Guild Wars 2

I heard someone over on guru saying that crit damage doesn’t effect phantasms, so I spent a few hours testing this.

From my tests, it looks like crit damage from traits and runes is very obviously applied to phantasms, however crit damage stats from equipment (I tested armour/weapons, not accessories) is ambiguous and may not be transferred at all. The higher crit damage set had both the highest (by a couple of points of damage) and the lowest (by hundreds of points of damage) critical hits compared to an armour set with 5% lower critical damage.

I made sure to only alter crit damage (and crit chance often unavoidably) and not power, condition damage or any other DPS stat. The only traits were defensive (toughness to help not die while testing) and granting fury to phantasms (crit chance, not damage) to make them faster. No buffs or conditions were used.
I tested using the Warlock phantasm as it only attacks once which made it easier to keep track since the combat log doesn’t log phantasm damage.

My methods weren’t exactly scientific being hampered by my resources and also my poor understanding of maths, so I’m eager to hear if anyone else has tested this, how, and their findings. I’ve posted this issue as a bug since the Wiki says the crit damage stat transfers and it seems very odd that it would only not transfer from armour/weapons and not anything else.

the only thing they might not transfer to are runes, as sigils effects dont transfer either.

Rune effects definitely do transfer. Base stats on armor/weapons appear not to.

perhaps the stats from runes transfer, but the special effects some runes give certainly dont. are you saying your illusions getting hit can summon rock dogs for you?

No, Wads. Considering the whole thread was about crit damage specifically transferring to phantasms I assumed you were on topic.

How are you conducting your tests? Phantasms seem to get 100% of my crit damage bonus. Keep in mind that base crit damage bonus is only 50% in this game. (At least for mesmers)

And remenber that crit damage is (stupidly) scaled by level.

I posted the details about the test over on Guru so I’ll quote it all.

I don’t have a head for numbers so I would be glad I hear if there was a better way to test. Please keep in mind that when I checked level 80 gear I was unable to find anything suitable to test vs zerker gear, as zerker has power as the first stat and the only other set I could find which did that was valkyrie, which also has crit damage.

For this reason I went with level 50 gear honed vs vigorous as it had the same power stat.

I just tested this out in the Mists on a Heavy dummy, using the Warlock phantasm (since it only hits once). No buffs, conditions, etc.
I changed the crit damage stat by having 10 Dueling the first time (crit chance and 10% crit damage increase) and refunding it the second time.

There was a clear and noticable difference. Every warlock crit with 10 dueling did more damage than the crits when I removed 10 dueling. I didn’t do the exact math, but the hits were:

With 10% Crit Damage: 2228, 2136, 2166. 2228
Without 10% Crit Damage: 1990, 2002, 2026, 1913

Was the test the previous poster did only checking crit damage stats on gear, or on runes, not traits? It seems like it would be much harder to test on gear/runes without altering the power stat so I’m interested to hear more from people who say that crit damage doesn’t transfer.

UPDATE:

I tested this again, with 10 dueling (for the crit chance so it didn’t take forever) and with swapping out rune stats. The first test had 33% crit damage, and the second test had 38% crit damage after swapping out a 25 healing rune stat for 5% critical damage. I chose healing so it wouldn’t interfere with the damage.

Used warlock again, since it attacks only once, and made sure the target had no conditions, and my phantasm and I no buffs.

33% Crit Damage : 2132, 2019, 2250, 2013, 2019
38% Crit Damage: 2314, 2143, 2266, 2153, 2254, 2167

Once again I see a noticeable increase. All I have left to test is the crit damage on armour, but I can’t do that in the Mists so it will take a while.

UPDATE:

10 Dueling for crit chance and Fury (20% crit chance) on Phantasms. 20 Chaos (staff skill cd reduction and damage reduction) and Signet of Illusions to help me and my phantasms survive the test.
Wearing green level 50 Honed (Power and crit damage) armour/staff, and then with level 50 green Vigorous (power and vitality) armour/staff. No runes, same accessories for both tests.

Using the Warlock on level 45 Blue Moas in Bloodtide Coast.

17% Crit Damage: 689, 760, 773, 717, 695, 716, 741, 725, 738, 724, 753, 701, 612, 614, 639, 615
12% Crit Damage: 682, 750, 716, 751, 718, 692, 724, 717, 692, 738, 695, 697, 686, 678, 740, 724

The highest crits were on the 17% crit damage set (760, 753 vs 751 and 750 on 12%) but the 17% crit damage set also had the lowest crits by far. I’ll need someone better at maths than I am to work out the significance, but it doesn’t look like 5% extra crit damage to me whereas the 5% increase on the runes was very noticeable. Of course for the previous two tests the crit damage overall was higher since it was in PvP on dummies and my level wasn’t scaling, so differences would be larger and more obvious.

