Good Idea/Bad Idea: Vengeance

I got my start in blogging back in November, when I wrote a guest post for Rhidach’s blog, Righteous Defense.  I posted at the time about Vengeance, and some of the problems that I foresaw at the time.  The other night, Theck and I had a long conversation about Vengeance, and I think there are some interesting things to say again.

Vengeance: The Problem It Purports to Solve

Vengeance, as far as I can tell, is aimed specifically at solving one problem: It exists to keep tank damage low in PVP while keeping tank damage high in PVE.  In this regard, it does succeed, but it has a whole host of follow-on problems that I do think should be addressed.

It’s worth understanding some of the historical context of Vengeance.  In Beta there was going to be threat decay, and tank threat multipliers were all 2x damage, rather than the current 3x.  Tricks and misdirect would tide you over until Vengeance ramped up, and since DPS threat was also decaying, it didn’t really matter that tank threat production was extremely low for the first few seconds of a fight.

When threat decay was removed, there was an obvious problem with tank threat – it ramps so slowly that by the time Tricks and MD threat fades away, at a 2x threat multiplier tanks had no chance of having surpassed early DPS threat.  So we get a 3x multiplier, in order to have enough peak threat production to not only stay ahead of DPS, but to catch up the 15 second lead time DPS has on threat production.  The natural consequence of this is trivializing threat after the first 30 seconds of the fight, because tanks aren’t just “ahead” on threat, they’re so far ahead that it’s nearly impossible to fail at holding aggro.

Vengeance: The current problems.

There’s a few things going on with Vengeance right now that I personally consider to be problems, of various levels of severity.  The first is the RNG nature of the way it stacks.  With a little bit of “luck” on avoidance, tank Vengeance stays low for too long, and suddenly when tricks and MD expire, there’s several DPS over the limit.  This is something that’s extremely difficult for DPS or tanks to judge and react to, because they have to know 15 seconds in advance that the tank isn’t getting hit and isn’t GOING to get hit, in order to make good decisions.

This particular problem is liable to get worse as tank avoidance increases – the rate at which Vengeance stacks is decreased, and the reliability of it stacking becomes more volatile.  It’s a mechanic that feels like it’s completely out of our control as tanks, we have to pray to get hit “enough” for the few seconds that it matters, and we can’t afford to gear around that because the boss damage levels assume survival oriented gearing on tanks.  Catch-22.

Another problem, one that’s perhaps less important, is that the offtank can only make very small contributions while not tanking.  Without Vengeance, a tank does a nearly trivial amount of damage, on the order of 4k DPS – or under 1% of raid DPS output in a 25man.  With it, the tank does 12-15k.  It’s tough to feel like your contributions matter when you don’t have a vengeance stack.  Before Vengeance, tanks were resource constrained while not tanking, and one could maximize around that constraint to feel like you were making a contribution.  Not so anymore.  You maximize because you can, because it’s the right thing to do, even while knowing that whatever contribution you make is really lost in the RNG of crit rates from the real DPS.

None of these problems are crippling, really.  They’re things that we have to work around, they’re annoyances, and they aren’t really fun to deal with.  They disengage from gameplay and make one think about gaming the system.  That’s a pretty good sign of a bad mechanic.

Fixing Vengeance

I firmly believe that threat production should matter, and Vengeance takes away from that as it stands, and it does so in a way that I feel like I can’t control.  In order to fix Vengeance, I think it needs to be an “always on” mechanic in PVE.  Set the tank threat modifier back to 2x, and tune the Vengeance level appropriately for tanks to be at roughly 110% of DPS threat.  DPS shouldn’t have to hold back, there’s no counter-intuitive ramp-time issues, and good play as a tank becomes important again.  If you can’t run your rotation adequately, your DPS will have to hold back.  If you can, they won’t.  Doing your job becomes important.

