Why Sunwell Radiance is a good mechanic

This is an extended version of a reply to a comment on Rhidach’s latest blog entry.  The poster (Selyndia) suggested that boss expertise, which is functionally not very different than a progressive Sunwell Radiance debuff, is a bad mechanic:

Also, for those that will bring up the “Boss Expertise” argument. Honestly, until I see it, it sounds like a brainstorming idea that won’t pan out. “Boss Expertise” sounds like a good way to keep tank itemization in check, but it’s a really bad end result, because it makes it so that no matter how geared a tank is, they perform exactly the same in the new tier of content. Imagine if, to handle over inflation of healing gear they had bosses put a blanket “All heals do X% less” aura out there, or if bosses all naturally took a fixed percentage less of damage than normal. It would be a similar situation. The biggest problem with this also, is that you don’t feel like you’re character is improving. Your avoidance and mitigation remain pretty constant; the only stat really changing is your HP total as Stamina scales higher, unless you are doing older content. Who really wants to play a character, that tier after tier isn’t really improving performance? No DPS would want to deal 10k DPS in the first tier of an expansion and 10k in the last tier due to passive boss abilities.

I’m not posting this to pick on Selyndia, because this is a common sentiment.  But I do want to raise awareness of a fact that most people overlook – the Sunwell Radiance mechanic is actually necessary as a balancing tool.

As gear improves, everyone gets better at their jobs.  DPS hits harder (INT/STR/AGI/SP/AP), hits more often (haste), and crits more frequently (crit).  Their overall DPS goes up significantly from the beginning of an expansion to the end.  In a similar fashion, healer throughput and longevity increases significantly, everyone’s health increases from extra stamina on gear.  Tanks get an even larger health increase, and they improve their mitigation and avoidance to reduce total damage taken.

To keep boss encounters interesting, each new tier of bosses has to raise the bar slightly.  They have to demand higher DPS output, higher tank survivability, and higher healer throughput to keep the game fun and challenging.  That’s the way the raiding amusement park works – you must be this tall to kill the boss, and if you’re not, go back to the previous ride and “grow” a little (gear up) before you come back.

The game already has natural ways to most of these things to keep boss encounters difficult.  I like to call them “counters,” because they’re numbers the developers can tweak to counteract the increase in player strength. Most of the counters are pretty straightforward and intuitive, which is good – it means the game design allows for scaling and expansion.

Most of the counters are pretty simple.  Player DPS went up?  Great, we’ll increase boss health to tune the encounters in order to our desired time-to-kill.  Note that it doesn’t matter how DPS increased.  The boss doesn’t care whether you hit it more often or hit it less often but crit more frequently.  It cares about one metric or “observable” – DPS (or in boss-speak, “OW OW OW THE LITTLE PEOPLE ARE HURTING ME”).

Similarly increased health, healing throughput, and mitigation in the form of armor and block are countered by making the boss hit harder.   For a tank this means that the bosses melee hits get larger or more frequent to increase the total damage taken per second (DTPS), which stresses the healers and tank health more.  For the raid, it means an increase in the amount and size of random raid or AoE damage events.

What the game doesn’t have a natural counter for is tank avoidance increasing, or in other words the boss’s success probability going down.  The developers can increase the amount a boss hits for, or how frequently he attacks, but those are both indirect solutions that cause collateral damage.  They’ve increased boss DPS output above what it takes to properly stress tank health and healer throughput/mana in order to combat the extra, unreliable damage reduction of avoidance. That puts a tank in a situation where you’re more likely to get a bad string of RNG and die to a larger-than-intended boss spike.  This is exactly what happened from Naxx-ToC in Wrath, which is what prompted the changes in 3.3.

