Sore's Stat Testing Guide

++ INTRO ++
I'm not really an expert in this area. I spend extremely little time actually testing abilities. I would spend more time doing so but life and time management does not permit me to. Others have more time to burn and of them, some test rigorously. I thought I'd take the time to bring the math and practice to the masses to see who it inspires to add to the testing effort. This isn't some art or practice that you want to corner the market on. The more people submitting information, the better. In many cases, the people that have the time simply don't know the math and the practice. I encourage others to come and give me advice as well on how to best test stats so I can incorporate the ideas here. I think of this as a community resource.

Power Descriptions, Tooltips, and Current Effects
To get started, begin by reading the description of the power you intend to test. It's common to overlook something stated there that challenges your existing understanding of that power. Maybe there's a power interaction you didn't know. Read the tooltips for all the relevants stats. After casting the power, check your current effects (Hit up on D-Pad and scroll to the bottom). If testing with a buddy, have him check his current effects after you cast a power on him. With that said, sometimes descriptions, tooltips, and current effects have been known to be wrong or misleading. In the end, let your testing be your guide. For example, it's been observed that if a tooltip shows a base range being applied then the highest value of that range is actually never achieved. So let descriptions, tooltips, and effects simply serve to guide what to test.

Combat Log Testing
One of the most effective methods of capturing data is switch your chat to the combat log. The challenge is that the log often scrolls through and overwrites older entries too fast. Capturing this information is best done by recording your game activity with a PVR and then replaying the results, gathering what went into the log a few frames at a time. It is best to attempt to execute your test more than once. You don't want to have to pull the video session just to realize your test didn't go as planned and have to go through the whole process again. PC players have the added benefit that their combat log is stored on their file system and can be parsed using text editors.
Depending on what you're testing, you might consider only having damage in or damage out enabled in your chat settings.

++ DAMAGE ++
Might:
4 points = 1.8%
1.8%/4 = 0.45% per point
0.45% is expressed mathematically as .0045 for multiplication purposes
(that saves us from dividing by 100 later)

> Applying Might
DAMAGE * (1+MIGHT*.0045)

> Reversing Might
DAMAGE / (1+MIGHT*.0045)

> Slop Intercept Equation (Y = MX + B)
B (Y-Intercept) = BASE DAMAGE
M (Slope) = B*.0045
X = MIGHT
Y = DAMAGE

This is interesting because the slope for lower base numbers is not as steep as the slope for higher base numbers. This provides the effect that the ranges of possible values will widen between low and high results of the same attack as your might increases.

Defense:
71 points = 1% mitigation
1/71% per point
1/71% is expressed mathematically as 1/7100 for multiplication purposes
NOTE: That's for level 30 and it equates to about DEFENSE x 0.000141 when the actual scalar used is DEFENSE x 0.000142. Similarly at level 1, DEFENSE / 1500 equates 0.000667 when the actual scalar used is DEFENSE x 0.000674.

Toughness:
The mitigation for PvP is TOUGHNESS x 0.000132. It's less mitigation than PvE.

It is assumed right now that training dummies have 1065 defense. It seems they mitigate about 15% of your incoming damage. The best way to test this is to run numbers on a buddy for which you know his TOUGHNESS. Reverse his damage mitigation to build base stat numbers you trust. Then run tests on training dummies to see how much those numbers have been mitigated. Keep adjusting your assumed training dummy DEFENSE mitigation until you arrive at numbers that align with the base numbers you worked out with your buddy. Forward calculating those base numbers with mitigation applied should then match what you saw.

> Applying Defense
DAMAGE * (1-DEFENSE x 0.000142)

> Reversing Defense
DAMAGE / (1-DEFENSE x 0.000142)

Modifier:
35%, 45%, 50%, or 65% depending on the power
People show this as a percent so you tend to have to divide by 100 in the forumula
> Applying a modifier
DAMAGE * (1+MODIFIER/100)

> Reversing a modifier
DAMAGE / (1+MODIFIER/100)
(It is common to test without modifiers applied so you rarely need to reverse a modifier)

Critical Hit:
Critical hit magnitude increased by 4% per skillpoint
Default critical hit magnitude is 0%
Critical hit = Normal Hit x (1.25 + CRITMAGNITUDE)

> Applying a crit
DAMAGE x (1.25 + CRIT MAGNITUDE)

