Balance Discussion

[Sensuki]

I don't think that's right because two stilettos doesn't do that much damage. You try using two stilettos and then using like two maces - huge difference, and the only reason you're winning at all with the Stilettos is the broken interrupts.

 

oh right, not actually in seconds - that's probably why.

 

@roguelike what formulas are you using for that data? can you post your base values and stuff

Edited March 2, 2015 by Sensuki

[roguelike]

Whoops, that blunderbuss value is off.  That's how good blunderbusses's would be if they if they had 4 Damage Penetration, lol.  Pretty good!

Here's an updated chart with some values fixed:

 

 

And here's the raw spreadsheet:

 

 

https://docs.google.com/spreadsheets/d/1e0JNpqfLNpFBDduJEwLxDm-0ix6iNunvJ3C2jeti82U/edit#gid=0

 

 

Formula was double counting penetration values on dual weapons, but even fixed it still says that dual maces and dual stilettos are good.  Ignoring weapon damage ranges and just using an average damage may be making two-handed weapons look worse than they are.  

[ushas]

roguelike, that's really interesting.

 

A few more stuff, I would be also interested to know, if you will consider Sensuki's request:

 

-- How did you estimate Two-Handed, in comparison to others.

 

-- How did you consider selection of armors.

 

Assuming the armors you listed there:

 

Plate  A. DR: Slash: 15.0, Crush: 12.0, Pierce: 15.0
Mail   A. DR: Slash: 13.5, Crush:   4.5, Pierce:   9.0
Scale A. DR: Slash:   8.8, Crush:   7.0, Pierce:   5.3
Lion       DR: Slash:   4.0, Crush:   4.0, Pierce:   4.0

 

As Maces, Stilettos and Estocs have 3 DR bypass, one can simply add three points to their damage range, when considering opponents like these. Maces are Crushing weapons, making them even better against armors like Mail or Plate. But you know that.

 

-- Why is there difference for Arquebus (6 DR bypass) between Scale Armor and Lion...

 

-- What are Deflection and Accuracy values of targets and attackers, respectively. (Also what about hit/graze/crit)

 

-- Whether the stilettos are faster than Maces and Sabres.

 

 

Can I ask what in Interrupt mechanism is making one deal more damage?

[Rafkos]

Is it possible to edit spells' effects? Maybe we should play with values a bit and show it to Devs?

[roguelike]

roguelike, that's really interesting.

 

A few more stuff, I would be also interested to know, if you will consider Sensuki's request:

 

-- How did you estimate Two-Handed, in comparison to others.

 

-- How did you consider selection of armors.

 

Assuming the armors you listed there:

 

Plate  A. DR: Slash: 15.0, Crush: 12.0, Pierce: 15.0

Mail   A. DR: Slash: 13.5, Crush:   4.5, Pierce:   9.0

Scale A. DR: Slash:   8.8, Crush:   7.0, Pierce:   5.3

Lion       DR: Slash:   4.0, Crush:   4.0, Pierce:   4.0

 

As Maces, Stilettos and Estocs have 3 DR bypass, one can simply add three points to their damage range, when considering opponents like these. Maces are Crushing weapons, making them even better against armors like Mail or Plate. But you know that.

 

-- Why is there difference for Arquebus (6 DR bypass) between Scale Armor and Lion...

 

-- What are Deflection and Accuracy values of targets and attackers, respectively. (Also what about hit/graze/crit)

 

-- Whether the stilettos are faster than Maces and Sabres.

 

 

Can I ask what in Interrupt mechanism is making one deal more damage?

 

It's not called a crude chart for nothing.  I'm actually ignoring damage types, the armor values are just for reference and mostly arbitrary.  I think it's easier to see the trend of the curve this way instead of splitting out every single value.  

 

Other things I'm doing are ignoring deflection and accuracy, spitballing accuracy to 1.5% damage per point of accuracy, even though that's quite right, ignoring crits and grazes(though I'm adding grazes to the latest version).  Interestingly, it seems if you compensate for graze damage benefiting from full damage penetration, pistols pull even with crossbows.

 

Two-Handed weapons are just 17 damage weapons that swing at the same speed as one-handed weapons at the moment.  Stilettos are faster than maces and sabres.  Maybe too much faster, I might have messed up the dual weapon attack speeds.

