Guild Wars Wiki talk:Projects/Armor rating

Interesting project. I'm probably to lazy, and don't play enough to help, but I'll check your progress. I have some ideas about some of the values. I'll add values I think are compound in square brackets. (check edits to see what you think). ANother question: When you find that the value is off with exactly 16, why not assume that this is due to a shield? Backsword 08:56, 26 December 2009 (UTC)

Avalanche: I'm not aware of any caster enemy that can be seen carrying a shield, although granted you couldn't see it either way. You're right though, it must have a sword/axe and shield combo since it attacks in melee. I didn't mark it as "assumes it has a shield" because (and I suppose I'm being pedantic here) we're not "supposed" to think that a caster enemy would use a shield, and anyone reading it may not bother checking what a superscript 1 refers to.

Agari Nahualli: You should be able to see their shield if they have one. Granted, there are oddities like Nuno who has a wand and apparently carries an invisible focus.

Ridgeback Skale: Actually has 16 less armor than it should on top of missing a shield.

I like the square brackets. Makes it a bit easier to see that, for example, cracked mesas have global +8 armor, but are also missing their +20 vs physical warrior bonus. Manifold 16:59, 26 December 2009 (UTC)

Hm, yeah. One thing about Shields: I'v seen reports that creatures with random shields will sometimes spawn without them. In cases where equipment models are not visible, this could lead to getting different results when testing against different induviduals.

Will keep an eye on it. Backsword 13:34, 28 December 2009 (UTC)

Suspicious Professions

Creature	Skills	Suspected
Skree Singer
Dragon Moss
Stone Crusher
Mutant Ooze

Some of these resultsmakes me wonder if their profession might not match their skills. If they are easily farmed, killing them repeatedly in HM for a tome would be a viable test. Backsword 12:09, 28 December 2009 (UTC)

Testing suggests that Scarab Nest Builders and Rock-Eater Scarabs are actually some sort of caster. --Knux 10:40, 25 January 2010 (UTC)

Concerns

-1s

I think it's somewhat unlikely that so many enemies really have one less armor rating than they should. I've been trying to find a fault in my methods, but haven't come up with anything. It comes out the same with low (38) and high (89) damage skills. It stays constant among physical and elemental damage, including on Warriors and Rangers. Looking at the enemies we have:

• level 20 Warrior with shield: 4 (96/116AR)
• level 20 Warrior with no shield: 1 (80/100AR)
• level 22 Warrior with shield: 1 (102/122AR)
• level 23 Warrior with shield: 1 (105/125AR)
• level 24 Warrior with no shield: 5 (92/112AR)
• level 28 Warrior with shield: 1 (120/140AR)
• level 28 Warrior with no shield: 5 (104/124AR)
• level 24 Ranger: 2 (82/112AR)
• level 24 caster: 17 (72AR)
• level 28 caster: 11 (84AR)
• level 20 Para with shield: 2 (96AR)
• level 24 Para with no shield: 2 (92AR)

No Assassins or Dervishes.

Very large damage skills may have an issue related to this. Sandstorm Crags seems to have a "global" +30 to armor rating, while being effected by the -1 issue, ending up with +29. However, Junundu Siege, which is listed as 400 earth damage, deals 200 damage to them. This is +28 from their expected armor rating.

I went to test against a Modniir Berseker, one of the enemies in your list who has the -1 armor rating. In Normal Mode, that's a level 20 warrior who's wearing a shield, so expected armor would be 96:

3 x 20 (level) + 20 (armor bonus since they're warriors) + 16 (shield)

I went with four Fire Magic skills, in order to test different levels of initial damage. The results I had were:

Initial damage	Final Damage	Armor rating
50	28	93 or 94
65	36	94
100	55	94 or 95
120	67	94

...Which gives 94 as the armor rating, -2 when compared with the expected value of 96. I thought that was due to the shields - armor may be at a fixed value, but the weapons the monsters have are not always the same. Originally Guild Wars had an "what you see is what you get" system in which enemies only dropped the items they had, so you would only get as a drop the exact item that enemy was using, and we could see that those items were not always the same, and often not maxed, be it weapons or shields or foci. This system isn't exactly the same now, but it still exists to a degree (a Sorrow's Furnace boss will only drop his green weapon if he's holding it before being defeated, for example).