> Reversing a crit
DAMAGE / (1.25 + CRIT MAGNITUDE)
(It is common to throw out crits in damage testing so you often do not need to reverse a crit)

Chaining It Together:
> Finding Base Numbers
BASE DAMAGE = DAMAGE > REVERSE DEFENSE > REVERSE MODIFIER > REVERSE MIGHT
BASE DAMAGE = DAMAGE / (1-DEFENSE*0.000142) / (1+MODIFIER/100) / (1+MIGHT*.0045)

Note 1: If you didn't have the modifier in collecting data, you don't need to reverse it. This can get confusing in spreadsheets that show a modifier as an input for applying damage, but don't expect that same modifier to be reversed since damage numbers were collected without a modifier.

BASE DAMAGE = DAMAGE > REVERSE DEFENSE > REVERSE MIGHT
BASE DAMAGE = DAMAGE / (1-DEFENSE*0.000142) / (1+MIGHT*.0045)

> Projecting Damage
DAMAGE = BASE DAMAGE > APPLY MIGHT > APPLY MODIFIER > APPLY CRIT > APPLY DEFENSE
DAMAGE = BASE DAMAGE * (1+MIGHT*.0045) * (1 + MODIFIER/100) * (1.25 +CRITMAGNITUDE) * (1-DEFENSE*0.000142)

Note 2: Round down after each element you apply. Most calculations leave this off which is fair for simplicity but can inflate values for long running ticks of low base numbers. This facet is much more relevant in vitalization calculations for POT and weapon regen.

DAMAGE = BASE DAMAGE > ROUND DOWN > APPLY MIGHT > ROUND DOWN > APPLY MODIFIER > ROUND DOWN > APPLY CRIT > ROUND DOWN > APPLY DEFENSE > ROUND DOWN

Damage Split
The issue with Damage Splits is that training dummies are singular and don't give you a chance to test damage splits. Not all AoE attacks split after the same number of attacks. To adequately test this, you need to have a buddy that can pull adds together for you. It's better a second person do this so that you don't have to worry about that move setting up power interactions or modifiers that add more complexity to what you need to reverse in your calculations. Since each calculation rounds down, the more that you reverse means the more you're becoming inaccurate.

Critical Chance
The default value for critical strike chance is 5%. This is actually two values, critical weapon chance and critical ability chance (might damage). Skill points will increase both values at the same time. Socket affinities can increase those values independantly. My model for projecting critical chance is to take the same ability and simulate it occuring 100 times. Assume that you will recieve a critical strike a number of times equal to your critical chance value. Similarly to reverse critical chance, execute 100 attacks. The number of critical strikes in that series is your critical chance. Of course randomization makes that inaccurate. That means you need to repeat the test more than once. Take the average of your results. It should become more accurate with the more tests you run. It may suffice to run a smaller sample.

ORIGINAL POST BY TUNSO:
https://forums.station.sony.com/dcuo/index.php?threads/webcast-highlights.117416/page-5#post-1229430

Spreadsheet Building
Here is a spreadsheet I've used to project values. In this case I'm comparing Might damage to Precision damage and trying to evaluate whether straight build vs. mixed builds vs. affinity builds vs. pure min/max builds make more sense. It is not perfect. Finding a reasonable base might value to compare to a base precision value is next to impossible. It is then up to the person using the spreadsheet to make their own assessment on what percent of their damage comes from might vs. precision.
http://files.enjin.com/208137/Files...20Modeling.xlsx

Here's another spreadsheet for straight assessing might damage without taking criticals into consideration. This is the type of calculations most people run to get a feel for their results.
http://files.enjin.com/208137/Files/Might.xlsx

Here is another spreadsheet useful for collecting numbers for determining base values. It is based off Electricity values. It was lent to me by DrunkBiker. I believe IBrutalLegend originally put it together. I made some tweaks to how the calculations work. It is used with no gear or skillpoints (542 might) ran on training dummies.
http://files.enjin.com/208137/Files/Electricity Average Base Damage.xlsx

++ Power ++

Vitalization
10 vitalization = 1% power generation increase
1 vitalization = 0.1% power generation increase

> Applying Vitalization
Regular Power = POWER * (1 + VIT / 1000) )

> Base Numbers
Testing for base numbers is not necessary as they are widely published. All POT moves initiate ticks of 70 base power. All power transfers distribute 100 base power to the three most needy players in the group.