I also don’t know how crit damage is calculated, but to me it looks like the crit damage stat on armour perhaps isn’t working. I’m fairly confident that the crit damage stats on runes and from traits is being applied to phantasms.

I did consider that:

A) Crit damage scaling by level would hugely reduce the amount my Warlock crit for in a level 45 map as opposed to being in the Mists on the dummy.

B) It would be harder to spot a 5% increase when the total crit damage was much lower, since 5% of less damage is a smaller and easily missed number.

C) I’m not particularly good at maths so I can’t tell what’s going on beyond ‘this doesn’t look any higher’.

That said, the armour/weapon test was ambiguous at best having the highest but also the significantly lowest crits. If someone who understands crit damage better can explain that, that would be great.

I posted this here hoping that others would have tested too or could spot flaws or suggestions in how to test the transfer on equipment better. It’s difficult trying to make the power the same, the crit damage vastly different and the crit chance high enough to test, and keep phantasms alive for the test!

Can anyone confirm if power effects crits at all? It would be a lot easier to test in full 80 gear if I didn’t have to keep the power equal.

Power level does not matter at all. What matters is the % increase of critical hits over the average normal attack.

For example…

Naked, no traits, iWarlock //THIS IS NOT ACTUAL DATA, JUST AN EXAMPLE
1000, 990, 1500 (crit), 1010, 1500 (crit)

From this data, the average normal attack would be 1000 and the average critical is 1500. This means the critical attacks are hitting for +50% (which is the normal baseline for critical damage)

Now lets say you throw on some Berserker’s and apply traits…
4300 (crit), 2000, 1900, 2100, 4290 (crit), 2000, 4310 (crit)

In this case, the average iWarlock attack is hitting for 2000 and the critical attacks are hitting for 4300. Basic math tells us the critical attacks are hitting for +115% damage in this example. Subtract the base 50% crit, and this means the attacks are getting +65% critical damage from gear. Now if the in game tool tip for +Critical Damage is +65%, you could be confident that the iWarlock is getting the full benefit of all the +Crit Damage on the character.

If that character happens to be showing +85% critical damage (on the tool tip), and the critical attacks were hitting for +115% instead of +135%, this would indicate that +20% of the critical damage is getting “lost”. Then you could use trial and error to figure out what was not being counted.

Also when testing this, it may be a good idea to do a test WITHOUT Sharper Images so you don’t have a Condition on the mob to further boost iWarlock damage.

Cliff notes: Power doesn’t matter. You should be averaging normal attacks and critical attacks separately, then figure out how much harder the critical attacks hit for.

Were the mists tests done with Steady weapons?

Pay, no I did them with regular PvP weapons since they don’t have stats aside from weapon damage anyway.

Ocellaris – Thank you for explaining! I did do it without sharper images (since conditions buff warlock damage anyway) but I will do it again and try to make a note of all the hits.
Maybe I’ll have to record it, because the combat log doesn’t show phantasm damage so that could get tricky.

iWarlock is probably the worst phantasm to be testing this with. It has a very wide damage range and this damage is randomized with each attack. You’d need to do hundreds (if not thousands) of tests to verify the effects of critical damage statistically.

A much easier way to do this would be to use one of the multi-hit phantasms like iWarden or iDuelist. The “dice roll” only happens once per attack cycle, so you can easily compare the difference between crits and non-crits within that attack. This eliminates the random factor in the testing, and the only thing you’d need to worry about are rounding errors.

You are really overestimating if you think it would take “thousands” of tests to reduce the margin of error on iWarlock spam. The damage range isn’t very large when there are not Conditions on the targeted mob.

Point taken, but its still drastically easier to work with a completely deterministic situation than a random one.

I tested this in pvp lobby with 1899 power, 30 dueling, berserker’s amulet/jewel and 6/6 rune of divinity, for a total of +62 critical damage. Using the defender I got 23 non-crits and 41 crits for mid-point damages of 171.5 non-crit and 365 crit. Which works out to crit being 213% of non-crit. That’s extremely close to the expected ratio. The data also fit the expected distribution for the defender (ie range of ±6% from mid-point) and the non-crit damage was pretty much exactly where expected, so it’s consistent with previous testing.

Then I recorded a single attack string from a warden under same conditions, it did 178 non-crit and 377 crit, for 212%.

So I’d probably say phantasms use +critical damage from gear.

Now, there’s always the open question of whether or not the tooltips for the various runes are actually correct. PvP Divinity runes, at least, seem to be accurate.

I’m pretty sure illusion stats are the same as your listed stats. Anything that doesn’t show on your stat sheet (trait effects, sigils, special rune effects) they don’t get, anything that does (increased stats from armor and traits) they do. It’s worth testing whether they get the bonuses from sigils that add stats for every kill.