[ushas]

Thanks for answers. It didn't show me your other post where you are providing the table, so my questions were off, sorry for that. 

[Sensuki]

Why is there a column called accuracy bonus in decimals? and where is the times attacking column from

 

Rapier also attacks at the same speed against hatchets and daggers (and other 'fast' melee weapons)

Edited March 2, 2015 by Sensuki

[aeonsim]

So the damage types do make a major difference if you look at the basic Dagger you may think it's equal to the Rapier based on the base damage, ACC and speed.

 

However if you look at the damage types and the types of armour it changes noticeably making the Rapier with it's piercing damage a better option in general compared to the Dagger with it's slashing damage.

 

Padded Armour, Slashing does +1 damage vs Piercing

Leather Armour, Slashing does +1 damage vs Piercing

Chain Armour, Slashing does -5 damage vs Piercing

Scale Armour, Slashing does -4 damage vs Piercing

Breastplate Armour, Slashing does +2 damage vs Piercing

Plate Armour, Slashing damage = Piercing

 

So in it's worse case the Rapier does 2 damage less than a dagger when against a breastplate while in it's best cases it does +5 damage vs Chain.

 

Against creatures the difference is even bigger:

Adra Beetle Rapier +9 damage vs dagger

Spiders +4/+6/+7/+9 damage vs dagger

Stone Beetles +15 damage vs Dagger

 

Now the numbers are simply the difference in Enemy DR when comparing a Dagger (Slashing) vs a Rapier (Piercing) so they're not entirely reflected in reality once minimal damage and damage multipliers are taken into account, but they give you a general idea...

 

VERY Crude modification to show the difference, by simply adjusting the damage by the difference between Piercing and slashing for the listed armour types and setting damage=0 to actually = 0.2*original damage.

Edited March 2, 2015 by aeonsim

[Dr. Hieronymous Alloy]

 

Two-Handed weapons are just 17 damage weapons that swing at the same speed as one-handed weapons at the moment.  Stilettos are faster than maces and sabres.  Maybe too much faster, I might have messed up the dual weapon attack speeds.

 

 

My understanding is that this chart is a fairly accurate depiction of weapon speeds:

 

 

http://i.imgur.com/IvOZuk0.jpg

 

In other words, a "fast" weapon like a rapier or stiletto has a 1 second attack time and a .66 second recovery, while other weapons have a 1 second recovery. Then there's also a "recovery penalty" depending on *how* you're weilding it (two handers get a 1 second recovery). Then armor recovery, etc. 

[roguelike]

Sorry the data is so bad and/or hard to understand.  What I was really hoping for was that someone else had already done all this, maybe with prettier graphs, who could explain that arquebuses actually do way less damage than great swords.

 

My understanding is that this chart is a fairly accurate depiction of weapon speeds:

 

 

http://i.imgur.com/IvOZuk0.jpg

 

In other words, a "fast" weapon like a rapier or stiletto has a 1 second attack time and a .66 second recovery, while other weapons have a 1 second recovery. Then there's also a "recovery penalty" depending on *how* you're weilding it (two handers get a 1 second recovery). Then armor recovery, etc. 

 

Yeah, I started out with testing the ranged weapons to figure out their speed since they're all different, but I should have just pulled the melee weapon speed data instead of trying to figure it out.  

 

Incidentally, why is the speed system so dense?  Plate armor's description says '-50% recovery speed' which going by that table would seem to increase your time to make an attack and recovery by 25% or so, giving you 4 actions in the time you used to have 5.  What it actually does is double your recovery block, increasing your time to make an attack and recover by 50%, giving you 2 actions in the time you used to get 3.  Unless you're using a ranged weapon, which have a reload time.  If the reload animation time equal to the recovery time (is it? no way to tell!) than plate armor with a ranged weapon increases your total time to attack and recover by 33%.  

 

The speed system may be a mechanical success, but it's a nightmare to understand.    Like, I literally have no idea if the +1.5 reload speed part of swift aim or 0.8 reload speed of penetrating shot are working properly or are bugged.

