Guild Wars 2

Warning This is going to be long and there will be no TLDR

I want to begin the basic math for engineers so we have a base to compare
builds and setups with and have a good way to check our hypothesis for any
changes we suggest.

What this thread is:

This is a place to post factual results tested in game. We will use these
numbers to compare and eventually arrive at specific and realistic
conclusions on abilities, traits, and equipment while avoiding what it
“feels like” is happening.

I would like this to become a discussion that eventually answers questions
such as the power VS precision debate and which is better for effective HP
toughness or vitality? Then eventually moving towards things thresholds and
balancing of stats.

What this thread is not:
* A suggestion thread: there are hundreds of other threads for that.
Suggestions on the class or class abilities will be ignored.

* A comment forum: Please keep posts about specific information
you have tested in game where you can verify the numbers from both
sides or in a controlled environment. So for example do not post
numbers from spvp or tpvp matches. Instead only against dummies in
HoM (just an example. Any proper control will be accepted).

* An ArenaNet message board: we will be operating under the unlikely idea
that ArenaNet is not reading this forum as to keep discussion between
interested parties and away from messages hopefully seen by an employee
and unrelated to the topic. This should not be taken to mean we do
not want ArenaNet’s input should they have something to further the
discussion but more intended as a guideline for players posting.

* A kitten stroker. Please keep on topic and factual. No one here will care
if you have bigger numbers except to refute/support a claim already made
in the previous posts in this thread.

Now that the guidelines and disclaimers are out of the way…. let us begin with the
facts as compiled by wiki.guildwars2.com to illustrate the building blocks we need
(in order to finish the puzzle we have to look at the pieces first).

Attributes:

Power(stat): Every point increases attack(stat) by one point.
- Attack(stat): Equal to weapon damage plus Power(stat), and together
determines amount of damage dealt.

Precision(stat): Increases critical hit chance (CHC).
#note: The amount of additional precision(stat) required to
increase the chance of a critical hit by 1% is determined by
the following formula:
(0.0024 × Level^2) + (0.06 × Level) + 1
***thank you Mario.4036 for the correction to precision ( bad copy pasta is bad :P )

The required precision(stat) to increase the CHC by 1%:
At level 20 is 3.16.
At level 40 is 7.24.
At level 60 is 13.24.
At level 80 is 21.16.

- CHC: chance of landing a critical hit with their attacks and spells.
#note: Applied conditions such as bleeds and poisons can never crit.
However, skills which cause direct damage in pulses such as turrets or
some ground-based AOEs can crit.

Toughness(stat): every point increases armor by one point which improves the character’s ability to withstand direct damage. Toughness(stat) has no effect on falling damage.
- Armor(stat): is equal to Defense plus Toughness attribute, and will determine
how much incoming direct damage will be reduced.
#note: Armor does not mitigate Condition Damage.
- Defense is a value primarily increased by armor equipment. As mentioned
above (see toughness) Defense + Toughness(stat)= Armor.

Vitality(stat): every point increases maximum health by ten points.
#note: Base Health is different for different classes. At level 1
all characters also have a base 24 Vitality(stat), and at level 80 a base
916 Vitality(stat). See this link ( wiki.guildwars2.com/wiki/Health ) for table.

Experiment 1 Rifle Weapon Abilities:

Naked Character (no armor no traits)
/ Power 916 / precision 916 / Toughness 916 / Vitality 916 / Attack 1061 /
CHC 4%, +0 critical dmg/ Armor 916 / Hp 15,082 /
Weapon: PVP Steady Rifle (attack 125 – 125)

Hip Shot hit results after 100 shots at Heavy Training Golem:
Hit 28 dmg
Crit 42 dmg

Results:
Piercing shots hit for the same damage for all targets they pierce
within the weapon’s range. From what I observed the critical hit chance was
independent for each target hit and was not based on the original target
or any previous targets hit (i.e. whether the first target in line crit or not had no
bearing on any successive pierced target’s crit chance).