The natural counter to a tank’s avoidance increasing is to make the boss more likely to hit you.  And that’s only possible by suppressing some of your avoidance somehow.  It doesn’t matter what you call it: boss expertise, Sunwell Radiance, Chill of the Throne, or Deathwingcrown Throniance: Electric Boogaloo; it’s the logical way to counter increasing avoidance without introducing side effects.  In that sense, it’s no different than boss health or boss damage increasing – it’s a fundamental mechanic that should be part of the core game mechanics.

It doesn’t have to counter all of your new-found avoidance, just enough to make sure that avoidance doesn’t reach levels that start leading to imbalance between tanks and/or excessive amounts of RNG.  If we gain 3% avoidance in each tier, the boss expertise mechanic might only offset 2% of it, leaving us with a small net gain and a slower, more gradual increase in total avoidance from tier to tier.

I do agree with parts of Selyndia’s comment.  This mechanic is not best-implemented as a debuff.  A debuff feels like you’re being nerfed.  You’re sad, because you spent all that time collecting dodge rating only to see your character sheet read 1% dodge.  That’s depressing, and in some part expected.  You can bet that DPS players would complain loudly if they got a -10% damage done debuff in each tier of content.

But interestingly nobody complains that boss health goes up each tier, despite having exactly the same net result as far as encounter balance is concerned.  The reasoning is pretty obvious – we’re hitting harder, faster, critting more, and generally feeling like more of a badass all-around.  The boss’s health is just the finish line we’re racing towards, and while the race might be exactly the same if we ran half as fast and half as far, it’s more fun to go faster and longer.  It lets us retain the feeling that our character has significantly improved from previous tiers.

The avoidance equivalent should be implemented in a similar fashion.  Instead of a debuff on the player, it should be a buff on the boss that reads “BossName’s attacks ignore X% of their target’s dodge and parry.”  That way you don’t feel like you’re being nerfed, but the game still raises the bar for each successive tier of content by increasing X.  Think “Firelands Precision” instead of Chill of the Suncrown.

It’s semantics, of course, but I think that the buff vs. debuff issue is one of the primary reasons players hated Icewell Radiance so much.  So even though the numbers work out the same, this is one of those cases where psychology makes one route much more appealing than another.  Because at the end of the day, what matters is that we’re having fun playing this game, and fun is not always something you can distill down into an equation.

This entry was posted in Tanking, Theck's Pounding Headaches, Theorycrafting and tagged , , . Bookmark the permalink.

14 Responses to Why Sunwell Radiance is a good mechanic

  1. Rhidach says:

    Well said, Theck. I’ve only ever looked at the issue from the viewpoint of “boy, it sucks to be nerfed” or how annoying it is to walk into a place and see a debuff informing you that you’re “x% less badass”. Consider me a convert.

    Although, frankly, I’m for whatever measures keep us as far away as possible from the Wrath system of our depending on some absurdly high avoidance percent as the thin red line between life and being a puddle. If this was system used to keep tank avoidance low–and correspondingly boss damage survivable for more than 2-3 hits–then that’s definitely a boon.

  2. Thisius says:

    You raise all valid points, but really all I’m seeing is why Sunwell Radiance is necessary, not why it’s a good mechanic. I might be making a distinction without a difference here, but to me things like Sunwell Radiance are examples of how the current mechanics today don’t scale well at all.

    That isn’t to say that WoW is broken or anything, but I think clearly there are some scaling issues with the way we approach content. Sunwell Radiance is necessary in that it keeps the game fun and enjoyable to a degree, but I’d prefer a world where Sunwell Radiance isn’t neccesary, because I don’t think it’s very fun.

    • Theck says:

      I think the distinction is more about how it’s implemented. Sunwell Radiance as a debuff in the last tier of content says to the players “oops, we screwed up the scaling, this is a quick fix.” But Sunwell Radiance as a core component of encounter mechanics, implemented slowly and methodically along the way as one of several tools to properly balance encounters, doesn’t carry that same stigma.

