Talk:Critical hit

It seems to me that only melee weapons and bows are subject to critical hits. Wands and staffs do not seem to generate them. If this is true, I'd like to add it to this article. Can anyone confirm one way or the other? --Squeg 06:31, 19 October 2005 (EST)

Ensign published results of some critical hit testing, 2 level 20s, one serving as a punching bag for the other, along with the raw data, in the thread here, and a linear fit gave a rate of y = 0.0144x + 0.0027, with y being the critical hit chance; roughly 1.44% per attribute level, and 17.5% or so at level 12. The excel file with his raw data can be found here, though the confidence intervals can be pulled in a bit by using +/- z(a/2)((p(1-p))/n)^0.5 instead. Ensign sampled nearly 8000 hits, so the data has a degree of accuracy surpassing most studies, and I feel confident correcting the page based on it. --Epinephrine 23:16, 16 March 2006 (CST)

How do critical hits work with dual dagger attacks? Do both hits get a critical or are those independent too?


 * I've used Twisting Fangs and gotten a crit with one attack and not the other before. As long as dual strike skills follow the same rules for dual strikes due to points in dagger mastery, I'd assume that a dual strike simply causes another attack, with randomly chosen damage and another check for a critical hit. --Armond Warblade (talk) 18:33, 17 August 2006 (CDT)

The skill Critical strike produces two critical hits, but I think they are calculated as two attacks. After all, the first one can mis/be blocked/evaded, and the second could still hit.

How about adding this article to the Game Mechanics category?

The table is inconsistent with the formula. The critical hit rates from the table and the original data are closer to 0.013x + 0.01. -- Gordon Ecker 00:24, 30 September 2006 (CDT)


 * And the values on Ensign's original table are closer to 0.0133x + 0.01. I've corrected the tables. -- Gordon Ecker 00:51, 30 September 2006 (CDT)

Circumstances of critical hits?
Do the chances for critical hits change when attacking moving foes? Knocked down foes? Foes from above? Foes from behind?
 * What I have found is that you are guaranteed a critical hit when striking a moving foe from behind. Don't know if this is considered fleeing, but anyway, it works regardless of which direction they are moving, as long as you hit them from behind. 80.202.245.195 19:08, 10 August 2006 (CDT)


 * What range of angles counts as "behind"? a civilian 20:26, 13 January 2007 (CST)


 * Skill level appears to mean the weilder's attribute level. Does the weapon's minimum attribute level have any effect; perhaps a 15-22 8-min sword may critical less often than a 15-22 11-min sword.  Perhaps also wands do critical but since there is no attribute level associated with 'wands' that the chance is as for swords presuming attr level is zero, 1.3% or whatever.  Or perhaps a fire wand criticals based on the wielder's attribute level in Fire. Oye 14:22, 12 December 2006 (CST)

Change to formula
For now, I've reverted the anon edit for the simple reason that it did not contain justification, yet changed the formula without modifying the results. If the new formula was correct, it should be explained in the edit summary box. --- Barek (talk • contribs) - 19:03, 7 August 2006 (CDT)

0 axe
Say you had 0 axe mastery and were using a req. 9 axe, and then got a critical hit. Would you still do the maximum damage of the weapon x √2? &#91;&#91;AOTT]] 00:05, 14 August 2006 (CDT)
 * Criticals don't simply deal weapon max times sqrt(2) points of damage. For criticals, damage is worked out like normal except the damage from the weapon's range is automatically the max and either the damage is scaled by sqrt(2) or your "attack rating" is raised by 20 (the two work out equivalently).  --68.142.14.39 00:27, 14 August 2006 (CDT)
 * take a look at Damage, the full equation, including criticals is there --Honorable Sarah [[image:Honorable_Icon.gif]] 00:29, 14 August 2006 (CDT)
 * You're doing the replying to me thing again. --68.142.14.61 18:53, 17 August 2006 (CDT)
 * After reading the Damage link, I think I get it AOTT 10:14, 18 August 2006 (CDT)

