true. 75% of the game mechanics were obscure and the descriptions either inaccurate, not clear enough or non-existent. But basically it's not like the mechanics are supercomplex.. It's just that PoE does a very poor job of explaining its mechanics.

 

If they just could have done this better then there wouldn't have been so many complaints a la "I don't get how XY works". I mean look at the PoE forum (char builds subforum especially). Even today there will be a new thread that discovers how some hidden mechanic works and everbody (even the hardcore players, including me) will go "whooot? Didn't know that!". That's a pretty good example of bad documentation. But so easily solved... so easily... 

Edited January 9, 2018 by Boeroer

[IndiraLightfoot]

Great point, Boeroer! In essence: Instead of remaking half a system, they could put all that effort into fine-tuning the latest PoE1 system, make a cool tutorial area (like NWN2 OC, for instance), and then have great info, tooltips and in-game wiki with mechanics etc, explained in a clear-cut way.

[Boeroer]

That would have been my preferred approach. But there are also some really nice things in the beta. I guess a mix of the new features (that I like) and the old ones would have been great. But that's just me.

 

The beta still feels a bit "tyranny-ish" to me. I didn't like Tyranny too much (only played it for 300 hours or so ).

Edited January 9, 2018 by Boeroer

[AndreaColombo]

I agree there was no need for a big overhaul, but some of the things they’re doing with Deadfire actually have merit. We need more internal consistency and a lot more UI completeness and clarity of information for a start, and a few basic balance changes as highlighted in this thread as well as KDubya’s weapon analysis.

[theBalthazar]

the only thing I would have wanted from  Tyranny is : spell modification. 

 

I saw it less strongly than Tyranny. But for example 2 modifiers on the spells.

 

In this situation, you can take 9 seconds. And allow a -30 % recovery on a "sigil".

 

Two slots : (No more, for avoid the chewing gum effect, and stay with a strong personnality of the spell)

-30 % recovery

-30 % cast time

 

And others choices. Like bounce +2, add burn effect. etc. This ideas of modification was amazing in Tyranny. The most apreciable part for me.

 

Plus, with that, you create a precious treasure = sigil. So the system works.

 

And with higher personality spells of pillars = win.

Edited January 9, 2018 by theBalthazar

[MaxQuest]

PoE2 is all or nothing.

If you understand the system, you max out penetration, armor and strengh and obliterate everything with dual wielding.

Until you understand this ( and some things are hard to understand ), enemies might destroy you just because you have 1 point pen or AR too little.

Perfectly matches my experience with the Beta so far.

> Create a beguiler/fury and travel with the 4 predefined mercenaries. Get completely crushed by anything that has a "lagufaeth" in it, despite of intensive micro.

> Create 3 wayfarer/berserkers. Max PEN. Max AR. Dual-wield sabres. 19/10/18/3/18/10. Watch them destroy anything on an auto-pilot AI behavior, in less time a druid would conjure a Blight.

 

1 penetration (x) ~ 25% dps ~ 8 might (x) ~ 12 perception (+/x) ~ 12 accuracy (+/x) ~ 8 dexterity (x)

It's kinda hard to estimate PEN (and AR), because the gain from +1 penetration keeps changing along the scale.

It would be much easier if it was gradual. But Josh is adamant about it being not: link.

The thing is it makes following estimations and conclusions a bit fuzzy.

 

Also would like to add a few thoughts/corrections:

- 12 perception is greater than 12 accuracy, since there is also 24 reflex.

- 5 constitution may not necessary have the same impact as 8 dexterity or 8 might; as the latter help finishing the current opponent faster and switch to another, creating 2x1 local overwhelming; plus in case of might less limited resources will be spent. And if there is a high amount of incoming healing, investing in constitution becomes a little moot.

But that's nitpicking) as I do agree with your conclusions about armor, dual-wielding and shield/deflection.

 

---

Btw, when thinking about such heuristics vs Josh's approach, I tend to think of:

- balancing a wood plank on a line (example), vs

- balancing a wood plank on a point (example)

 

In the first case it's easier to come up with a balanced state, and even if you shift some value, overall balance won't get altered if the values are tied by common heuristics.

 

In the second case it's harder to balance stuff, but you can achieve some exotic state where unexpectedly everything works fine enough.

Tbh I had this feeling when estimating (additive) MIG vs (multiplicative) DEX in PoE1. Unlike DEX, MIG was useful only for damagers and healers; and additionally it's effect was getting dilluted by other damage coefficients (like crits and weapon quality), but it still was on-par with DEX, because of the flat DR in general AND DR of the enemies we met in particular.

 

Personally I like having some heuristics / association formulas. They make the whole system which is being constructed more robust. And only when the construction is almost finished / has a solid fundament, one can add one or two point-balanced features like an icing on the cake. (without going: oh we can't change x this way, because it will unbalance y, which will unbalance z, etc) (and the icing may be required such that the system won't feel too rigid, dull or boring)

 

The beta still feels a bit "tyranny-ish" to me. I didn't like Tyranny too much (only played it for 300 hours or so ).

Speaking of game systems, neither did I. I mean I liked the story, the implemented ramifications, the idea of spell creation. But completely disliked the attack resolution and balance. It was too easy to break, i.e. come with a single best build, check it in solo PotD, and at that point replay'ability from powerbuilding point of view was gone, because trying anything else would be just gimping yourself. And one more thing: playing with a good build in Tyranny felt unfair for me, as if the systems were not thought/play-tested well enough and you are just exploiting them; while comming with a good build in PoE feels way more rewarding and without that guilt feeling. Edited January 9, 2018 by MaxQuest

[Madscientist]

Actually I liked Tyranny.

 

There were some things that I did not like:

- Learning by doing leveling system

- Cooldowns as the only resource. Once you have some abilities you can spam them all the time. Until then you have to fill the time between abilities with auto attacks.

- Your main char can learn everything without restrictions. This leads to completely OP combinations. The Divinity games have the same problem. PoE1+2 is limited by the class system.

 

In a perfect game, I would like:

- The class system of PoE2

- The resource system of PoE2 martial classes (you have a pool of resources and each ability uses this pool)

- The damage system of PoE1 ( maybe a few changes, but the general system of PoE1 is better than PoE2)

- maybe we combine the resource system of PoE2 with spell crafting of tyranny? Like a mage has magic as resource. A fireball costs 1 magic, a faster or stronger fireball costs 2 magic and a faster and stronger fireball costs 3 magic. You have only 6 magic per encounter.

- game mechanics should be more consistent and better explained to the player.

 

In general I like PoE2 a lot, with the big exception of the penetration/armor system which I really do not like.

[Boeroer]

Actually I liked Tyranny.

Ok, you lost me there...

[the_dog_days]

I liked Tyranny but I didn't like its combat systems.