      I firmly believe that if this mechanic existed in classic, with Blackwing Lair bosses having a little more dodge/parry suppression than Molten Core bosses, we’d accept the idea as a completely logical balancing technique in the same way as increasing boss health. “Of course every new tier should be more likely to hit us, the bosses are more badass as we go further, so we need more avoidance to take the extra beating!”

      If Blackwing Lair bosses had exactly the same health as Molten Core bosses, but had higher armor/resistance/etc. so that your spells and abilities did less damage, would there be public outcry? Probably not. It would feel awful if it were done just for one tier (say, Naxx 1.0), but if it were done progressively and consistently we would expect it by now, and we’d accept it as “how the game works.”

      All of that aside, your point about “a world where Sunwell Radiance isn’t necessary” is well-taken. It’s not strictly necessary as a mechanic. If avoidance is tuned such that we start out the expansion low and never get above a certain “tolerable” value – say 40-50% total miss+dodge+parry – then you can probably get the whole thing to work without it. But that’s not really a different “world,” it’s just constraining the stat to a very fine region of parameter space. And if plans change along the way you might end up outside that region, like we did in Wrath with herioc modes accelerating the ilvl progression ahead their initial projections.

      Basically, the point is that you can try and balance around it if you have to, but you’re trying to do a job without having the right tools. Sure, I can try and drive a nail in with a shoe, but wouldn’t it be easier and more logical to do it with a hammer?

  3. Ryan says:

    “fun is not always something you can distill down into an equation.” I’ve been reading the AT&C for a few years now so if Theck and the others haven’t got an equation yet then this must be true. I like your point about a buff vs debuff. Good stuff.

  4. anafielle says:

    What? No equation for fun? I’m disappointed. You’ll have to work on that. I’ll expect a blog next week that includes at least three MATLAB sims modeling “fun”. Get cracking.

  5. Tomaj says:

    Deathwingcrown Throniance: Electric Boogaloo

    Win.

    Anyway, I think a lot of DPS (and some tanks) don’t realize exactly what kind of fiasco Wrath as a whole created for healing, and why there was so much stress on changing healing styles from Wrath to Cataclysm. Arguably, one could include Sunwell in this with Circle of Healing/Chain Heal spam bots. Essentially, things like Chill of the Thron are necessary more than anything to keep healing in check. It’s technically feasible to have a DPS tank a boss in some manner (doesn’t necessarily make it a good idea, but it’s possible – see Chimaeron). It is not possible in any way, shape or form to complete encounters without healers.

    So, to stress healers in Wrath, they gave us Chill of the Throne, a zonewide debuff that reduced dodge by 20%. Since tank dodge continued to go on the diminishing returns table regardless of the 20% less dodge, what happened is that this caused scenarios where a tank could get one-shotted if a healer was not constantly spamming their fastest heal possible. Enter the mana regeneration variable.

    Throughout Wrath, healers were essentially incapable of completely running out of mana (unless you were a holy priest) because of how regeneration mechanics worked, and certain spells. Tack on the general style of healing in Wrath (say, Rejuv x5, Wild Growth, repeat), and you have something that comes out to be very problematic because there’s no thought in it. When we were doing ICC in Wrath, our healing assignments were, “paladin(s) on the tanks, everyone else heal anything that’s missing health.”

    To get back to my point, you have to plug in this variable along with the tanking avoidance (DPS is a non-issue, and is balanced around fire – you can blame DPS for Alysrazor) in order to get the bigger picture, which as I said, many people miss. But wait, there’s more!

    Did you know that in PvP, healers already have a 10% minus healing applied in all PvP areas (Wintergrasp, Tol Barad, arenas and battlegrounds)? Yep. If you PvP, you get to deal with Sunwell Radiance.

    So really, in the end, these mechanics are not really even to balance tanks – they’re more to balance healers. At least, that’s how I see it. Mostly because it seems Blizzard can’t seem to understand that they didn’t get it right the first time (see, incoming nerfs to Mana Tide/Innervate).

