The other things you point out in your post - mainly Sunder vs FF/Mangle, Thunderclap and Demoralizing Shout - are irrelevant to bring up if you're comparing tanks on *most fights* because any raid worth anything will have a Warrior put up Sunders (or a Rogue put up imp. EW) and a Druid put up FF/Mangle. Same goes with Thunderclap and Demoralizing Shout: they'll be put up in their improved states in any semi-serious raid (not always 5/5 demo, sometimes 2/5).
Sure, you could do all of that. No question. The point is that it isn't optimal and it really limits your flexibility as a raid group.
Another warrior needs to tclap. Ok, who's doing that? The MS/slam warrior who is also providing your demo shout? How complicated is his dps cycle now that he needs to stance dance or how much dps did you just cost him because he can't do this in zerker stance? What happens to his cycle if he gets a resist? Is it a fury warrior? Why are you bringing a fury warrior over another rogue (higher dps) or a ret paladin (high dps and lots of buff/debuff raid synergy benefits)? Is it a protection warrior? Why is he dps'ing when the feral druid does as much or more dps and can provide buffs to the raid (battle rez, innervate) if he isn't tanking?
Imp expose armor eliminates rupture from a rogue's dps cycle. It also takes points away from lethality or poison talents for the rogue. For a well-geared Sunwell rogue that's ~120-150 personal dps they lose for 475 less net armor on the boss. Assuming you're running physical dps heavy, that's ~3% more damage and might be a small net gain, but the rogue has to be able to melee that mob to keep it up. Tanks usually have a much easier time staying in harms way than rogues do.
The main reason warriors tank encounters is that they have comparable overall mitigation to a druid (magic damage is extremely common in most boss encounters) and they have the lowest opportunity cost while providing the most benefits. You can certainly gain a small threat/mitigation benefit by having a druid tank (and that will be optimal some of the time), but it's not usually optimal.
I recently went from no-macro tanking (save for a focus-set devastate for tanking two) to working with several macros. these are two of the macros I wrote:
The idea being to increase the amount of time that Shield Block is up. I also left Shield Block on its own in case I need to choose between the two. The last line prevents me from seeing "not ready" warnings when Shield Block is still on cool down, but I do see warnings when there isn't enough rage for both actions to be taken. I'm pleased with the results, since Shield Block is up nearly 100% of the time in endless rage situations (when not on cool down), and up much more frequently during trash pulls.
Question: By embedding Shield Block in Revenge and Devastate am I setting myself up for problems in later instances such as Kara, ZA, and 25-Mans?
On bosses where I need to keep Shield Block up because a crush could be lethal, this macro is ideal. They have the same cooldown so it's one less thing to worry about. I *only* use this macro for that; everything else, I take it off my main bar. I put Shield Block first so that it always uses it first incase I don't have enough rage for both. I also still have Shield Block and Revenge on my bar if I get rage starved and don't want to use both at any given time. However 99% of the time any fight where a crush is lethal, I have enough already because of how hard the boss hits that this won't be an issue.
Veneratio went over such a macro in one of his tankingtips.com podcasts.
I just got an infraction for not checking something on the tanking spreadsheet. I didn't realize there was one or I would have already gotten it, does someone have a link? I did a search and didn't find a thread.
It seems to be pretty good for survival and mitigation choices, but seems to be out of whack in the threat gear choices, based on a few hours of playing with it. Definitely good for sandboxing some upgrades though.
There's not some hidden "but he tries really hard" variable built into the game. -Slake
I always love the "it doesn't fit my style of play" line. There are only two styles of play; Correct, and Incorrect. The only people that ever use this line are people with the incorrect style of play. -Sebudai
It seems to be pretty good for survival and mitigation choices, but seems to be out of whack in the threat gear choices, based on a few hours of playing with it. Definitely good for sandboxing some upgrades though.
What seems to be wrong with the threat numbers? Too high, too low? Keep in mind the threat value in the gear chart is an average of max threat and limited threat (max threat is heroic strike every swing, limited threat is no heroic strikes). A couple of buffs I haven't added are blood lust and drums for threat. Also it doesn't model healing threat (prayer of mending, earth shield, etc). So your max threat could be higher than rawr says, and of course it's only doing average threat, so during a normal fight you might get more crits than on average, might have less attacks parried, might get more rage from taking damage. I'd love to make the model as accurate as possible so if you have suggestions on the threat side of things, feel free to post them.