I believe, then, that the armor rating of the shields is not always the same. I also believe some casters are actually using a wand and a shield (but still a warrior shield, so it would have a very low armor rating due to the lack of the proper attribute). In order to test this, I have used a 120 damage Fire Magic skill against two Modniir Bersekers together, and the result was different damage values, which IMO show how the armor ratings were different probably thanks to different shields.

I think it explains why most enemies with -1 values are either warriors or paragons, and why dervishes and assassins, who always use two handed weapons, don't have strange armor values. Erasculio 10:19, 30 December 2009 (UTC)

That's a good point. There's also that gw stores damage as integers between caculations, so I would expect some systemic errors anyway. Backsword 12:20, 30 December 2009 (UTC)

Very interesting test with the Modniir. I generally hit something several times, in case there's some effect I'm not aware of messing with the results, and I recall a couple times where one enemy seemed slightly different from the others. Star Blades especially seem to vary quite often. Perhaps Yeti (ranger)s are really using a one-handed bow or a special wand? They only use one hand in their attack animation and have no bow attack skills. Doesn't explain Avicara Fierce, though.

It seemed like every level 24 caster had the -1 issue, but I just retested Saltspray Dragons and they are taking 72 damage from an 89 damage Stoning, which correlates to the expected 72 armor, so it doesn't seem to be a blanket effect. There are few enemies wielding a visible focus, perhaps that has something to do with it. Manifold 16:20, 30 December 2009 (UTC)

Shing Jea and Istan

I haven't done much testing in these areas, but from what I have, it seems that most of the enemies have their armor ratings tinkered with. Some of these alterations are very small, all 6 or so casters I tested on Shing Jea seemed to have +1 armor. Troublesomely, one of these was an Afflicted Mesmer, which had 0 change from expected on the mainland.

I'd like to be sure there isn't some issue with relatively big or small damage involved

I just tested a level 12 Wind Rider, which had normal, expected armor ratings when I tested it at level 20. It had +1 armor. It seems this and the -1 issue are opposite sides of the same coin. Higher armor or level creatures may lose 1 armor, and lower armor or level creatures may gain 1. Whether this would have happened if I was level 12 as well, I don't know. To keep things from getting multiple results, I'm adding a note about testing enemies that appear at multiple levels at the level closest to 20. Manifold 17:42, 31 December 2009 (UTC)

Hard mode

I've tested only a few enemies in hard mode. Most scale up as expected. However, Shing Jea and Istan enemies seem to regain their "expected" armor, those weird +5s and such go to 0.

Azure Shadows' piercing armor rating (and perhaps other damage types) doesn't change at all in hard mode compared to normal mode, despite being 4 levels higher and being expected to gain 12 armor.

Mantid Drones' (Kaineng City) and Mantid Monitors' armor rating goes from the peculiar +17 in normal mode to +24 in hard mode (that may not be the exact number, but it was something like that).

I might go to Guild_Wars_Wiki:Requests_for_research with these questions, but maybe I've overlooked something obvious. Manifold 22:01, 28 December 2009 (UTC)

Those Mantids have +26 armor rating in Hard Mode. Mandragor Smoke Devils have -17 armor versus earth instead of -24. Manifold 23:53, 28 December 2009 (UTC)

Mm. Take a look at Spirit for a clear case of HM mattering. Backsword 12:21, 30 December 2009 (UTC)

Nice job!