It fixes ramp time, it fixes the RNG nature of Vengeance, and it fixes the off-tank contribution problem.  It allows tanks to gear for threat stats if their execution is poor, and feel like doing so has a meaningful impact.  The very simplest solution is just to redesign Vengeance to say “Tanks do double damage to NPCs.”  It’s not very well lore driven, but neither is toggling raid difficulty with a UI switch.  In this case, I think function is more important than form, and the current implementation has major problems with function.

I’m being slightly facetious with the suggestion above, but that version of Vengeance would work seamlessly with the PVP/PVE constraints on design, and it would get rid of all of the problems I’ve highlighted so far.  It would be a functional solution to the problem.  It might not be “elegant”, but sometimes the simplest solution has an elegance of it’s own.

Other options include Vengeance on avoidance, Vengeance with no decay while in combat, giving every tank a Vigilance-like mechanic to keep Vengeance up while off-tanking, or just giving up entirely and going back to the wrath model of having tanks be resource constrained while not tanking, and accept that tanks will do ~30% of the damage of a real DPS to keep them balanced in PVP.  That’s fine – it’s still better than the situation we have now.

Final Words

I understand that there are some design constraints that mandate a solution like Vengeance.  It does very effectively solve the PVP problem as it stands.  But tanks are primarily a PVE role, and the mechanic has serious flaws in a PVE environment.  I would like to see it fixed.  I would rather see the mechanic work well in PVE, the primary focus of a tank spec, and it’s okay if it’s a little kludgey in PVP.  But it should be a design priority for the mechanic to work well in PVE, and it currently doesn’t.

This entry was posted in Mel's Random Musings, Tanking, Theorycrafting. Bookmark the permalink.

29 Responses to Good Idea/Bad Idea: Vengeance

  1. Lynn says:

    Nice post. But what makes me sad is the fact that this should have been written one year ago when Vengeance was introduced for the first time in Cataclysm beta (and when most of the experienced tanks already realized it is a bad mechanic) and not now, 7 months after the Cataclysm release. Better late than never, I guess.

    On the other hand, I am glad you mentioned the pvp issue, as most of the tanks are not aware of this and as much as I despise Vengeance, I must say that Blizzard solved this nicely. Pre-3.2 warriors/paladins stacking tons of Block Value and two-shotting everything was stupid and I do not wish anything similar to happen again.

    I personally dislike Vengeance a lot, but I also know very well that Blizzard loves this mechanic and it will not be removed. My only hope is that they will tune it as best as they can and make it less intrusive than now so tanking can become more interesting and fun once again.

    Vengeance probably solved one or two problems, but many more have arisen ever since its introduction.
    I am looking forward to reading your future articles.

  2. Rhidach says:

    I completely agree with all your points here, but I don’t think Blizzard is inclined to fix Vengeance. They seem very ok with it as is.

  3. Antigen says:

    Good post, Mel.

    Vengeance is a band-aid; I would really like to see Blizzard show some long-term solutions here.

  4. Wocka says:

    Well said Mel. I can’t wait until late 4.2 or 4.3 when our damage taken gets to the point that heroics are a nightmare. I suppose we’ll just have to Pull Moar! Bliz has never given us a mechanic that would successfully last an entire expansion without being changed. In BC they didn’t even bother fixing it, we just tanked with no pants (stop looking at my tail like that).

    Off tanking is unfortunate right now, only on hateful strike fights style fights can you develop any meaningful dps. Even then, vengeance never produces super results. I mean, we already have a thankless job, at least we can see pretty big yellow numbers every now and again.

    • Meloree says:

      Heroic 5-mans and normal mode raids have, I suspect, even more of a problem than heroic-mode raids, especially when there’s any kind of gear discrepancy. I can’t prove that, having not really done any of that content in a while, but given that I’m starting to see issues on certain heroic bosses with poor vengeance stacking and bad avoidance streaks, I’d be stunned if it was better in content with a lower damage intake.

      • Wocka says:

        Best tolerated with a Blood DK alt- vengeance is nice with low mitigation and constant self healing.

      • Meloree says:

        Interesting point – I hadn’t really considered that the different damage intake levels of the different tanks obviously have an impact on vengeance stacking. I wonder if it’s significant for the first 30 second problem, in any kind of statistically verifiable way.