> Short Math
1200 VIT = 1200/10 = +120%
100% + 120% = 220%
220% = x2.2
Base Power * 2.2

Critical Power
> Applying Critical Power
Regular Power = POWER x (1.25 + CRITICALMAGNITUDE)

Weapon Regen
Power is generated through achieving combo tiers. Other than tier 4, power is not generated for each hit.
T1 (no label) : 9 ticks of 10 power
T2 (Good!) : 9 ticks of 12 power
T3 (Incredible!) : 9 ticks of 15 power
T4 (Superhuman!) : 20 power per hit
> Testing Weapon-based Power Regen
You can achieve a combo tier and then check your current effects. It will be an effect called Recovery. Sometimes weapon effects are also called Recovery so you'll have to differentiate that. Again remember that when they specify a range, the highest number is never actually achieved. Another option is to check the Damage Out combat channel as it will indicate that Recovery is healing you for so much supercharge and so much power. This is hard to test at higher tiers on training dummies as they reset your combo meter as they give you power. Testing on mobs are difficult because your damage out spams the combat log. Recording the log with a PvR is a good option.

Spreadsheet Building
This spreadsheet can be used for projecting power out numbers based on different vit.
http://files.enjin.com/208137/Files...talization.xlsx

This spreadsheet can be used for projecting power out for assessing the effects of critical chance and magnitude.
http://files.enjin.com/208137/Files/Vit Modeling.xlsx

++ Healing ++

Healer Bonus
If in healer stance, all healing is boosted by 65%.

> Applying the Bonus
BOOSTED HEALING = HEALING * 1.65

> Reversing the Bonus
HEALING = BOOSTED HEALING / 1.65

Restoration
4 points of restoration increase healing by 1%
1 point of restoration increases healing by .25%

> Dominance
Each point in dominance acts like an additional point of restoration

> Applying Restoration
HEALING = BASE * (1 + (RESTORATION + DOMINANCE) * .0025)

> Reversing Restoration
BASE = HEALING / (1 + (RESTORATION + DOMINANCE) * .0025)

Critical Healing Magnitude
> Applying Criticals
CRITHEALING = HEALING * (1.25 + CRIT MAGNITUDE)

> Reversing Criticals
HEALING = CRIT HEALING / (1.25 * (1 + CRIT MAGNITUDE))

Chaining It Together
> Calculating Healing
HEALING = BASE > APPLY HEALER BONUS > APPLY RESTORATION > APPLY CRIT

> Reversing Healing
BASE = HEALING > REVERSE CRIT > REVERSE RESTORATION > REVERSE HEALER BONUS

Spreadsheet Building
This spreadsheet can be used for projecting healing out values based on different restoration.
http://files.enjin.com/208137/Files...estoration.xlsx

This spreadsheet can be used for projecting healing out for assessing the effects of critical chance and magnitude.
http://files.enjin.com/208137/Files/Resto Modeling.xlsx

++ Tanking ++
I will likely need the most help with this section. I've never played a tank and don't have a tank character I can test with. Plus some of the fundamentals are different between Earth, Fire, and Ice.

Defense:
71 points = 1% mitigation
1/71% per point
1/71% is expressed mathematically as 1/7100 for multiplication purposes
NOTE: That's for level 30 and it equates to about DEFENSE x 0.000141 when the actual scalar used is DEFENSE x 0.000142. Similarly at level 1, DEFENSE / 1500 equates 0.000667 when the actual scalar used is DEFENSE x 0.000674.

Toughness:
The mitigation for PvP is TOUGHNESS x 0.000132. It's less mitigation than PvE.

The best way to test this is to run numbers on a buddy for which you know his TOUGHNESS. Knowing the base damage numbers, apply how much damage you expect to land. Take the proportion of what should land and what does land to evaluate the mitigation. That proportion should lead you to his mitigation percent. Reverse that mitigation to what his TOUGHNESS is.