This is a great project, one that hopefully we would be able to add to the creatures' pages. You have also done a lot of research very quickly, nice work. I think it would be a bit easier for more people to help, though, if you had listed all creatures in the project page, as opposed to only the creatures you had some data for; that would make it easier for other users to see which creatures still need to be tested.
I have the feeling we could also use this project to estimate the total HP of the creatures, given how one could do it after dealing damage (as this project requires, anyway) and then taking a screenshot of the monster's health bar. Erasculio 23:36, 29 December 2009 (UTC)

Adding enemies with no data is a good idea. It was my impression that HP increased the same way for NPCs based on their level as it did for PCs, with the exception of certain bosses and boss-like creatures. I would like to add this data to the individual pages, but I have a few worries, as explained above that I'd like to get nailed down first. Manifold 01:58, 30 December 2009 (UTC)

Yeah, Health works the same: base decided by level, but can then be modified. My impression was that this is not done as much as with AR. Backsword 12:16, 30 December 2009 (UTC)

I left out some enemies that only appear during one obscure quest, and the "unique stuff" category is being handled on a case-by-case basis, but otherwise think I have every enemy listed now. Manifold 05:26, 4 January 2010 (UTC)

Except Titans, I just noticed. Tomorrow, tomorrow. Manifold 05:29, 4 January 2010 (UTC)

And Destroyers. Don't worry, I'll add the destroyers later myself. Erasculio 09:43, 4 January 2010 (UTC)

I forgot a bunch of humans too. I'd like to get the carriers in JQ too, come to think of it. Oh, and I think some demons and maybe some undead need to be added too. Manifold 16:55, 4 January 2010 (UTC)

I'm wondering, what do you think about also listing each armor value, instead of only listing the modifier (say, 79, 60, 100 together with the -1, 0, +20)? If we listed the armor value (which we have to calculate anyway), IMO it would make it easier to eventually implement this information in the articles about each enemy. Erasculio 01:48, 5 January 2010 (UTC)

I've been pondering making some creature page mock-ups with armor data, some with "absolute" armor, some with "relative" armor. If people would rather have relative armor displayed on the pages then it's just more work to write them all down. If you or anyone else wants to add them, though, I don't mind. Manifold 04:08, 5 January 2010 (UTC)

By "relative" do you mean values like 100, 96, 60, etc? I like those values more (not in the tables here, but rather if this data is moved to the creature articles) as I think it would be easier for people to understand if we said "this creature has 84 armor" than if we said "this creature has +4 when compared to a warrior of this level without a shield). I'll begin adding those values later today, then. Erasculio 10:54, 5 January 2010 (UTC)

By relative I mean "+/-xx" and by absolute I mean "xx". I'm not sure how you're planning to do it, but it would be easiest for me if you wrote it after the creature's name, a bit less convenient if it was a column, and a lot less if it was added into every cell. Manifold 16:39, 5 January 2010 (UTC)

Never mind, then, it won't be so hard to add that information when we move this data to the creatures' pages. Adding it now would only be a convenience for later, but if that would create an inconvenience to you now, it's not really worth it. Erasculio 17:27, 5 January 2010 (UTC)

Chaos damage

That seems to be excellent way to test "global armor bonuses". What skill are you using, Backsword? Manifold 16:27, 30 December 2009 (UTC)

Seeping Wound. Justbecuase it is effectivly testing base AR. However, it has a major issue in that the base damage of 16 means results are crude, and willbe the same for a range of ARs. Therefor, I only used itto test values with a clear expectation and assumed it to be correct if it matched.