Blunderbuss results after 50 shots against 3 Heavy Training Golem in varying range
(closest:~100 furthest:~400):
Dot 43 regardless of range (see results)
Hit 70(~100 range), 52(>100 & <400 range), 44(~400 range) crit 105(~100 range), 61(@>100 & <400 range), never crit at ~400 range

Results:
Interestingly enough the dot seems to tick faster the closer you are
but does the same amount of damage each tick. I was consistently ticking the closest
target for 43 dmg * 11 ticks and the furthest target for 43 dmg * 4 ticks.
The dmg per tick never changed. The only thing that did change was the tick frequency in the same total bleed time. Also shots are not blocked by previous targets. You could look at this as all shots pierce 100% of the time or that they have a sort of cone area effect regardless of what targets are in the way. However I can not
determine if obstacles completely block the cone effect or if the ability allows for a partial effect if some of shot can hit an enemy but some cannot.

Overcharged Shot results after 50 shots against Heavy Training Golem:
Hit 44 dmg
Crit 66 dmg

Results:
Regardless of range within its limit of 400. Shots only hit the first
target it comes into contact with even if target is further down the path and
within range. There were no pierce or AoE behaviors observed during this test.

Jump Shot results after 50 shots against grouped Heavy Training Golems:
Crit on leap 61
Crit on landing 126
Hit on leap 40
Hit on landing 84

Results:
Vulnerability effects all targets within the radius of it’s leap AND
landing (tested up to 5 grouped targets). This ability can apply it’s
vulnerability effect to two separate groups of enemies. In addition,
if the lept from target and leapt to target are one and the same, the
vulnerability effect stacks and the duration refreshes on landing. So it is
possible to get 3 stacks of vulnerability on a target for 1 second then up it
to 6 stacks of vulnerability for the next 7 seconds on landing.

Reserved for Further testing (as soon as time allows).

Bleed has been confirmed to stack ticks of damage, not increase the damage of each tick.

It has not been confirmed, however, whether players with lower condition damage affect your ticks or if there is an internal tick damage that distinguishes each player’s tick amount.

Sporadicus.1028:

Bleed has been confirmed to stack ticks of damage, not increase the damage of each tick.

It has not been confirmed, however, whether players with lower condition damage affect your ticks or if there is an internal tick damage that distinguishes each player’s tick amount.

That is an interesting question. It would be pretty terrible if a team is focusing an individual enemy and the highest damage per tick condition is being overwritten by another one based on which lands first.

The only thing I could find on the matter was this quote from the official GW2 wiki:
“Condition damage that is already affecting a target will update per tick according to the source’s stats. For example, An Engineer may inflict Bleed with an Elixer Gun toolkit and then swap to a pair of +condition damage pistols in order to increase the damage over time, even if no further conditions are inflicted. "

This only seems to be talking about your own damage though. I will need to get another person to help me test that in the near future in the HoM.

Adding to the quote above, I found a little more information on conditions that may give us some insight until I can test the values themselves.
“A condition is a negative effect that can be inflicted against enemies through the use of skills and affecting traits. Conditions are characterized by their effect and duration. When a condition is reapplied (to a foe who already has it), some conditions will stack increasing in effect, some conditions will increase in duration, and a couple will have no change. When a condition is removed using a skill, the entire stack of that condition is removed. Condition duration is accurately added and will not be rounded.”
(from http://wiki.guildwars2.com/wiki/Condition)
There is also a table with each condition and if it stacks in duration upon reapplying or stacks in intensity.

Sporadicus.1028:

Bleed has been confirmed to stack ticks of damage, not increase the damage of each tick.

It has not been confirmed, however, whether players with lower condition damage affect your ticks or if there is an internal tick damage that distinguishes each player’s tick amount.

Why do you state that this is not confirmed? It has been confirmed for some time, and well discussed. Perhaps you simply missed it……anyway bleeds stack up to a set number like 25 or something might be 35 I do not remember. Basically it stacks the top 25 if that is the cap. In other word, if I have higher condition damage then you and I am stacking 5 stacks and players A, B, C, and D are also higher condition damage then you and stacking 5 stacks a piece as well, then your bleeds will not even register.