Edited March 2, 2015 by roguelike

[Sensuki]

My understanding is that this chart is a fairly accurate depiction of weapon speeds:

 

http://i.imgur.com/IvOZuk0.jpg

I think that's a bit out of date. The general idea is the same but the specifics have changed. Would be nice to get some new formulas from Josh.

[Dr. Hieronymous Alloy]

 

My understanding is that this chart is a fairly accurate depiction of weapon speeds:

 

http://i.imgur.com/IvOZuk0.jpg

I think that's a bit out of date. The general idea is the same but the specifics have changed. Would be nice to get some new formulas from Josh.

 

 

Yeah, I asked on SA but he didn't respond; that chart also has a few obvious holes like medium weapons and missile weapons and that some spells are labelled "fast" etc.

 

 My guess is that either A) it's changing again in the next patch or B) since we're getting close to release that info's gonna come out in the "strategy guide." 

[Sensuki]

I doubt that information will be in the strategy guide or the manual as those have already been finalized. They will probably include no information about actual weapon speeds and just have very vague information.

[roguelike]

I think I'll let dps lie until I figure out some way to square the predictions with reality.  In the meantime, here is an incredibly difficult to read chart for predicting the damage that talents give you based on the attack roll modifiers (accuracy - deflection).  

 

 

https://docs.google.com/spreadsheets/d/1eRETDzBeyOwmGYKvyFie810HnCIpq7fpBS6tVlYHe_Y/pubchart?oid=1416148723&format=interactive

 

 

Summary of predictions:

1) Dirty fighting is really bad.  On it's own it's worth about +1 accuracy.  Vicious fighting is worth another +1.  Crit damage is just too low for what's effectively +5% crit to matter.  

2)The wood elf racial bonus, which is equivalent to marksman, is far superior to the other bonuses and gives ~5-10% ranged damage and reflex/ranged mitigation, seemingly making them both the best ranged attackers and the best tanks, since tanks are the ones who receive most ranged attacks.

3)Accuracy returns fall off sharply once you have 15 more than opponent's deflection.  Before +15, each point is eliminating a miss and adding a crit, after 15, you're replacing grazes with crits.  So when you pass +15, 1 point of accuracy goes from being worth %1.5 damage, to only worth %0.8.  One-handed weapons may be suffering from this.  It also means that weapon focusing in weapons that give +accuracy is generally going to be less good since you're more likely to pass that +15 threshold.  Note that this threshold tends to disappear against enemies whose DR is effective against grazes but not hits and crits, meaning accuracy beyond +15 continues to be most useful when fighting enemies who's DR is about 1/2 your weapon damage.

4)Lol, one-handed weapon style.

5)Bloody slaughter is a respectable +damage skill if you already have good accuracy. 

[Dr. Hieronymous Alloy]

What I'd really like to see is a chart of the break points for the dps tradeoff between Dex and Might for different weapons.

[Sensuki]

3)Accuracy returns fall off sharply once you have 15 more than opponent's deflection. Before +15, each point is eliminating a miss and adding a crit, after 15, you're replacing grazes with crits. So when you pass +15, 1 point of accuracy goes from being worth %1.5 damage, to only worth %0.8. One-handed weapons may be suffering from this. It also means that weapon focusing in weapons that give +accuracy is generally going to be less good since you're more likely to pass that +15 threshold. Note that this threshold tends to disappear against enemies whose DR is effective against grazes but not hits and crits, meaning accuracy beyond +15 continues to be most useful when fighting enemies who's DR is about 1/2 your weapon damage.

I still chuckle at the accuracy removal from attributes after the damage calculation chance. It appears no maths/logic was applied !

Edited March 4, 2015 by Sensuki

[roguelike]

I made another chart, this time it makes the amazing prediction that fireball is a bad spell.  Don't take fireball.  This has been your public service announcement for the day.  Consider Fan of Flames or Wall of Flame for your flame needs instead.

 

 

 

In fact, you're probably better off just not being a wizard at all unless you really love attacking things with wands.

 

Also Wall of Fire is amazing right now with the accuracy bonus/bug from being a trap.  I kinda hope they don't fix it, it would actually give wizards a decent spell.

 

Does anyone know whether the DR bypass on spells implemented as 0 second DoTs is intentional?  Seems weird that Ionic Projection just randomly ignores DR.