        Paging Theck.

      • Theck says:

        It probably is, but the differences are likely to be pretty minor. Avoidance is the real culprit, not mitigation. The DK will definitely build Vengeance faster since he’s taking unblocked+self-heal compared to our blocked hits. But the Vengeance starvation problems we’ve been discussing generally aren’t caused by a series of blocks, that should still be enough damage to stack it reasonably well. They’re caused by strings of (avoid, avoid, block, avoid, avoid). Due to the decay mechanisms, I have trouble believing that replacing blocks with hits is going to have enough of an impact to offset the avoidance problem.

        Low-level or undergeared alts of any variety should have no trouble building Vengeance even in Heroics, because their passive mitigation still sucks.

  5. Kihra says:

    Nice post. I think it’s also worth noting that Vengeance promotes bad play as well, Raise your hand if you’ve ever stood in fire, or deliberately not used a CD, etc. just so that your Vengeance will stack faster? I know I have.

    What’s especially annoying is that Hit/Expertise are actually really valuable…. for about 30 seconds. Then you don’t care at all whether or not you had those stats. That just feels really silly and broken to me.

    One quibble:
    “…we can’t afford to gear around that because the boss damage levels assume survival oriented gearing on tanks. Catch-22.”

    The survivability checks in this tier for tanks were not high at all. You actually can reforge avoidance into Hit and Expertise and you’ll have no trouble surviving. Not that I’m advising people to do so… I’m just pointing out that the encounters in T11 at least were not built around boss damage that was so extreme that you couldn’t spare any room for threat stats.

    • Wocka says:

      I used to pre-pot armor. That’s doing it wrong these days. If anything use Strength, but I’m not sure it’s really enough to matter.

      • Meloree says:

        It helps to prepot strength. When you’ve got a 30% chance of having a fury warrior or ret paladin overaggrod at the 30 second mark, it helps.

  6. Meloree says:

    I diagree with your quibble, sir. The survivability checks weren’t particularly high from a Wrath healing standpoint, but they were fairly significant. Just generally deployed over a longer timescale (~4-5s, rather than 1s), and survival gearing became all about weighting the dice against the situations that can and do happen.

    Certainly Al’akir, Cho’gall (Flame Orders or tanking adds), Sinestra, and Nefarian are all capable of killing tanks if the tank and healer aren’t all taking care of business. Less of a problem for paladins, we’re a bit OP :P

    More importantly, bosses have never hit so hard that you couldn’t wear threat stats, so much as it’s always been inadvisable not to weight the dice in your favor as much as possible until the fight was completely understood. Bosses certainly don’t hit so weakly that there’s NO risk to tanks, therefore it’s best to mitigate that risk as much as possible.

  7. Kihra says:

    My perceptions may admittedly be a bit colored by doing most of my progression when WoG was ridiculously OP.

    This may end up being less of an issue in Firelands with only 7 bosses, but with a 13-boss tier, by the time most guilds get to the hardest end bosses, their tanks effectively outgeared the content already by being in mostly 372. There was definitely survivability wiggle room.

    You can give up 4% avoidance or so to get your Hit and Expertise up to respectable levels (not capped but respectable), and your healers won’t even notice the difference. Again, I’m not advocating that anyone do this, but you’re not going to fall over dead because you gave up a few % avoidance. The checks just are not that unforgiving.

    Smart cooldown use just dominated this tier as far as being the primary survivability tool, and paladins just have such a great set of cooldown tools. Used well, they could definitely make up for slightly increased physical damage taken.

    • Meloree says:

      I don’t disagree – and I did most of my progression at the same time as you. Also, WoG is STILL ridiculously overpowered.

      That doesn’t mean that I’m willing to trade avoidance for hit/expertise that I don’t need for 95% of a fight. Especially when that avoidance increases the likelihood that I survive any given dangerous scenario – including things like healer DCs, or just outright fail from someone. It may be a minor effect, but it’s still trading something I value (living) for something I don’t (sustained threat).