One thing I did to increase fidelity was to hit the same target again, this time with it beingg under Cracked Armor, in order toprovde a second reading. Backsword 14:22, 4 January 2010 (UTC)

isn't putting for all <damage type> AR (<damage type> AR = 110... (for example)) instead of giving base AR=90.. and <damage type> AR offsets +/- XX , less accurate and/or leading to loss of information ? Elephant 22:22, 5 February 2010 (UTC)

talking about computed and newly displayed data on creature pages. Elephant 22:23, 5 February 2010 (UTC)

I'm not sure how absolute armor ratings (versus relative) would be less accurate or lead to a loss of information. I record how much damage I do to various creatures with various damage types, see how much armor rating that damage corresponds to, and then compare that to how much armor rating is expected for that creature. When I add the data to creature's pages I figure how much armor rating is expected of them, and then add or subtract the relative numbers on this page. If there's multiple levels of the creature possible I note which level the data is for. Manifold 23:43, 5 February 2010 (UTC)

Creature page formatting

I've started the discussion of how to add this data to creature pages at Guild_Wars_Wiki_talk:Formatting/NPCs. Manifold 03:52, 9 January 2010 (UTC)

Pre-Searing

I'm going to try to test the Pre-Searing enemies, since I'm farming them anyway and have nothing to do, but how should I go about trying for damage when I don't have a weapon that always deals the same amount of damage?--Hawkins 19:52, 3 February 2010 (UTC)

Things get more complicated with a non-level 20 character using a variable damage weapon, so I never looked too closely into it. Damage_calculation#Weapon_Damage could help, although I'd advise you to double check anything you read there, as I couldn't get the formulas to work for myself. Manifold 20:17, 3 February 2010 (UTC)

No need to test for some foes if the Prima strategy guide still holds true. I can add these if ANet confirms it is still the same for these foes as that era.Yumiko ^,~ 05:32, 15 May 2011 (UTC)

Data Table Suggestion

First let me say that this is an incredibly ambitious project; lots of props to Manifold. Well done. The data tables on the main project page are just begging to be sorted in ascending/descending order. It would be nice if someone with Wikiskills could add sorting functionality to the tables while still maintaining the current view as default. Keep up the good work. -Ninjatek 03:58, 13 February 2010 (UTC)

Done, if I understood you correctly. Manifold 04:37, 13 February 2010 (UTC)

That's exactly what I had in mind; very convenient. Thanks again! -Ninjatek 21:24, 13 February 2010 (UTC)

Portal Wraith

Is the armor rating information in the Notes section of Portal Wraith correct? --Silver Edge 09:47, 18 March 2010 (UTC)

Well, they have less elemental armor than expected and more physical than expected, but if you average it all out they come out below expected. The data there is for at least the Hell's Precipice version, it's possible but unlikely the Gate of Madness version is different. I haven't tested either in Hard mode, but from what I've seen so far of Hard mode armor ratings, their elemental armor is probably pretty close to 60, and their physical would probably be greater. I'll do some testing, but I'd remove the note. Manifold 15:57, 18 March 2010 (UTC)

Status of the project

I'll probably fill in a few gaps here and there eventually but I'm pretty much done with this project. When new creatures are introduced I'll do my best to record them, but otherwise there's few creatures left, and almost all of them are a pain to record and not very important. HM versions seem to hold few surprises, so I'm not too keen on going after them all. Creatures from The Deep, Urgoz's, parts of DoA, etc don't have many entries due to requiring humans to reach them. There's quite a few unrecorded humans, but they never have specific weaknesses or strengths and rarely have overall bonuses.

Everyone is welcome to add their findings, of course. Manifold 17:13, 4 May 2010 (UTC)

Holy Damage

Shouldn't there be a spot for holy damage in the table?
I mean, Smiters would probably like to know what they'd have trouble with(...Ice imps and Ice Golems in Mineral Springs come to mind), and what would be a breeze(undead of course)... - Stimpson J Kat 17:35, 13 June 2010 (UTC)

Undead (with a couple exceptions) take double. Everything else takes normal damage. Manifold 17:51, 13 June 2010 (UTC)

Would it be too much to ask for the data specifically because of those exceptions? If it doesn't seem worth it, so be it. You put a lot of effort into this already, it seems, and we appreciate that. Demanding more or complaining are not my intentions. 76.106.245.213 19:13, 26 July 2011 (UTC)