    • Theck says:

      I agree, the tank/healer dichotomy is one that you really do have to treat together. Though I see Icecrown Radiance as a response to the possibility of tanks getting globaled, not the cause. The cause is avoidance scaling, which is exactly what Chill of the Sunwell is meant to combat.

      Early in the expansion, avoidance is low and boss hit size is tuned appropriately for tank health and healer throughput. As gear gets better, they have to increase boss hit size and/or swing speed to keep average DTPS high enough so that tank gearing and healing isn’t too easy. But because of avoidance increasing, they have to tweak those higher than what’s necessary to counter tank health and healer througput alone. Thus, the average DTPS after avoidance is tuned right, but it means that they’ve increased the DTPS ceiling (or the maximum “spike” DTPS, which is what you’ll get when you fail to avoid several hits in a row).

      That’s the real problem – if you tune average DTPS around healer throughput, then the spikes get very high and result in a tank being globaled, which forces the healing style of ICC (heal has to be landing on the tank every global). If you tune spike DTPS around healer throughput, then the healers can heal reactively, but average DTPS is very low and healing feels too easy.

      The Cata system really is the best of both worlds, I think. You do have triage, and you do have spikes, but tank healers can spend globals cross-healing and using raid utility without fear of tanksplosion. The worry is that the avoidance scaling issue hasn’t been addressed, so without the “boss expertise” mechanic we might inch back towards tanks being globaled in later tiers. I’m sure they’re going to try and design it such that that situation doesn’t happen, but I’m not sure it’s avoidable without a natural counter.

      The hope is that by introducing it naturally as a boss mechanic, it becomes something the player base takes for granted. We expect bosses in later tiers to hit harder and have more health, so why can’t we expect them to be more likely to hit us? If it’s not done gradually, we end up with Deathwing Radiance in the last tier and then have to listen to players whine and QQ about Blizzard failing.

      • Tomaj says:

        Well, it’s more of a chain of events. Tank avoidance gets too high, then we get Nerfbomb Radiance to put it back down again, and this on its own (tanks not being able to avoid damage) causes the possibility of globaled tanks, and thus Wrath-esque healing. The tanks actually being globaled, however, is because of the absurd amounts of damage bosses were doing, in addition to the avoidance issues.

        All in all, tanks and healers have to be balanced around each other. With the tank set bonuses (more avoidance) and the healer set bonuses (mana regen/AoE throughput), I think T12 is shaping up to be more of an AoE-fest than anything, which (in addition to this tier, short of a few specific phases of certain encounters) goes completely against triage healing as a whole. In other words, from what I can tell from set bonuses alone, for healers, it’s back to ToC again. Hooray.

        To get back on topic, though, since they have to be balanced around each other, and all the plate classes are actually getting parry bonuses instead of dodge, it also stands to reason that maybe dodge is a little bit over the top in terms of where they want it? As far as I remember, they didn’t get to implement their original planned changes for parry (which is a shame), but it’s been a while since I’ve seriously tanked, and I’m not completely up to par on how diminishing returns work right now.

        Still, the other problem with bosses having a higher chance to connect (with a non-dodged/parried attack), this could still lead to more spiky damage taken, and again, we end up back with Wrath healing because healer throughput needs to outweigh longevity, especially as I said with nerfs to mana regeneration cooldowns and what’s going to be considered “efficient” healing at any point. What might end up happening is that like Mastery > all for some tanks, Intellect > all for most, if not all, healers (in fact, that’s how it is already, with some healers even reforging out of spirit). I guess basically it goes back to what I said earlier – they didn’t get it right the first time, and now they can’t figure out how to fix it.

        I also apologize if this was a little rambly, I’m tired and haven’t slept yet.

      • Theck says:

        The point I was trying to make (perhaps badly) is that it’s not the Dodgeball Radiance debuff that causes the spikiness and allows tanks to be globaled. Whether you have 50% avoidance or 10%, you will occasionally run into strings of 2-3 melee hits in a row. If you assume that average DTPS is balanced in both cases (i.e. the boss hits 1.8 times harder on the high-avoidance tank), then the high-avoidance situation will have a spikier damage profile than the low-avoidance situation.