> Applying Defense
DAMAGE * (1-DEFENSE*0.000142)

> Reversing Defense
DAMAGE / (1-DEFENSE*0.000142)

> Spreadsheet Building
Here I'm trying to model how many hits you can take before death of the same constant damage. Then I evaluate whether increasing your defense/toughness or increasing your health will have a greater effect. We historically know the answer to that. It seems it takes roughly 10000 health before any scenarios emerge where more defense/toughness is preferred over health.
http://files.enjin.com/208137/Files...sToughness.xlsx

Fire Health Buff:
Health x 1.6 + Dom x 2.2 = Total health after using a power

Fire Self Healing:
Healing = Dom + Resto
> Spreadsheet Building
http://files.enjin.com/208137/Files/DOM vs RESTO using Burnout.xlsx

++ Precision Damage ++

Short Version
All damage calculations are basically linear progressions. That means the standard slope-intercept formula applies.
Y = SLOPE * X + YIntercept
Y=MX+B

Might Damage (for comparison)
Because of the might calculation, slope is actually as follows:
Y=M*X + B
Y = damage, X = might, B = base damage, M = the scaling factor

M = .0045*B
Y = (.0045*B)*X + B
Y = B + .0045*B*X
Y = B*(1+.0045*X) <-- what we're used to seeing
This is why talking about might, we're interested in base numbers. We're interested in the Y-Intercept because the Slope is actually derived from that.

Precision Damage
Damage-per-second implies that if you have 0 DPS, you do zero damage. That means there is no base. The Y-Intercept is zero.
Y = damage, X = your DPS stat, B = base damage, M = the scaling factor (I call it ATK Percent)

B = 0
Y = M*X + 0
Y = M*X
Damage = ATK Percent * DPS

That means each of your weapon attacks do not have a low or high base value. They have a low or high slope, i.e. they have a low or high scaling factor. If we wanted to do research on weapon damage, we wouldn't search for or record base numbers. We just need to record the ATK Percent. To do that, all we need is two samplings of damage with different DPS.

Y1 = damage sampling with lower DPS
X1 = your lower DPS stat

Y2 = damage sampling with higher DPS
X2 = your higher DPS stat

M = (Y2 - Y1)/(X2 - X1)

Now you want to compare apples to apples. In your test you want to compare your lowest damage to your lowest damage and your highest damage to your highest damage. That's because each weapon attack has a range of possible ATK Percents it may choose between it's high and low. That means the ATK Percent is different between low and high. That means the higher your DPS is, the more possible damage result numbers emerge. That's why it is better to compare results with two relatively low DPS stats, e.g. 20 DPS and 40 DPS. You will be able to find the HIGH and LOW results in a smaller sampling.

The ATK Percent you calculate from your findings might be slightly off because DCUO rounds down. After getting that value, try to alter it to something that a DEV was likely to pick. Instead of 24.85%, try the ATK Percent as 25%. Project values forward and see if you would end up getting the same results. Another hint that your sampling doesn't have accurate numbers is if you try to derive your Y-Intercept and it doesn't turn out to be zero. That's a hint you probably need to toy with your ATK Percent slightly.

Spreadsheet Building
This spreadsheet provides tools for capturing precision-based (weapon-based) damage as well as projecting precision-based damage.
http://files.enjin.com/208137/Files/Precision.xlsx

Damage Builds

Might Builds

Might for all affinities

• Might: 230 SP + 238 Mods + 41 Affinities = +509
• Precision: 14 Affinity + 30 SP = +44
• Weapon Chance: 13 SP + 3 Iconic + 5 Affinity = +21%
• Ability Chance: 13 SP + 3 Iconic + 5 Affinity = +21%
• Magnitude: 60 SP + 10 Iconic = +70%

Might for ability affinity only

• Might: 230 SP + 252 Mods + 41 Affinities = +523
• Precision: 14 Affinity + 30 SP = +44
• Weapon Chance: 13 SP + 3 Iconic = +16%
• Ability Chance: 13 SP + 3 Iconic + 5 Affinity = +21%
• Magnitude: 60 SP + 10 Iconic = +70%

Might no affinities

• Might: 230 SP + 270 Mods + 41 Affinities = +541
• Precision: 14 Affinity + 30 SP = +44
• Weapon Chance: 13 SP + 3 Iconic = +16%
• Ability Chance: 13 SP + 3 Iconic = +16%
• Magnitude: 60 SP + 10 Iconic = +70%

Might/Prec Builds

Might/Prec no affinities

• Might: 230 SP + 190 Mods + 41 Affinities = +361
• Precision: 14 Affinity + 30 SP + 200 Mods = +244
• Weapon Chance: 13 SP + 3 Iconic = +16%
• Ability Chance: 13 SP + 3 Iconic = +16%
• Magnitude: 60 SP + 10 Iconic = +70%

Might/Prec for all affinities (Leaning Might)