Which sucks if there are several conditioned focus players at the particular fight.

Experiment 2 Flamethrower kit Weapon Abilities:

Naked Character (no armor no traits)
/ Power 916 / precision 916 / Toughness 916 / Vitality 916 / Attack 916 /
CHC 4%, +0 critical dmg/ Armor 916 / Hp 15,082 /

Weapon: none equipped (flamethrower kit only)

Napalm hit results after 100 shots at lines of 5-6 Heavy Training Golem:
hit 328 dmg * 10 ticks (1 tick per sec over 10 seconds)
crit: n/a cant crit

Results:
This ability is confusing. In my tests I lined up 5 or 6 heavy training golems in a tightly spaced line and used the ability. In some cases all 6 golems would get hit by the fire they were standing in and take damage as normal. The majority of the time (about 2/3) only four would be hit by the fire and would then take damage as normal. No amount of repositioning of the enemies or cast location/ground target would change these results. It is possible the ground target is incorrect and is in fact much more thin than it shows and is missing golems completely or it may be bugged. Either way getting more than 4 enemies to line up in a real fight is highly unlikely. The other thing to note is that the enemy must stay in this area to receive the entire duration. It does not apply the full effect simply by traversing through it. As soon as the enemy leaves the line of effect the damage stops (even though their icon cycles till the single burn falls off).

Flame Blast results after 100 shots at lines of 5-6 Heavy Training Golem:
This ability has two stages. I will call them travel and blast.
Travel Hit: 230 -250 dmg per enemy
Travel Crit: 349 – 360 dmg
Blast Hit: 533 – 592 dmg in a small aoe
Blast Crit: 840 – 894 dmg

Results:
From the tests I did I found that there was no limit to enemies hit in either stage of the ability(tested up to 6). Also crits on travel and on blast are based individually per enemy hit (in the same way hip shot is). Each enemy hit by this attack in both the travel and blast phases has a chance to get crit by the corresponding damage. If an enemy is hit by both the travel and blast phases it seems to effectively roll twice (once for each attack it was hit by) as I have crit on both in a single shot to enemies in the travel path and close enough to the epicenter of the blast when it goes off.

I skipped Air-blast and Smoke-vent as there is no damage component or secondary to them and in the limited time I tested them (about 20 uses each) they consistently applied as described in the tool-tip. This is why I skipped net shot in section one as well.

Experiment 3 Grenade kit Weapon Abilities:
Note: though the tool-tip says throw ‘A’ grenade in fact two are thrown when using any skill in the grenade kit (without traits). It would seem the grenade itself caps at 10 enemies effected whether it is in range of more or not though.

Naked Character (no armor no traits)
/ Power 916 / precision 916 / Toughness 916 / Vitality 916 / Attack 916 /
CHC 4%, +0 critical dmg/ Armor 916 / Hp 15,082 /

Weapon: none equipped (grenade kit only)

Grenade skill results after 100 throws at groups of 5-6 Heavy Training Golem:
Hit: 154 – 170 dmg
Crit: 242 – 249 dmg

Results:
This skill is a direct damage multiple grenade toss where each grenade goes off in a small aoe. Everything hit by each grenade explosion has a separate chance to crit but interestingly enough if it is not a crit the same damage is applied to all targets in the all the grenades blast radii. It would seem grenades do not calculate the ‘white dmg’ (or non crit hit) with individual rolls. I would guess that conceptually three rolls are happening (one for crit chance, one for crit damage, and one for white damage) but the roll determining if the ability crits or not is in fact rolled individually (as you can have more than one crit in the bunch), then damage for the crit is the same on all crits that happened in that toss.

Shrapnel Grenade skill results after 100 throws at groups of 5-6 Heavy Training Golem:
Hit: 169 – 182
Crit: 259-270
Condition: 12 seconds of bleed (43 dmg per tick * 12 ticks)
Results:
The behavior of this grenade skill was exactly the same as the previous grenade skill. Outside of the instance of a crit, the damage for both grenades is the same and effects all enemies in the target area. The bleed is consistent with other tested bleeds and ticks in the same manner described in those tests as well. Consistently hitting 24 ticks in 12 seconds of 43 damage in all 100 throws.