      It’s easier to influence on-pull threat by turning around when you’re on GCDand in between swing timers a couple of times, until you’ve got a nice healthy dose of Vengeance, imo.

      • Kihra says:

        Right, I’m not going to trade avoidance for Hit/Expertise during early progression either. That said, I do wish they would make Hit/Expertise matter more. All they would have to do is somehow connect survivability to successfully landed attacks to make Hit/Expertise vastly more interesting.

      • Meloree says:

        The DK model. Death Strike is their primary tool for controlling their own survival, and it has to hit (most of the time, I don’t think they aim to hardcap expertise).

        It works, imo. I’m willing to be it’d be pretty frustrating when it fails, though.

      • Theck says:

        I think there’s a fundamental flaw in that line of thinking, personally. Sure, you can connect threat stats to survivability, but at that point they cease to be threat stats. They’re survivability stats, and you get the threat “for free,” because you wouldn’t be taking them otherwise. In that sense, it’s no different than giving us a talent that gives us some sort of avoidance->hit/exp conversion.

        Unless I’m mistaken, most progression DKs don’t stack hit/exp in any appreciable amount to increase DS healing, they still gear like we do – ignore threat and focus on maxing mitigation/avoidance/health. That should be enough evidence that the “convert threat stats to survivability” paradigm doesn’t generally work – you either tune it so that the survivability bonus of the threat stat is too low, and people ignore it, or so high that they favor it over *real* survivability stats.

        No, in the end, threat stats should matter to us *for threat* and only for threat. Which means threat has to matter, which means Vengeance has to change. We need an implementation that doesn’t let us /afk after the first 30 seconds of a fight, so that we have an incentive to perform well. Sure, there will be tank-swap fights and what not that trivialize it in certain situations, but for the vast majority of content we should care about threat and care about threat stats. That worked rather well in Wrath, because we had long conversations about how much hit/exp it was safe to run with (though it was generally still “not very much”). So we know it *can* be done. Just not via Vengeance.

      • Donthealme says:

        Awesome blog and conversation, it’s funny because I forgot how much I hate vengeance lately until I started reading Kihra’s blog and the talk about abusing it for high dps tanking, I would love for something to happen so threat matters again…I still have never seen the issue with the wrath threat model.

      • Zaephod says:

        I look at the threat stat discussion like this: Currently, we are supposed to care about Health, Armor, Dodge, Parry, and Mastery. Add in some threat stats too: Hit, Expertise, Crit, Haste. Ok, not those last two. Essentially, you want to have to gear for 2 additional ratings. Why? There’s a limit of just how many ratings you can have on your gear, especially with how little dodge and parry scale.

        Gearing as melee is easy, Hit to cap, Expertise to soft cap, then throw on your crit and haste as theorycraft says you should. Why should you make tanks gear that way in addition to gearing for survivability?

        If they really want us to care about threat stats, they need to help us out somewhere. As was mentioned earlier, they could give us some avoidance -> threat stat conversion. That’d be rather similar to Int for casters. Int -> SP and Crit %. Conversely, they could reverse that arrow. Hit and expertise converting to dodge or parry rating at some ratio.

        None of the above really addresses the vengeance debate. Maybe they should just allow vengeance to stack when we land an attack too. It’d be much slower stacking than when we get hit, but at least then an OT could contribute damage and/or be capable of picking up those adds that need to be bursted down quickly. Brainstorming is fun, but I need lunch.

        Y’all have a good one.

  8. Good thread, but all moot point unfortunately. Blizz said quite plainly in ‘Meet the dev’s #9 – Tanking’ that they are fine with Vengeance the way it is and it ain’t gonna change.
    Also they arn’t even considering how it works in pvp or low end overgeared content or 5mans.

    Just gotta do what i do and reforge for the hit/exp caps as appropriate.

    • Meloree says:

      Blizzard says a lot of things, but they’ll also change their minds. There’s no reason not to write about my perspective on Vengeance, maybe it tilts the scales and gets us a better design some day.