• Might: 230 SP + 190 Mods + 41 Affinities = +361
• Precision: 14 Affinity + 30 SP + 120 Mods = +164
• Weapon Chance: 13 SP + 3 Iconic + 5 Affinity = +21%
• Ability Chance: 13 SP + 3 Iconic + 5 Affinity = +21%
• Magnitude: 60 SP + 10 Iconic = +70%

Might/Prec for ability affinities only (Leaning Might)

• Might: 230 SP + 190 Mods + 41 Affinities = +361
• Precision: 14 Affinity + 30 SP + 140 Mods = +184
• Weapon Chance: 13 SP + 3 Iconic = +16%
• Ability Chance: 13 SP + 3 Iconic + 5 Affinity = +21%
• Magnitude: 60 SP + 10 Iconic = +70%

Might/Prec for all affinities (Leaning Even)

• Might: 230 SP + 152 Mods + 41 Affinities = +323
• Precision: 14 Affinity + 30 SP + 160 Mods = +204
• Weapon Chance: 13 SP + 3 Iconic + 5 Affinity = +21%
• Ability Chance: 13 SP + 3 Iconic + 5 Affinity = +21%
• Magnitude: 60 SP + 10 Iconic = +70%

Might/Prec for all affinities (Leaning Prec)

• Might: 230 SP + 114 Mods + 41 Affinities = +285
• Precision: 14 Affinity + 30 SP + 200 Mods = +244
• Weapon Chance: 13 SP + 3 Iconic + 5 Affinity = +21%
• Ability Chance: 13 SP + 3 Iconic + 5 Affinity = +21%
• Magnitude: 60 SP + 10 Iconic = +70%

Might/Prec for weapon affinity only (Leaning Prec)

• Might: 230 SP + 152 Mods + 41 Affinities = +323
• Precision: 14 Affinity + 30 SP + 200 Mods = +244
• Weapon Chance: 13 SP + 3 Iconic + 5 Affinity = +21%
• Ability Chance: 13 SP + 3 Iconic = +16%
• Magnitude: 60 SP + 10 Iconic = +70%

Precision Builds

Prec for all affinities

• Might: 230 SP + 41 Affinities = +271
• Precision: 14 Affinity + 30 SP + 248 Mods = +292
• Weapon Chance: 13 SP + 3 Iconic + 5 Affinity = +21%
• Ability Chance: 13 SP + 3 Iconic + 5 Affinity = +21%
• Magnitude: 60 SP + 10 Iconic = +70%

Prec for weapon affinities only

• Might: 230 SP + 41 Affinities = +271
• Precision: 14 Affinity + 30 SP + 264 Mods = +308
• Weapon Chance: 13 SP + 3 Iconic + 5 Affinity = +21%
• Ability Chance: 13 SP + 3 Iconic = +16%
• Magnitude: 60 SP + 10 Iconic = +70%

Prec no affinities

• Might: 230 SP + 41 Affinities = +271
• Precision: 14 Affinity + 30 SP + 280 Mods = +324
• Weapon Chance: 13 SP + 3 Iconic = +16%
• Ability Chance: 13 SP + 3 Iconic = +16%
• Magnitude: 60 SP + 10 Iconic = +70%

DAMAGE BUILD RESULTS

75% Might Damage / 25% Precision Damage
Build % Increase
Might Ability Affinity 63%
Might Full Affinity 62%
Full Might 59%
Mix Ability Affinity (Might) 55%
Mix Full Affinity (Might) 55%
Mix Full Affinity (Even) 53%
Precision Full Affinity 51%
Full Mix 51%
Mix Full Affinity (Precision) 51%
Mix Weapon Affinity (Precision) 49%
Full Precision 47%
Full Precision Weapon Affinity 47%

50% Might Damage / 50% Precision Damage
Build % Increase
Might Ability Affinity 62%
Might Full Affinity 61%
Precision Full Affinity 60%
Mix Ability Affinity (Might) 60%
Full Might 59%
Mix Full Affinity (Might) 59%
Mix Full Affinity (Even) 59%
Full Mix 58%
Mix Full Affinity (Precision) 58%
Full Precision 57%
Mix Weapon Affinity (Precision) 57%
Full Precision Weapon Affinity 56%