Flash Grenade skill results after 100 throws at groups of 5-6 Heavy Training Golem:
Condition: 10 sec (5 sec per grenade explosion * 2 grenades thrown) of blind
Results:
Works as described. Throws two grenades that effect all enemies in the explosion radius with the blind condition. Each grenade hit gives five seconds of blind on the target (as blind stacks duration not intensity). So this ability will consistently stack a 10 second blind on all enemies it effects (without traits).

Freeze Grenade skill results after 100 throws at groups of 7-10 Heavy Training Golem:
Hit: 154 – 169 dmg
Crit: 240 – 254 dmg
Condition: 4 sec (2 sec per grenade explosion * 2 grenades thrown) of chilled

Result:
This grenade ability had a much larger radius and applies the freeze condition to all targets it hits. The behavior of the grenade itself is the same as the others described as far as white damage and critical hits go.

Poison Grenade skill results after 100 throws at groups of 7-10 Heavy Training Golem:
Condition: 10 sec (5 sec per grenade explosion * 2 grenades thrown) of poison + 2 poison clouds that reapply (and add thereby add to the duration of) the condition every second for 5 seconds ( as long as enemy stays within one of the clouds).

Result:
This is my favorite grenade. While there is no blast damage to the grenade itself, there is a blast cloud effect. Unlike the other grenades, only one of the two will typically hit where you target. The other grenade of the two as far as I can tell places itself overlapping and at random at an area close to the original impact. The damage ticks and stacks duration much like the bleed on shrapnel grenade does but with one added bonus. The cloud that is left will continue to add a 5 second duration poison to the subject as long as they are in the cloud once every second. This typically resulted in a poison duration of around twice the blast duration ( ~ 15-20 seconds) with an enemy in the cloud the whole time during my tests.

That concludes my testing for today as that much has already taken a significant amount of time to test and compose.

Attack stat as listed in the hero panel, and wiki seems to be useless and wrong.

Damage is factored as:
skill coeff*weapon damage*power/2600

Skill coeff is hidden, but easy to figure out with the other stats as constants. Most of these have been tested and listed. Weapon damage is your weapons max damage. Kits have a weapon damage of 969 at level 80.

Power most certainly does NOT increase attack 1 to 1.
Power increases damage by a % of your weapon damage.
Meaning 1000 power is like + sixty damage when using a 1000 damage weapon.
But 1000 power is only +30 damage when using a 500 damage weapon.

Also, tooltip listed damage is shown as your MAX damage vs a heavy armor target.
You will do more then listed vs low armor targets.

Consider the damage you do vs critters in WvW and about the world. You hit rabbits for 200,000. They have no armor.
So, high armor does not reduce the damage you take. low armor increases it.

The toughness of the opponent is also very relevant to your damage out put

@Casia

I think I misunderstood you, so if you would bear with me… can you choose between the three options below as to what you are trying to say:

1) A weapon with max damage of a 1000. Power of 1000 (lets say that means 60% since I have no idea how power actually affects anything…). My damage to a heavily armored like-leveled character is now 600 on a consistent basis not accounting for crits?

2) Same Scenario, but instead of power revealing how big a percentage of the max damage one gets from their weapon, now it is added to weapon damage, and so the damage I would get against heavily armored like-leveled enemies is now 1600?

or 3) None of the above, I misunderstood you, and am a moron.

power is not added.

If you had a weapon that deals 1 damage, but 5000 power. you would still deal like 1 damage. literally, 1.1 damage As power is multiplied into weapon damage to determine your output.
Example:
Rifle.
1205 max damage on an exotic rifle. (986-1205)
If you have 1450 power.
the formula is:
skill coeff * 1205 * 1450/2600

Skill coeff for rifle are:
1- .794
3- 1.46/.91
4- .91
5- .819/1.64

Your tooltip with these stats will thus read:
1. 397 damage (.794*1205*1450/2600)
3. 978 damage close, 611 damage far.
4. 611 damage.
5. 550 on jump, 1100 on land.