Grunting after being critted :)
I'm pretty sure that the noises PCs make are caused by critical hits. Like when the female mesmer makes that funny highpitched "HA!" shriek after being cracked by a hammer. Can anyone else confirm?
 * Yeah I believe there is a sound effect made by victim's of critical hits, including yourself. At least I've always assumed this was true but never paid enough attention to be sure. -- BrianG 01:32, 2 January 2007 (CST)
 * I'm not sure if every grunt indicates a critical hit, though. I'm being attacked by a Strider, whose damage usually ranged between 6-9. I got hit with a 6 and grunted. --Nova 16:46, 30 January 2007 (CST)

Wands
I don't know about pets, but it's definitely possible to get a critical hit with wands. --Fyren 00:20, 8 September 2006 (CDT)

What about spells? Do any spells have a chance to critical? I'm sure that damage spells *without* projectiles such as Deathly Swarm cannot critical. But I think I've seen projectile spells such as Lightning Orb crit.
 * Pretty sure no spell can. --Fyren 19:31, 21 September 2006 (CDT)

Well, I removed the note completely, incidentally removing that pets cannot crit. I have no idea this is true. Gordon never responded to this after adding that wands can't and he's also the one that added that pets can't crit in pet. If he's (or anyone) is more sure about that, feel free to re-add the pets note. --Fyren 12:03, 25 September 2006 (CDT)


 * Isn't there a skill that makes your pet crit? --Armond Warblade (talk) 00:28, 8 December 2006 (CST)

I carry a starter wand on my assassin, funny when that criticals. Mass damage! &mdash; Skuld 12:08, 25 September 2006 (CDT)

---Do the noises differ from a regualar grunt? Guild Wars 23:07, 6 January 2007 (CST)

I just wanded two Practice Targets to death with 16 in the wand's attribute, and only got a single critical hit, which suggests a critical hit rate of around 2%. -- Gordon Ecker 02:57, 15 February 2007 (CST)


 * I did some further testing. With 12 Beast Mastery, a level 20 Hearty Phoenix against an AL 60 practice target with a hero spamming "Go For The Eyes!" the lowest damage value was 15 and the highest three damage values were 35, 34 and 33. With a 12 Beast Mastery, a level 20 PvP Elder Wolf against an AL 60 practice target with a hero spamming "Go For The Eyes!" the lowest damage value was 17 and the highest three damage values were 37, 38 and 40. If critical hits were being inflicted, the highest damage value would be roughly 40% greater than the second highest damage value. In both tests I killed about half a dozen AL 60 practice targets. -- Gordon Ecker 18:10, 22 February 2007 (CST)


 * I just did some more testing with "Find Their Weakness!". It was triggered by a 30 damage hit, a 35 damage hit and a 40 damage hit. -- Gordon Ecker 18:27, 22 February 2007 (CST)

Minions
Are Minions capable of a Critical Hit? If so, what's the rate? --Zinger314 15:17, 1 October 2006 (CDT)


 * The last time I tested it they couldn't get criticals. Details are in talk:minion. -- Gordon Ecker 18:11, 22 February 2007 (CST)

Attacker/Target level difference
The topic indicates that "Attacker and target levels are also factors," however there's no mention of the effect. Has anyone done any testing to see how this effects the critical percentage? Dfscott 10:45, 6 October 2006 (CDT)
 * You have a higher chance to crit against a lower level target. I don't know if you have a lower chance to crit against a higher level target.  I have no idea about the exact numbers involved.  --Fyren 15:50, 6 October 2006 (CDT)
 * I've found that when your fighting a foe with a level low enough you don't gain experience, you critical about 75% of the time. I figure there's some ratio for each suceeding level/ armor level. Equal level foes hit about 1/20 of the time with a maxed attribute or halfway and modest critical strikes.THIS IS UNCONFIRMED!Dark0805 22:14, 24 November 2006 (CST)
 * Testing with a Candy Cane Hammer, 12 Hammer Mastery and 12 Critical Strikes resulted in a critical rate of over 90% against level 5 practice targets. -- Gordon Ecker 03:22, 10 January 2007 (CST)
 * Beating up level 2-3 Mantid's outside of Shing Jea Monastary with 13 Critical Strikes gives me about a 99% chance of a critical. Might be 100%, I'm not sure. But I think I remember getting a single non-critical hit sometime between when I hit the first Mantid and when I landed the final blow to the last living creature on the entire island. :D Jioruji Derako 03:13, 11 January 2007 (CST)
 * One the same subject, I'm quite sure I didn't get anywhere near consistant damage ratings when I was a mere level 5 out there. I can run a bit of testing on that sometime, with 3 points in Critical Strikes so I can tell when I critical... I'm quite certain it has everything to do with the difference in level between the attacker and the target, though. Jioruji Derako 03:18, 11 January 2007 (CST)