25% Might Damage / 75% Precision Damage
Build % Increase
Precision Full Affinity 71%
Full Precision 70%
Full Precision Weapon Affinity 68%
Full Mix 67%
Mix Full Affinity (Precision) 67%
Mix Weapon Affinity (Precision) 66%
Mix Full Affinity (Even) 66%
Mix Ability Affinity (Might) 65%
Mix Full Affinity (Might) 64%
Might Ability Affinity 61%
Might Full Affinity 61%
Full Might 59%

SUPPORT BUILD RESULTS

I ran a similar spreadsheet for restoration. I took a heal and modeled 100 ticks, assuming it will crit the exact number of times as our crit chance. Keep in mind that the value of Resto vs. RestoPow mods is a separate debate. You'd have to compare the results of your heals with your power bar, with the likelihood of over healing, with heal spamming, with HoT, etc. This assessment is based on pure healing out which will favor Resto.
http://files.enjin.com/208137/Files/Resto Modeling.xlsx

Build % Increase
Resto Full Affinity 53%
RestoPow Full Affinity 50%
Resto Chance Afinity 49%
Resto Mag Affinity 48%
Full Resto 45%
RestoPow Chance Afinity 44%
RestoPow Mag Affinity 43%
Full Resto/Power 39%


I ran a similar spreadsheet for controller modding. I took a POT tick and modeled 100 ticks, assuming it will crit the exact number of times as your crit chance. Keep in mind that the value of Vit vs. VitPow mods is a separate debate. You'd have to compare tick rates, time out of combat, sodas, weapon combo regen, and power transfer echoing. This assessment is based on pure power generation which will favor Vit.
http://files.enjin.com/208137/Files/Vit Modeling.xlsx

Build % Increase
Vit Full Affinity 16.4%
VitPow Full Affinity 14.9%
Vit Mag Affinity 14.7%
Vit Chance Afinity 13.1%
VitPow Mag Affinity 12.5%
Full Vit 12.4%
VitPow Chance Afinity 10.8%
Full Vit/Power 9.5%

RESERVED

RESERVED

I like this guide.

Sore, I'm trying to work out the might formula for a base of 2-5 and I'm pretty sure I'm not doing it right.

Explain like I'm 5?

Copy-pasting this:

> Applying Might
BASE DAMAGE x (1+ (MIGHT x .0045))

If your might was 1000 and base damage is 2.
2 x (1 + (1000 x .0045)) <-- Plug in the numbers
2 x (1 + 4.5) <-- Multiple
2 x 5.5 <-- Add
11 <--Multiple
11 <-- Round down

If your might was 1000 and base damage is 5.
5 x (1 + (1000 x .0045)) <-- Plug in the numbers
5 x (1 + 4.5) <-- Multiple
5 x 5.5 <-- Add
27.5 <-- Multiple
27 <-- Round down

EDIT: In the might section of the guide, I have spreadsheets you can just put your numbers into.

I added a link to a spreadsheet for capturing and projecting precision damage.

Your Precision section is not accurate. For starters having 0 DPS does not mean you do 0 damage per hit. You can observe this if you take off your weapon and use your fists, you will still do damage (small but some). This implies that there are indeed base numbers for weapon attacks. This is true but I will not go into it in this post.

Second weapon attacks do not "scale" like Might, Restoration and Vitalization do (which all use the same general equation). I like to think of Precision as "upgrades." This observation is also seen with pets and Combat Rating. Pets that are based on Combat Rating upgrade just like weapon attacks.

Third if you test weapon attacks long enough you will notice that you see all numbers between the lowest range and the highest range unlike the other stats. This is much different compared to Might that derives its final numbers directly from the base range. In the case of weapon attacks (as pointed out above) the number of possible values remains intact as you raise your DPS (and Precision as Precision translates into DPS).

I'm not going to post an upgrade equation at the moment (need to finalize testing) but just be aware that weapon attacks do have a base and DPS/Precision does not scale like Might.

Sore PM'd me about 2 weeks ago asking me to post how I test and I have to say that it is pretty similar to what Sore is implying.

First I never factor in criticals in my testing. All I care about is base numbers and a equation. If I cannot relate my output with an equation than my testing is not complete. Everything in a predictable environment has some sort of equation and it is likely that similar actions (Restoration and Vitalization for example) will use the same general equation. Knowing how to reverse the numbers you see in game is very helpful in finding the base numbers but using what Sore provided that is not that hard to do.