Edited March 5, 2015 by roguelike

[Veevoir]

 

In fact, you're probably better off just not being a wizard at all unless you really love attacking things with wands.

 

This is the true public service announcement in your post. Any chance chart will get improved with spell levels? This is an important distinction, to see if dmg progression from spell level to another is at least consistent.

[ushas]

Also Wall of Fire is amazing right now with the accuracy bonus/bug from being a trap.  I kinda hope they don't fix it, it would actually give wizards a decent spell.

 

 

Well I have a list of issues the trap spells share with normal traps (including accuracy) in queue to report. You don't want me to do that? (but it may use the same mechanism anyway)

 

One may bet you will beg developers to fix it when first spell an enemy uses on your 4 lvl  grouped (eg. by cinematics) party is Searing Seal trap... When it mixes together with other bugs, eg. with the >100 enemy defence stat bug and even other high accuracy bugs, very usefull indeed;) I got wiped like this several times, also due to Criticals of this trap.

 

But anyway, thanks for the chart.

 

 

Any chance chart will get improved with spell levels? This is an important distinction, to see if dmg progression from spell level to another is at least consistent.

 

Does it matter now, when half of these spells may be bugged and subject of change? (just asking, I don't know)

[Veevoir]

@ushas: I did not assume it is *that* bad, thought it is mostly an issue with numbers that need a bit of a tweak at this point..
If most spells are bugged - there is no point

[Luckmann]

I made another chart, this time it makes the amazing prediction that fireball is a bad spell.  Don't take fireball.  This has been your public service announcement for the day.  Consider Fan of Flames or Wall of Flame for your flame needs instead.

 

 

 

In fact, you're probably better off just not being a wizard at all unless you really love attacking things with wands.

 

Also Wall of Fire is amazing right now with the accuracy bonus/bug from being a trap.  I kinda hope they don't fix it, it would actually give wizards a decent spell.

 

Does anyone know whether the DR bypass on spells implemented as 0 second DoTs is intentional?  Seems weird that Ionic Projection just randomly ignores DR.

It should be noted that "right now", Wall of Fire is bugged, and it only tics once, when someone touches it. Then it doesn't do anything anymore.

 

I'm assuming this is not intentional, though.

[ushas]

@ushas: I did not assume it is *that* bad, thought it is mostly an issue with numbers that need a bit of a tweak at this point..

If most spells are bugged - there is no point

 

I don't know how bad it is, that's why asking. Others who reported issues would know more.

 

If the changes will consist of damage and accuracy tweaking, it may change this chart in some way...

It also depends how roguelike estimates the numbers. Besides these are different area AOEs (eg. differ in size and friendly-fire), some should have damage over time, some nondamaging secondary effects, some are combat-only, etc... So it depends what you are comparing it for.

 

 

And what about Rolling Flames? Perhaps it doesn't have stellar damage, but it's providing some interesting tactical options, like scounting ahead by fire

Check how far one can fire and roll:

 

 

[Veevoir]

Holy crap!
This Rolling Flame range is borderline on exploit - if enemies who don't see you wont go looking for whomever sent the rolling fire - you can kill anything outside of visual range... It is also conveniently about the right size for a dungeon corridor
And if the do it is still strong - set up traps, roll a fire, wait for them to stumble upon traps :D

 

Edited March 5, 2015 by Veevoir

[ushas]

It should be noted that "right now", Wall of Fire is bugged, and it only tics once, when someone touches it. Then it doesn't do anything anymore.

 

I'm assuming this is not intentional, though.

 

Yeah. I think it may be in the traps department by mistake. As compare to other trap spells it doesn't disappear after triggering and you cannot use it before combat...

[ushas]

Holy crap!

This Rolling Flame range is borderline on exploit - if enemies who don't see you wont go looking for whomever sent the rolling fire - you can kill anything outside of visual range... It is also conveniently about the right size for a dungeon corridor

And if the do it is still strong - set up traps, roll a fire, wait for them to stumble upon traps :D

 

 

 

They know who did it, across whole map. But it can be useful as you said. Any other ideas?

 

Also with this bug, you can watch their approach

http://forums.obsidian.net/topic/70386-435-fog-of-war-issues/?p=1573357

Edited March 5, 2015 by ushas