      And no, I plan to continue to reforge away from hit/expertise wherever possible. There remain ways to deal with the first 30 second problem – they’re just frustrating and unintuitive. But in the long run, survival is more important.

    • Theck says:

      What basis do you have for saying that they don’t consider PVP or low-end content? Because we have direct evidence that they *do* consider both of those situations when it comes to Vengeance, our mastery, and pretty much every other change they make to us.

      Blizzard’s dev team is made up of humans – they can make mistakes, and they can change their minds. And they’ve admitted to both of those in the past. So it’s clear that if presented with a reasoned argument, change can happen. See, for example, their recent comments about how the Holy Shield changes came about. People asked for it, they decided it was a good idea and delivered.

  9. Pseudo says:

    Alternatively, they could:

    1. Reduce the effect of vengeance by X%, and bump base damage a proportional amount of damage (so it evens out.. I think it’s safe to say Prot whatever is an annoyance and not a ‘threat’ in PVP unless they have a healer and numerous people attacking them.. but they are still not as deadly as anyone else + a healer.. so not a huge PVP hit there with just a small 20-30% adjustment.
    2. Vengeance is based on “damage you would have taken (e.g. a blocked hit gets vengeance equal to the pre-block damage, and dodge/parry obviously gives the full value) — This gives equal (vengeance/threat) to any tank regardless of d/p/b/m percentages, and helps over geared CTA tanks as gear improves
    3. Give all tanks some form of “vigilance” mechanic.. Different tanks can do it different ways for fun combinations
    - Paladins get vengeance on crusader strike/hammer if they hit a target on someone else
    - Warriors already covered, but perhaps will need tweaking or an ability to also boost their generation
    - Bears & DK’s I don’t know well enough to really even guess what they can use — but anything used semi-frequently can perhaps generate some vengeance similar to paladins
    #3 might even remove the need for #1′s (since they would actually *have* vengeance and could contribute

    So i guess in short, I agree with most of what you said. I personally don’t think Prot paladins are very effective in PVP beyond annoyance levels and perhaps out-last strategies so I don’t think a PVP Boost is really an issue, especially with how things will continue to scale with DPS numbers.

    I will say vengeance as is needs tweaks, and right now is *boring* as an off-tank.. then again off-tank in general is boring with 20%+ miss chance if you need to be next to/in front of the boss

    • Meloree says:

      Indeed, prot paladins aren’t very effective in PVP – and that’s the way Blizzard likes it, I think.

      Either way, I agree – there are a lot of options to make Vengeance “better”. Anything that continues to include ramp/decay in combat is going to remain annoying and frustrating for offtanks – maybe that’s intended, but I’m still not a fan.

      I absolutely agree that Vigiliance is a good tool for prot warriors offtanking right now – I’m a little jealous up them.

      Sorry for the moderation delay, I wasn’t near my computer all day.

  10. Pingback: Tanking Circle Questionnaire: The Melbot Edition | Sacred Duty

  11. David Tell says:

    Until they fix the hit/expertise vs avoidance threat issue it im not warrior tanking 5-mans (I may tank with other classes with easier and faster rotations like bear).. Im sick of not being able to get threat easily at the start of fights with stupid dps people not realizing how vengeance works (I dont like taunt spamming for no good reason). I wont run to stay ahead of the party either (even if that works it sucks and makes no sense from a design perspective – i derisively call it run and gun tanking). Developers assumed every dps would give tanks a few seconds to get threat (lol at idiotic devs). I have heard they are testing giving some built in bonuses to hit and expertise off defense stats on gear. If they do that it will help and I might start tanking. Ill stick with arms/fury for now. Im in a good casual guild but most of the players are just flat out scrubs – friendly scrubs so I dont like the idea of leaving. They just wont listen as few of them have ever tanked lol.

  12. Pingback: Vengeance: Once More, With Feeling | Sacred Duty

  13. Pingback: We Didn’t Need Pants Anyhow | Sacred Duty

Leave a Reply