What seems to be wrong with the threat numbers? Too high, too low? Keep in mind the threat value in the gear chart is an average of max threat and limited threat (max threat is heroic strike every swing, limited threat is no heroic strikes). A couple of buffs I haven't added are blood lust and drums for threat. Also it doesn't model healing threat (prayer of mending, earth shield, etc). So your max threat could be higher than rawr says, and of course it's only doing average threat, so during a normal fight you might get more crits than on average, might have less attacks parried, might get more rage from taking damage. I'd love to make the model as accurate as possible so if you have suggestions on the threat side of things, feel free to post them.
The numbers for each item look good, but the Optimizer for Threat puts out results that no tank would end up wearing. I'd recommend putting an option to set levels for minimum defense, armor and health so that you don't just get back a full dps/pvp set as a result. Also, the Main Hand enchants need to have Executioner and Potency (+20 Str) added to it, and +35 Agi taken out.
There's not some hidden "but he tries really hard" variable built into the game. -Slake
I always love the "it doesn't fit my style of play" line. There are only two styles of play; Correct, and Incorrect. The only people that ever use this line are people with the incorrect style of play. -Sebudai
Well if you only optimize for threat it will be a mostly dps gear set. Add an additional requirement when using the optimizer of survival, mitigation, chance to be crit, chance to be crushed, and/or health minimums you want while still optimizing for threat. Or increase the threat scale on the options tab and then optimize for overall rating.
Unfortunately I cannot remove 35 agi from the main hand enchant at this time. The main hand enchant slot is shared with 2h weapons. I'll add Executioner and Potency for the next release.
I was curious about 'Rawr' so I downloaded it and tried it out. I put in all of my threat gear stats (or as many as I could, anyway, it doesn't have Steely Naaru Sliver or the Gnomeregan Auto Blocker...), and checked all of the buffs that I had in my stacked group parses. The projected unlimited-rage output was showing 1388, which was about 700-800 threat off the realistic amount. Heck, 1388 tps is easily achieved in a naked group with non-threat gear on, so something's really off with the estimated threat output.
I'd love to see some WWS parses for warriors who can do 2100+ personal TPS. Keep in mind rawr doesn't account for healing threat or misdirection threat so if you are comparing rawr to omen in game rawr will be lower. If you look at the highest Gorefiend TPS on Coolyo Coolyo[dot]org they have 2040 TPS. Almost 200 TPS of that was from healing. So 1850 personal TPS. Rawr doesn't account for cooldown stacking, however the auto blocker is modeled, you might have to add it through the menu. It assumes average block value, while in game you can stack it and get 2 uses during a 2.5 minute fight. Also bloodlust/drums/executioner aren't modeled.
The highest TPS on coolyo which still has an active WWS is Coolyo[dot]org at 1799 TPS. 250 of that was from healing. He used two threat trinkets, Shard of Contempt and Dragonspine Trophy. He had a bloodlust, drums, 2 trinket uses in a 3 minute fight, and executioner all which aren't modeled yet in Rawr.ProtWarr. He also got ~25% crit without a feral which means he got lucky on crits or was wearing a lot more dps gear than I normally wear.
I've taken the threat numbers directly from this thread and from omen. So in theory over a long enough fight and ignoring healing + misdirect and using gear/enchants that are modeled it should be very close. I am working on adding the other options in (BL, drums, executioner) so hopefully will have those for the next release. If you have WWS links and what exact gear they were using I'd love to analyze Rawr.ProtWarr's threat numbers to make it more accurate.
I'd love to see some WWS parses for warriors who can do 2100+ personal TPS. Keep in mind rawr doesn't account for healing threat or misdirection threat so if you are comparing rawr to omen in game rawr will be lower. If you look at the highest Gorefiend TPS on Coolyo Coolyo[dot]org they have 2040 TPS. Almost 200 TPS of that was from healing. So 1850 personal TPS. Rawr doesn't account for cooldown stacking, however the auto blocker is modeled, you might have to add it through the menu. It assumes average block value, while in game you can stack it and get 2 uses during a 2.5 minute fight. Also bloodlust/drums/executioner aren't modeled.
The highest TPS on coolyo which still has an active WWS is Coolyo[dot]org at 1799 TPS. 250 of that was from healing. He used two threat trinkets, Shard of Contempt and Dragonspine Trophy. He had a bloodlust, drums, 2 trinket uses in a 3 minute fight, and executioner all which aren't modeled yet in Rawr.ProtWarr. He also got ~25% crit without a feral which means he got lucky on crits or was wearing a lot more dps gear than I normally wear.