I'm pretty sure everything that takes double holy damage is documented, if not well known. I wouldn't be opposed to adding a holy stat to Template:NPC_statistics, a bot could add "x1" to everything that's not undead, and then the undead could be manually gone through and "x2" added for the relevant creatures. It really wouldn't be much work at all. I'm just not sure how valuable it would be for the few cases where it might be counter-intuitive, namely ghosts, carven effigies, and ancient statues Manifold 02:55, 28 July 2011 (UTC)

That's the question. I don't want to add to it for my personal sake. I'm curious about the value of completeness versus the efficiency of conciseness. This could be a one-stop guide for armor ratings versus all damage types for all creature types, or it could be a cleaner answer to those ratings that weren't previously documented. As the creator, it's obviously your call. Availability vs redundancy/necessity, perhaps. 76.106.245.213 14:10, 28 July 2011 (UTC)

Aloe Husk

According to the old prima guide it has -40 armor to fire and slashing damage and its armor rating is 3. It is a monk. Any way to confirm it? I put the armor modifiers on the main project page anyways.Yumiko ^,~ 06:15, 15 May 2011 (UTC)

I had no idea the Prima guide had armor ratings! Is it just a few creatures or what? 3 armor on a level 0 would be consistent other level 0s like Corsair Mage. I unfortunately don't have access to a pre-searing character, nor is it easy to test physical damage in pre. Manifold 14:55, 15 May 2011 (UTC)

It does actually give the armor ratings for every creature at time of release of prophesies from beginning of Pre to Hell Precipice. But it doesn't give all notable armor modifiers like for most plants. Almost all plants seem to have this inherent fire and slashing weakness (Not sure about juggernauts, I will have to test it in Luxon FA. Any ideas?). So should aloe Husk be -37/-37 Fire/sla because of armor = 3? Furthermore, since creatures are usually not changed mechanically except for skill updates, we should expect the same armor ratings for all creatures even now. I can give you the zipped PDF link as I'll be repacking it for wiki research use.Yumiko ^,~ 17:43, 15 May 2011 (UTC)

The values in these tables are relative to expected, and expected for casters is level * 3, so 0 would be expected, and the value in the table here would be +3 or -37. (Side thought:GW2 devs have said that currently there are no strengths or weaknesses to damage types, perhaps they were added late in development of GW1 as well) I think someone would have noticed if the Jugs were vulnerable to fire damage by now, but the elements aren't too hard to test. The easiest way for the physicals is to use Wild Blow with 12 in the relevant mastery.

I would absolutely love to take a look at the guide to double check and see if anything's changed. I'm not sure about putting the armor ratings for the pre creatures on their respective pages, the guide has a lot of outdated info, and apparently doesn't note individual strengths and weaknesses against damage types. You can e-mail me via the "E-mail this user" link under the "toolbox" section while on my userpage. Manifold 19:02, 15 May 2011 (UTC)

Seem you disallowed email,also I had to find your feedback page. Just email me at hirohito6565@yahoo.comYumiko ^,~ 22:06, 15 May 2011 (UTC)

Dervish change

Do we need to re-evaluate the armor level of dervish enemies since the mysticism change? Or not since it's reliant on an enchant? --JonTheMon 21:02, 27 May 2011 (UTC)

I see it as a temporary buff given by certain skills. A dervish enemy won't necessarily be enchanted. It would be a bit like listing the attributes of an elementalist with Glyph of Elemental Power as two higher, or listing the altered armor of a mesmer using Physical Resistance. Manifold 00:01, 28 May 2011 (UTC)

Reliability of results?