This is the damage you will deal to a “heavy” armored target, specifically the heavy golems. (I do not currently know what the normalized “heavy” golem is. probably round 2k.) 1836 is default light armored player in exotics with no extra toughness.
While when you attack lighter armored targets you will deal more damage.

.794 * 1 weapon damage * 5000power / 2600 (level 80 heavy armor)

*Note edited to correct the formula, which after talking to someone, I see how it works better.
The 916 constant was inaccurate. The 916 represented the armor of your target. The tooltip assumes a level 80 armor of 2600. so 2600 is the value you divide by for a tooltip. Note, most players do not have that much. so you will in fact do more to a lower armored player.
Again, note 1836 is light armor exotic gear, with no extra toughness via traits or gear.

Casia.4281:

Attack stat as listed in the hero panel, and wiki seems to be useless and wrong.

Damage is factored as:
skill coeff*weapon damage*power/2600

Skill coeff is hidden, but easy to figure out with the other stats as constants. Most of these have been tested and listed. Weapon damage is your weapons max damage. Kits have a weapon damage of 969 at level 80.

That is correct. A real test I just did to illustrate this is as follows: If it was as simple as power + weapon damage and attack exactly equaled our non-crit damage we would do a base of 1041 dmg with steady weapons using an ability such as hip-shot in the heart of mists. Currently we do about 28.

Here is more information on damage from the wiki itself:
http://wiki.guildwars2.com/wiki/Damage

Casia.4281:

Power most certainly does NOT increase attack 1 to 1.
Power increases damage by a % of your weapon damage.
Meaning 1000 power is like + sixty damage when using a 1000 damage weapon.
But 1000 power is only +30 damage when using a 500 damage weapon.

Not sure if this is a typo or not but Power + weapon dmg DOES in fact = attack. Attack =/= DMG however. Just throwing that out there. The above test I mentioned would be much higher dmg if attack = dmg on a 1:1 ratio (and we would rival backstab with our auto-attack alone.. oh to dream).

To clarify, the only reason I included dmg numbers in all the above tests is to show differences and similarities in behavior of skills. For example, I could not have known for sure grenade damage rolled the way it did without a proper control variable. It seemed the easiest way for an initial test. We have however moved forward a bit into stat weights (which is great!). Thank you for the additional information!

ok, yes. power increases the “attack” stat 1 to 1.
The “attack” stat has no effect on anything. it is not used in any equation.

Attack does not effect your attacks.

Casia.4281:

ok, yes. power increases the “attack” stat 1 to 1.
The “attack” stat has no effect on anything. it is not used in any equation.

Attack does not effect your attacks.

Agreed. Interestingly enough the wiki does state:
“Power attribute directly increases Attack (statistic) which affects all damage done by attacks. "
Though it does not have any bearing on anything currently used in the game I can see. Attack is simply an unused variable for storing the result of weapon damage + power. It seems a little strange to me that attack is this way and is not something more along the lines of (damage * power) to coincide with the actual math being used for dmg.

Experiment 4 Elixir-gun kit Weapon Abilities:
Naked Character (no armor no traits)
/ Power 916 / precision 916 / Toughness 916 / Vitality 916 / Attack 916 /
CHC 4%, +0 critical dmg/ Armor 916 / Hp 15,082 /

Weapon: none equipped (Elixer-gun kit only)

Tranquilizer dart skill results after 100 shots a Heavy Training Golem:

hit: 124 – 139
crit: 190 -206
condition: 1 sec weakness & 4 secs bleeding (42 dmg per tick and 4 ticks over 4 secs till 5 stacks when it hits for 43 dmg per tick for one second then drops to 4 stacks and again 42 dmg per tick and 10 ticks over 4 seconds)

Results:
What I noticed about this ability was that the majority of damage was coming from the bleed stacking. While you hit for 124 – 139 (without crit) you can keep a 4 – 5 stack bleed on the opponent just with the auto-attack. At this stack amount the golem bled for three ticks of 42-43 dmg every second. That is ~129 dmg a second plus the 124 – 139 per hit. A fairly strong auto-attack even without factoring in it also keeps up a weakness dmg reduction on the enemy. This attack is a single target effect. If something moves between you and your target it will get hit instead.