I've taken the threat numbers directly from this thread and from omen. So in theory over a long enough fight and ignoring healing + misdirect and using gear/enchants that are modeled it should be very close. I am working on adding the other options in (BL, drums, executioner) so hopefully will have those for the next release. If you have WWS links and what exact gear they were using I'd love to analyze Rawr.ProtWarr's threat numbers to make it more accurate.
Ok, here's a log I have that expires in 3 days, I may be able to re-upload it, not sure I saved it. This was done to refute someone's claim that greater than 1200 tps without mega group buffs is extremely hard/impossible, etc. Anyways -
Xavastrasz produced 1528 tps is the link where you can see the threat without healing accounted for. It's showing 1344 tps, which is 'low'. However, that's with no group buffs or heroisms/bloodlusts in any way, as I was (literally) in a group by myself.
Rawr is claiming I'd generate approximately the same TPS as that, but while being in an extremely stacked group (I checked everything possible, pretty much.) [also note the extremely low crit rate on my parse, much lower than even my character sheet, and we had imp. jotc too]
Currently the only way to get talents into Rawr.ProtWarr is to import from armory. After that you can modify them, but if you don't import initially you will have no talents. So make sure you import your toon then change his gear. You can change the talent spec after importing if needed.
I'm assuming you had (gotten from wws and some assumptions): imp mark, regular battle shout, imp BoM, Kings, Agi food, regular Faerie Fire, CoR, Imp Hunters mark (not sure on this one, but you had expose weakness which isn't modeled in ProtWarr yet so it should at least even out), Imp Judgement of the Crusader, Blood Frenzy. With all those buffs, Rawr.ProtWarr says your unlimited threat is 1428 TPS. That is without a mainhand enchant since executioner isn't modeled in the release.
Healing threat, thorns, rage gains, Felsteel Shield Spike all are not modeled in ProtWarr, and honestly it's low priority since it's a small amount of threat or not controlled by you and depends on the boss attacking you. So if I uncheck all those from your TPS report I get 1324 TPS. So thats 100 TPS lower than Rawr says is your unlimited threat. It seems pretty close to me on what Rawr models. This does assume you were using the same spec you currently have (8/7/46).
If I stack your group in Rawr with an enhancement shaman using totem twisting, using agi elixir, a feral, and BM hunter, I get 1740 TPS unlimited. Add in another 250 TPS from healing aggro, and another 100 TPS from misdirect threat and we get to the magic 2100 TPS number. So the numbers seem pretty close to me.
Edit:
If you compare the threat per skill from Rawr and your TPS report
Your numbers:
HS: 535.6
SS: 358.6
DEV: 249.5
REV: 119.7
AUTO: 60.4
So shield slam threat is within 10 TPS, revenge threat is within 15 TPS, devastate threat is within 20 TPS, and your HS threat is about 150 TPS less than the unlimited value from rawr, but you had 60 TPS from white attacks because every swing wasn't a HS.
Last edited by ezweb : 06/17/08 at 12:48 PM.
Reason: Add threat number comparisions
Okay, my mistake, I think what you said about the talents, I didn't notice. I think I imported my character, who was PvP specced for the weekend, and then I forgot to check to see how that carried over / was modified. Thus it probably was showing threat numbers without me even having Devastate or something, which is really interesting.
I haven't re-ran it since respeccing back to prot, but you did, and I'll take your word for it. That looks right. Side note, though, no parsing programs (obviously) take Misdirect into consideration.
Anyway, sorry for the derail, I'll fiddle around with Rawr more in the future now that I know it appears to be more right than wrong after all.
Another thing on my todo list is not allowing threat cycles for talents that the current character doesn't have. Right now it assumes you have shield slam and devastate. It also only supports one threat cycle. It's far from complete but for most tanks it should work very well.
Imp expose armor eliminates rupture from a rogue's dps cycle. It also takes points away from lethality or poison talents for the rogue. For a well-geared Sunwell rogue that's ~120-150 personal dps they lose for 475 less net armor on the boss. Assuming you're running physical dps heavy, that's ~3% more damage and might be a small net gain, but the rogue has to be able to melee that mob to keep it up. Tanks usually have a much easier time staying in harms way than rogues do.
How (ignoring the whole devastate/sunder threat issue) any different from a fury warrior losing far more than 120-150 dps for a 4% dps increase in physical dps by changing spec/rotation?