It seemed unlikely to me that so many enemies have such small alterations to their armor. I tested a Stygian Hunger in HM, and it's armor is definately 100 not 101. I used 3 damage packets, piercing 76 did 38, piercing 832 did 412, and cold 42 did 21, i.e. the damage was exactly halved. I have not tested any others yet, but it kind of makes me doubt all the results :/ -- 94.174.41.204

You happened to choose three even numbers, and 100 armor seems to be a special case. See this discussion. Manifold 15:48, 6 December 2011 (UTC)

the large damage packet distinguishes, 832 dmg @ AL 101 would give 408.8448 rounded to 409 dmg, while what actually happens is you deal 412 damage. I think that "Armor rating can change a bit, depending on how much damage is done" seems very unlikely, with the link you posted looking at the attribute line for Mantra of Earth, it seems possible that the dmg reduction is not exactly 50% but slightly less (e.g. 49.9998%), which then makes it round up to 59 damage. It just makes no sense that the enemies would be individually coded and given armor changes of +1 or 2 and so on, and the methods of testing are not totally accurate. The armor information is probably stored in the memory somewhere, which would be a far easier and more reliable method -- 94.174.41.204

I'm not arguing that the enemies most likely have "nice" armor values coded, but it was decided it would be best and make the most sense to display the armor rating they essentially have during play. If it was "far easier" to just look up all the values directly, I'm sure someone would have done it by now. Manifold 18:09, 6 December 2011 (UTC)

The armor values that are "coded" and the ones that are "during play" are not 2 separate concepts, they are precisely the same thing. I didn't look around enough to make a well informed view right now, but if as you say there seem to be small discrepancies and "special cases" then that means the understanding of the mechanic is incomplete, perhaps there is something else at play such as inherent dmg reduction or the rounding & multiplying mechanics are different to what we think. For example, I know that if you have 5pips of health regen, the game does not calculate that regen as 5 * 1 pip, but as 1pip+1pip+1pip+1pip+1pip, which due to the rounding nature of floats can give a slightly different value. -- 94.174.41.204

Hard mode armor rating

The leaked balance update states "Reduced the armor level for Hard Mode enemies", so this project page and the individual NPC pages that list hard mode armor ratings will be affected by this future update. --Silver Edge 20:39, 9 December 2011 (UTC)

Yeah, I suppose I'll have to wipe all of the HM values. Glad I didn't do more than Canthan enemies for the most part. It will be interesting to see how it's done, since just "10 less armor than they had before" would make the HM versions have less armor than their NM counterparts for creatures that gain less than 4 levels. For the most part. Manifold 21:01, 9 December 2011 (UTC)

Preliminary data after the HM armor: The Wardens I tested (Tree, Trunk, Leaf, Mind, Spirit, Summer) have the armor of a level 20 in HM, except for Tree. Stone Scale Kirin have 69 armor rating (previously 78), normal for their NM level of 23. Dragon Moss have 82 armor now, the same as their NM armor. Mantid Monitors and Mantid Drones have 60 armor. Blood Drinkers have 60 armor.

I think the rule is: If the level is below 20 in NM, then it has level 20 armor in HM. If the level is above 20 in NM, then armor is the same in HM as NM. Manifold 00:09, 6 January 2012 (UTC)

Nice (and quick) research; thanks. Did you manage to guess at the increased health rules while you were at it? – Tennessee Ernie Ford (TEF) 00:19, 6 January 2012 (UTC)

I haven't looked into it, want to do some 4 ele vanquishes first. Manifold 00:27, 6 January 2012 (UTC)

I just bet you do. :-) (I do hope the Hero AI is up to the task.) – Tennessee Ernie Ford (TEF) 00:55, 6 January 2012 (UTC)

Doppleganger seems to be the only found exception so far. Stone Elementals lose their cold bonus, though. Dragon Moss kept their "global" +10 armor boost, so perhaps all specific type vulnerabilities/strengths are wiped clean in HM? It would be a shame. Manifold 03:32, 6 January 2012 (UTC)

Forgot that Dragon Moss still took extra fire damage. Didn't specifically check slashing. Manifold 04:32, 6 January 2012 (UTC)