Elixir F skill results after 100 shots at a Heavy Training Golem:

Hit: 234 – 256
Crit: 361-379
Effect: Hits four targets (you can count as one of them) applying cripple to enemies and swiftness to ally.

Results:
This ability is very interesting and complex. It seems to roll dmg only once hitting all targets with the same dmg unless a crit happens. When two crits happened in a single fire, the crit dmg was the same for both crits as well. The swiftness is relatively pointless since firing the weapon at an enemy puts you in combat rendering the swiftness gain greatly reduced and barely noticeable. The cripple on the other hand is very nice. I would venture to say using this ability at ranges of 300+ is best (until its pathing can be deduced). If it hits yourself (or an ally) you miss both damage and an additional target for it to effect (as yourself or an ally will occupy a target). The 3 sec swiftness duration rules out using on your party out of combat as a swiftness buff as an alternative as well. This ability will make the most difference when no allies are hit with it and four enemies are groups and hit. It does not effect a single enemy multiple times (it will bounce to you from them then dissipate) but sometimes would effect the same enemy twice if bounced back to from another enemy and yourself.

Fumigate skill results after 100 shots at a Heavy Training Golem:
Hit: 5 ticks of 25 -27 dmg
Crit: 39 – 41
Condition: 3 ticks for 84 dmg poison & Vulnerability & Cures conditions if allies are hit.

Results:
The ability effects enemies in a cone pattern and ticks 5 times over the course of its animation. If all five hit the target a cumulative dmg of ~125 – 135 will be done to the target. Each tick also applies vulnerability and poison to the target. In practice, the vulnerability did not increase the actual dmg taken till the last tick (at 5 stacks) but lasted long enough that another ability could be used to take advantage of it. The poison ticked 5 times for 84 dmg consistently over the course of 5 seconds (1 second per tick). This skill seems to not be able to effect more than 5 targets regardless of how many are in the effect of the cone.

Acid Bomb skill results after 100 shots at a Heavy Training Golem:
Hit: 266 – 290 * 6 total applications (1 on use and 5 more over 5 seconds)
Crit: 402 – 435

Results:
Acid bomb decides dmg on the first tick but rolls for crit on each hit. So if you have 4 enemies being effected and standing in the area, they will be hit for the same dmg every tick but will each have a chance to crit every tick. The ability itself hits upon application then ticks the area for an additional 5 ticks. As stated previously, all dmg from the initial hit to the last tick will be the same number but each enemy has a chance to crit for every tick/pulse they are affected by.

If you enjoy discussing and testing class mechanics and love this profession as much as I do, please feel free to correct any discrepancies and add your own test results to compare. The more data we get the better we can understand the mechanics and get past the rumors, QQ, and kitten stroking in order to really understand and utilize the engineer to it’s full potential.

Some skill coeffs.
skill coeff*weapon damage*power/2600

kits are 969 weapon damage at 80. 2600 is armor used for tooltip. A player may have more or less.

grenade:
1: .478 (x2 default .956, x3 with grenadier. 1.434)
2: .524 (x2 default 1.05, x3 with grenadier. 1.57)
4: .478 (x2 default .956, x3 with grenadier. 1.434)

Rifle:
1: .598 (pierces)
3: 1.45/.91 (near/far aoe)
4: .91 (aoe)
5: .82/1.64 (jump/land aoe)

P/P
1: .324 (aoe, with piercing x2 happens often. for .65) bleed scales at .05
2: .37 x5 over 2seconds. 1.85 over 2s. (auto-a is .75s or 1.3a/s, not 1 attack/s.)
3: .37 (5 hit bounce, can bounce back)
4: .305 (x3 .915 aoe.)