Before I start testing I figure out what the game is telling me. For example for healing I know that for every 4 Restoration I should get a 1% increase, that I heal 65% better in Healer stance and I will heal a Fire Tank 80% better. I can also tell from testing that I don't see a range of numbers in my final output and that some of the numbers are critical (higher) values. Because I am interested in an equation this information gives me the following terms:

1. I don't see a range in my numbers so multiplication must be used over addition for at least one term. At the same time my healing is not constant so there has to be range of numbers somewhere in the equation (1st term)
2. Restoration is a percentage increase (2nd term)
3. I have to add on a stance bonus which is also a percentage (3rd term)
4. There is some sort of receive bonus that factors in (4th term)
5. There is a chance for a second value which is higher than the first. Because the chance is not 100% you need 2 equations to describe both. This is also a percentage (5th term)

Now that does not directly show me the equation but because I studied mathematics I do know some basic mathematical properties. What I know is that a) multiplying a percentage by a number makes that number smaller not bigger, b) a number + a percentage = the number * (100% + the percentage), c) a percentage expressed as a number is the percentage / 100 and d) A * B = B * A (meaning that terms using multiplication can be rearranged).

Long lesson short the game is telling me that the healing equation must be something in the form: H = B * (1 + R / 400) * 1.65 * (1 + HR) * (1 + C). So just by listening to what the game has told me I already know the equation which I can than use (by ignoring criticals) to find base values if they exist (which they must in this scenario). After I reverse the numbers I see for each trait I use my testing is complete and I can project other numbers.

Now Sore asked me specifically how I tested healing. I test healing in the Magic Shop instance. In there you will find a fire that will constantly damage you and enemies that won't threaten to kill you. What I do is I aggro some of the enemies and stand on the edge of the fire. This keeps me in combat and ensures that I always take damage. To stay in combat simply walk out of the fire and into the enemies sights from time to time.

From here I can proceed with testing my heals. If I need Power I would use Canine for Nature or just go to the back of the flame where the enemies cannot hit me to get out of combat.

Using the information about equations above I use this fire to find base values. To test if a heal has been affected with an update I simply compare before and after base values.

As for splitting there really isn't an easy way to do this by yourself. If you have a bunch of friends willing to help you use a variant of the fire method by rotating players in and out of the fire. Once they are damaged enough heal them, find the base values per person and compare them to single target numbers.

Max targets are a bit easier. Find a group of friendly NPCs fighting some enemy NPCs (find some opposite faction content) and use your heals on them. All heals have an animation that you can see around the target use that figure out max targets. For example Sorcery's Circle of Protection attaches white energy around the target and the initial hit of Nature heals shows green vertical lights and leafs around the target.

Hope that helps Sore. I know it is long but you don't read a testing guide if you expect summaries.

P.S. You should mentioned the Vitalization Registration Bug in the weapon regen section as well. If you wish you may also simply link to my guide about it.

Very interesting point about the fists, and I once thought of it like that. However, weapon damage is definitely linear off your DPS stat with a 0 y-intercept. This doesn't really make sense because you can have 0 precision and do damage with fists but I did some digging. Fists have a damage distribution that is static and will not change based on your level, CR, or precision stat, though it can be boosted with damage modifiers. Because fists are constant and weapons are linear with DPS with a 0 y-intercept, any base damage weapons have will not be related to the damage you see from your fists. I thought weapons had base damage for a long time, but conceded to the probable explanation that weapons can't have 0 DPS anyways so it won't matter.

I used it as an example coupled with the other points. It still holds true because weapon damage is not constant. If weapon damage were constant (meaning no base) than you would not see a range at all. Having a range means that something is causing that range and because you can see all numbers between that range means that it is being added on.

I'll post more about it when I complete my testing but weapon attacks do have a base.

P.S. Fist damage is not constant it too has a range and like you said is alterable.

okay.. bad wording then . I meant constant with DPS, not literally constant. The parameters - min and max - of the uniform distribution for fist damage will be the same regardless of your precision stat, while the min and max parameters for weapon damage scale linearly with DPS.

I'm curious about what exactly you mean by "base damage" though. I think we're thinking of different things. With my definition, having a base damage will not affect whether or not you see a range of values. The probability distribution already guarantees this. Okay, maybe I was being confusing again because I'm not thinking "base damage" to be a single value here. It's more like what the min and max of weapon damage distributions, the range of damage values, would be with a DPS stat of zero. Since the min and max scale linearly with DPS with a y-intercept of zero, "base damage" in my definition should be zero.