Still a stub?
This article has grown a lot since it was created. Does it still warrant the stub? There's really not all that much more we could add to it. --Armond Warblade (talk) 00:26, 8 December 2006 (CST)

Added Stoneflesh Aura to the list -- (-Da*Man-) 19:46, 11 December 2006 (CST)

Removed Stub section. I too think it's far from being a stub right now. Feel free to revert, but do tell us why ^^ Finrod 01:45, 13 December 2006 (CST)

Scythes, Critical Hits and You
While I was testing a build in the Isle of the Nameless, whenever I used any attack skill and it was a critical it was the same number on both the Practice Target and the Adjacent one. Could anyone help me do more testing in regards to this so we can get some clarification? IGN: Angel of Arnor Finrod 02:10, 13 December 2006 (CST)


 * What kind of testing would be needed? It seems to me that when a scythe does a crit, it does it to all the targets (instead of having separate chances to crit for each target). Remember that scythes attack all foes adjacent to you (or something like that). I'm willing to help, though. --Armond Warblade (talk) 00:43, 14 December 2006 (CST)


 * Well, I just wanna prove it. I know it works that way, and I think it's important enough to mention in the main article.. [[Image:Light of Deliverance.jpg|20px]] Finrod 12:05, 18 December 2006 (CST)


 * Take wild blow and hit the target a lot of times. "A lot" depending on what'll satisfy you.  --Fyren 12:23, 18 December 2006 (CST)


 * Wild Blow seems to be referenced here a lot... But is there a spot where you can stand in the middle and be surrounded by targets, as opposed to a target surrounded by targets? Been too long since I've been there. --Armond Warblade (talk) 23:43, 20 December 2006 (CST)


 * You don't need a bunch of targets; two is enough. Use dummies on the isle of the nameless.  --Fyren 00:00, 21 December 2006 (CST)

Residual Plot
Did anyone plot the residuals and see if the graph warranted a linear equation? High r^2 means shit in statistics. 67.34.59.40


 * Does it actually say that it's linear? At one point it even goes down and then back up. --Armond Warblade (talk) 11:26, 22 December 2006 (CST)


 * Someone has tried to fit it into a linear table. My stat skills are limited to sociology, but a high r^2 has always been important there.  In fact, we get excited if we get r^2 of 0.4.  I don't see why a high r^2 means crap here, but in strictly theoretical applications, it might.  Residuals, ummm... thinking.  Can't remeber.  Don't r^2 include residuals?  What is the relation between r^2 and linear regression?  I'm pretty sure a linear regression is a correlation based on residuals, in a way.  No?  Maybe not, I'm kinda foggy on this stat, I just remeber what I have to.  The rest the computer does, or I look up.--[[image:Windjammer Icon1.jpg|12px]] Windjammer 02:19, 10 January 2007 (CST)


 * The idea he is suggesting is to look for a trend in the residuals. If the regression is good they will be distributed evenly.  In defense of the idea of a linear fit, nearly everything in GW is linear, and simply plotting the data looks pretty linear.  More data could be useful, especially for the rare events areas; it's likely much more accurate on the higher end of the curve, but estimating exact rates of critical hits on the low end requires many more trials.  Think about it, there is 1 recorded critical for an attribute of 0; had he been unlucky it would have been a 0, had he gotten even one more it would have doubled to a 2%.
 * To be fair, most will only care about crit rates from attribute 9 and up, and more specifically for 12+. The fact that a single critical hit difference on the lower end greatly swings the value makes it hard to trust the bottom end of the curve.  And of course, he is wrong, R^2 means a great deal, given that it is the measure of the strength of a fit.  --[[image:Epinephrine.jpg]] ~ Epinephrine 09:59, 24 February 2007 (CST)