In "Custom>Relative Stat Values," there are mostly defensive stats, including agility and block value. Is there a way to add hit, expertise, haste, strength, and attack power? To some degree, you can already see some of this info by going to the normal gems section and looking at only pure stat gems (this will add strength/ap, hit, and crit).
One question about threat. Either there is an error or I have incorrect preconceptions about relative stat values. My example is in the cloaks section. Pepe's has a threat value of 1188 where the badge block value cloak has a value of 1614. I would have assumed that 25 hit was better than 38 block value. I do have +block value meta & talent. I'm not sure how the model works exactly. Does it take into account lost rage from avoidance and armor? This set has 6% +hit, so I don't think I'm going over the cap. Is block value really that good?
Not really specific to the ProtWarr model, but I really like how easily you can add new items. This is one of the things that is often tricky (particularly if you're unfamiliar with the language or spreadsheet software) in other tools/sheets.
The threat stats will be in the Relative Stat Value chart in the next release. I have not modeled rage at all. Meaning if you make a set where you are immune to melee your unlimited threat will still be the same as if you had zero avoidance/armor. I haven't decided how or when I'll get that modeled.
38 block value * 1.3 * 1.1 = 54 block value which is 54 * 1.3 * 1.15 * 1.02 = 82 threat / 6 secs = 13.7 TPS assuming no missed/dodged/parried and no armor. So block value is not an insignificant amount of threat when you think of all the scaling it can get. Feel free to download the source code and check my math.
My warrior is an alt with T4ish gear and a few SSC/TK items, besides the good tanking stuff you can get for badges. He will probably never loot MH/BT gear since we're quite abundant on tanks atm, and lack healers alot (my main). I'm looking to optmimize my aggro for some heroics, Kara and ZA cause I sometimes tank it for my T6 geared guildies, and there's where the problems start: I'm having great difficulty maintaining the often 1500 TPS (sometimes more) our warlocks and mages can do (sustained sadly) on trash or even bosses. I realize I will never match a full T6 tank in terms of threat (with equal weapon and shield), but am trying to improve at least what I can. I'm sitting over 15.5k unbuffed so HP shouldn't be the problem when I am toying around with a threat set, question is, which trinkets will give me the best TPS for my gear lvl? Im currently sitting on 80ish hit, 22 expertise.
I found the Autoblocker a nobrainer for so long, am I still right in using it always? Then for my second trinket, I'm in doubt between Shard of Contempt and another hit trinket (Romulo's or the one from BF heroic giving armor penetration on use). I'm already sitting at 5.50% dodge/parry removed so will that last, extra % of both give me better aggro than another 30 hit? I've been trying to swap them around a few heroics I tanked but since mob armor values seem so different I really don't think that's good way to compare them - at least I'm still clueless
My warrior is an alt with T4ish gear and a few SSC/TK items, besides the good tanking stuff you can get for badges. He will probably never loot MH/BT gear since we're quite abundant on tanks atm, and lack healers alot (my main). I'm looking to optmimize my aggro for some heroics, Kara and ZA cause I sometimes tank it for my T6 geared guildies, and there's where the problems start: I'm having great difficulty maintaining the often 1500 TPS (sometimes more) our warlocks and mages can do (sustained sadly) on trash or even bosses. I realize I will never match a full T6 tank in terms of threat (with equal weapon and shield), but am trying to improve at least what I can. I'm sitting over 15.5k unbuffed so HP shouldn't be the problem when I am toying around with a threat set, question is, which trinkets will give me the best TPS for my gear lvl? Im currently sitting on 80ish hit, 22 expertise.
I found the Autoblocker a nobrainer for so long, am I still right in using it always? Then for my second trinket, I'm in doubt between Shard of Contempt and another hit trinket (Romulo's or the one from BF heroic giving armor penetration on use). I'm already sitting at 5.50% dodge/parry removed so will that last, extra % of both give me better aggro than another 30 hit? I've been trying to swap them around a few heroics I tanked but since mob armor values seem so different I really don't think that's good way to compare them - at least I'm still clueless
Well, even a T6 tank isn't maintaining 1500 tps in every fight, especially in Kara or ZA where we can't queue up heroic strike on every attack.
You don't have a gear issue, it's a raid strategy issue. Your dps needs blessing of salvation, it will make their damage more threat efficient. They also probably need to be using their threat dumps (invis, soul shatter, vanish, feign) more frequently. Most well-geared T6 mages aren't pushing 1500 sustained tps (pre 20% mob health) with salvation, if they use their invis/soul shatter later in the fight, they should have an unlimited threat ceiling when a boss hits 20%.