        As you noted, it’s the absurd amounts of damage the bosses put out. And the only reason that the damage is absurd is because it’s compensating for high avoidance in an attempt to balance DTPS. So instead of hitting us for 50k per swing and us only avoiding 10% of them, the bosses are hitting for 90k per swing because we’re avoiding 50%. But that just means that instead of a relatively steady 22.5k DTPS (2-second swing timer, 50*0.9/2 or 90*0.5/2), the high-avoidance tank takes large bursts of back-to-back(-to-back) melees for 180k in 4 seconds, or 45k DTPS. The damage becomes less predictable, varies in spike magnitude much more significantly, and becomes more difficult for healers to react to.

  6. Wocka says:

    Nice Post Theck.

    I’ve often wondered whether avoidance is a good mechanic to have in a game like this. I guess some level of avoidance is good- it creates a feel of RNG that is exciting. But when the reaction required to deal with an unfortunate chain of events goes from 5 seconds to 2 seconds, that’s where the trouble starts.

    It seems to me that the best way to deal with it is make avoidance more predictable, maybe as a talent or baked in mechanic and let tanks grow by scaling health and mitigation. In ICC, I think scaling armor was fun. While I don’t like the idea of giving up a tool, I don’t see it as functionally different. An encounter has a large physical component- increase avoidance/mitigation to compensate. Another has more unavoidable damage- increase health. That’s essentially what we do anyway, it’s just we pick the tool that works better at that point in the game (bc-block rate/value, icc- armor, cata-mastery) and pad with the necessary stamina to make the fight work. Avoidance has had such severe DR that we don’t gem it or seek it out these days anyway- let it go the way of defense.

    The second part of this has to deal with healing. From my position, the solution seems straightforward- increase health pools at a rate that scales with potential hps. I think Blizzard just never finished the equation- Healing too easy? Increase Damage- this part they figured out. Damage starts feeling to RNG? Buff Stam of ALL players- either passively through primary stats on gear, or better yet make it where players other than tanks become concerned about their health pools and have the tools to deal with it (it’s a concern in pvp- why not pve as well?) or both. Then healers can throw bigger heals- feel like they are making progress, but providing larger buckets to cast into balances everything out and the arms race can continue in harmony. I interpreted this as a design intent with cata- how pvp was supposed to have a slower pace and healing was about mana management, not about speed, yet here we are going back to old habits in less than a tier.

  7. Pingback: Cataclysm Tanking/Protection Field Manual :: Updated Feb 9, 2011 (4.0.6) - Page 18 - Elitist Jerks

  8. snoweagle says:

    The boss proprty vs zone debuff is different in the respect that different bosses could have differing amounts of precision, depending on the context of their other damage, possibly providing tanks with an interesting decision regarding how much avoidance is appropriate for a given encounter, if boss precision is implemented in such a way that avoidance isn’t flatly nerfed across all encounters. Without going overboard it would provide more of a ‘hard earned character stats vs boss’s stats’ feel.

  9. Rohan says:

    There is another option: not have stats which increase avoidance. If tank avoidance doesn’t increase, if it stayed at somthing like 20-25%, then a lot of these issues go away.

    Of course, you would need replacement stats to keep tank gear interesting, or you could just have tanks use DPS gear, the way feral druids do.

  10. kal says:

    Presumably, the easiest solution to this sort of thing is to simply have a certain amount of damage be off the combat table. Via bleeds, or magic, or whatever – you could do this and have zero change to anyone’s stats. The effect would be the same, but players wouldn’t be nearly so upset, and it’s not like a lot of bosses don’t have these effects already.

    It’s also a lot more discoverable and intuitive – one can find out these abilties easily via the encounter journal.

Leave a Reply