If you're just looking for suggestions of good pieces of threat gear, Quignon has a bunch of them listed earlier in the the thread. When you're looking at certain pieces you need to ask yourself if you're having a problem with boss threat or trash threat. Trash has a much lower hit and expertise cap than bosses do, so your gear emphasis will be different depending on which one you're trying to improve. Until you're at ~6% reduction to parry/dodge you are always better off stacking more expertise, after that point, you need to start looking at what you're tanking.
The autoblocker is probably one of the best threat trinkets in the game. Shard of contempt, romulo's poison vial and madness of the betrayer are also very good (I like romulo's better than the blood furnace trinket). Shapeshifter's signet is good if you're not expertise capped on trash. Angelista's revenge is also a very good hybrid ring (~1% dodge, a little under 1% crit, 58 AP, stamina, and armor penetration).
How (ignoring the whole devastate/sunder threat issue) any different from a fury warrior losing far more than 120-150 dps for a 4% dps increase in physical dps by changing spec/rotation?
Well for starters, the bloodfrenzy debuff affects all physical attacks including those that are armor piercing like bleeds. Increasing the armor reduction only affects certain attacks. In addition, you may have people who approach 0 mob armor with their armor penetration gear, they gain nothing from imp EA, while they do gain from bloodfrenzy. They're 2 different raid dps debuffs with different mechanics. It's not a huge difference, but it's big enough to matter.
I also think you're probably overestimating the base dps difference for a well-geared fury warrior vs what an MS/Slam build can do with similar optimal gear (if you're trying to compare a warglaive fury warrior to a cat's edge warrior that's really not a fair comparison).
There are situations where the gains from bloodfrenzy are less raid dps than fury. M'uru (door dps, pre phase 2) and most 10 man's there is very little if any gain by being MS slam instead of fury.
I tested out the ProtWarr model's block value and hit threat values by adding small amounts, observing the results, and comparing to what I expected based on some quick calculations. Seems pretty much right on.
Thinking more about relative stat values for threat, I think the advantage of hit/expertise is not that it produces the most threat gain, it's more about reliability -- particularly no miss/dodge/parry strings early on in a boss fight or after an aggro drop.
Well, even a T6 tank isn't maintaining 1500 tps in every fight, especially in Kara or ZA where we can't queue up heroic strike on every attack.
You don't have a gear issue, it's a raid strategy issue. Your dps needs blessing of salvation, it will make their damage more threat efficient. They also probably need to be using their threat dumps (invis, soul shatter, vanish, feign) more frequently. Most well-geared T6 mages aren't pushing 1500 sustained tps (pre 20% mob health) with salvation, if they use their invis/soul shatter later in the fight, they should have an unlimited threat ceiling when a boss hits 20%.
If you're just looking for suggestions of good pieces of threat gear, Quignon has a bunch of them listed earlier in the the thread. When you're looking at certain pieces you need to ask yourself if you're having a problem with boss threat or trash threat. Trash has a much lower hit and expertise cap than bosses do, so your gear emphasis will be different depending on which one you're trying to improve. Until you're at ~6% reduction to parry/dodge you are always better off stacking more expertise, after that point, you need to start looking at what you're tanking.
The autoblocker is probably one of the best threat trinkets in the game. Shard of contempt, romulo's poison vial and madness of the betrayer are also very good (I like romulo's better than the blood furnace trinket). Shapeshifter's signet is good if you're not expertise capped on trash. Angelista's revenge is also a very good hybrid ring (~1% dodge, a little under 1% crit, 58 AP, stamina, and armor penetration).
Well I know they should hold back when threat goes through the roof, but we're doing it kinda for the fun of it so they can one day (hopefully) go all out. Pally buffs are usually available in raids but not always in heroics. Just trying to figure out which would benefit me the most at this time, more expertise or more hit. I'll take a look at that list thanks
Earlier today an issue came up when Felstrength legplates dropped.
The chance for you to hit the boss, as stated in the first post of the thread, is decided by one die that is being rolled a single time. What this does as is stated by the first post is making hit and expertise stack. Hit lowers the chance for your attacks to get missed, expertise lowers the chance for your attacks to get parried or dodge, so you need both for the die to get the highest chance to hit the boss.
The issue is this: With the itemization in Sunwell being heavily favoring expertise, there are few or none items with actual hit on them. I have read through this thread and seen a lot of posts why expertise is good and why hit is good, but, no posts where you compare hit and expertise to one another and when to choose the former or the latter.
06/13/08, 5:20 PM
Similar Threads
