Sacred Duty

Yesterday, we discussed some of the pros and cons of our current mastery implementation, and briefly considered the limitations of a bounded system like the combat table.  Today I’m going to toss around some ideas for new mastery implementations that could be used in MoP to avoid the complications caused by block-capping.

Diminishing Returns

The first, and most obvious answer, is that they could apply diminishing returns (DR) to block.  There are certainly some advantages to that approach.  It would give block a similarity to dodge and parry, and would allow for more complicated interplay between the three because the “best” stat would depend on the amount of each that you currently have.  If the numbers are tuned properly, it would completely eliminate the ability to reach block cap.  All in all, it seems like a fairly easy solution to implement.

However, there are some problems.  For starters, no other classes have DR on their mastery bonuses.  Paladins and Warriors would be the only ones who are punished for stacking more of their “awesome” stat.  Admittedly, the same is true for dodge and parry, so tanks are used to it, but it’s a valid criticism of simply slapping DR on block and calling it a day.  It would mean that mastery procs and buffs would have less value for a shield tank than a DK or bear, which makes itemization trickier.  That may not be a critical problem, but it’s enough to make us consider some of the other cool ways they could change mastery.

Expandable Combat Table

For a slightly more radical idea, we could completely change how the combat table works.  Right now, the roll is made between 0 and 100, and the combat table is adjusted according to current stats.  What if, instead, the combat table expanded to fit the possible options?

For example, you start out with 5 “points” of dodge, parry, miss, and block, and about 30 “points” of taking a regular hit.  We’ll ignore crits for now, since those get removed by spec bonuses.  Instead of rolling from 1-100, the combat roll would be from 0-50 (5+5+5+5+30=50).  So without any gear on, you’d have ~60% chance to take a regular hit, and a 10% chance to dodge, parry, or block.

As you load yourself up with gear, all of your stats increase, but so does the combat table.  So if you end up with 20 dodge, 20 parry, 5 miss, and 50 block, your roll is between 0 and 125 (20+20+5+50+30=125).  Your chance to take a regular hit has decreased to 30/125, or 24%, and your chance to block has increased from 10% to 40%.

This system is very similar to putting the stats on DR, because there’s no way to push “hit” off of the table anymore.  You can increase your stats indefinitely, but there will always be that small chance of taking a full hit.  Except in this model, the stats all scale linearly in “points,” but share the same DR formula when converting from points to percentages.

It’s a neat idea, but ultimately I think it’s worse than just putting DR on block.  For one thing, the system is a little less obvious.  A player may understand what “20% dodge” means, and the character sheet could still provide that number, but it’s unintuitive and confusing that adding more block reduces their chance to dodge.  At least with diminishing returns, you get the independent scaling of each stat, so that adding extra mastery doesn’t cause collateral damage to avoidance.

Neuter Block

As we discussed yesterday, one of the big reasons that mastery is so attractive has to do with the strength of blocking itself.  Shaving 30% off of an attack makes a very significant difference.  But if block was weaker, then mastery would be as well.  They could “nerf” blocking to prevent only 10%-20% of the incoming melee attack, which would reduce the impact of blocking on survivability.  In that system, you’d still be able to block cap, but it would no longer be as super-powerful to do so.

Unfortunately, this idea has more problems than diminishing returns.  It makes mastery our weakest damage prevention stat by a wide margin.  It doesn’t fix the discontinuity you arrive at when reaching block cap.  Even though this is not all that unlike the discontinuity that DPS players have to juggle with hit/exp caps, there are also no proc- or buff-based hit and expertise effects in the game.  But there are mastery procs, and it would be advantageous for those procs to be attractive to all five tanking classes.

Block Chance Fixed (BCF), Mastery Increases Value

In this idea, block chance becomes a fixed amount for warriors and paladins.  Instead of increasing your block chance, it’s set at 30%-40% as soon as you choose a tanking spec.  Mastery then has to play a different role entirely, and the logical choice is to focus on increasing the amount blocked.  Warriors could keep an incarnation of Critical Block, where additional mastery simply increases the chance of doubling the amount blocked.  Paladins could simply get an increase in the amount blocked, at roughly half the rate warriors get Critical Block.  For example, the same amount of mastery rating it would take to give warriors a 50% to critical block would give paladins an extra 15% block value, so both would yield an average of 45% damage blocked.

There’s a lot of merit to this idea.  It’s simple.  There’s no obvious bounding issues, because the scaling can easily be tweaked to make sure that 100% Critical Block chance isn’t achievable.  While paladins get the more reliable block amount, neither class has a particularly “smoother” damage profile because block-capping is out, so full-sized hits still occur at a more-or-less fixed rate.  Interestingly, dodge and parry get more attractive here as well, because they suddenly provide the only source of increased Combat Table Coverage.  That may devalue mastery enough that we don’t pursue it at all costs, which will keep us from lusting after high-mastery DPS plate.

I can’t really think of that many drawbacks, either, aside from mastery falling behind dodge and parry in attractiveness.  Again though, that could be considered a positive, and there are other classes who value mastery less than other secondary stats.  The only fear is that this makes it so weak that we actively avoid it, which is unlikely.  However, there are ways to fix that….

BCF, Mastery Generates a Proc

Using the same “fixed block chance” base as above, in this case mastery also generates a proc-like effect.  This could be its own stand-alone effect, or it could be in addition to an amount of bock value (or crit block) to make sure that mastery remains attractive.

The proc itself could be any number of things.  A proc that gave us a few seconds of a Holy Shield buff, increasing our block value during that time, is one interesting option.  For warriors, it would either be a buff that grants a chance for a Critical Block (with a longer duration or higher crit chance to compensate for the double-RNG), or simply a shorter-duration buff that grants the full 60% block value.  Not only does this idea have enough parameters to tweak to make it attractive compared to dodge and parry, but if the proc trigger is melee attacks it gives us a reason to like hit, expertise, and even a traditionally-eschewed stat like haste, something Mel has written about before.

On the other hand, this system brings block chance back to the table as an option.  A proc that gave us +30% chance to block for a short duration could work; the uptime probabilities involved produce a natural diminishing returns effect, such that you can never guarantee full CTC, and extra mastery rating gets less and less valuable automatically.  Or the proc could be a Shield Block-esque buff, which guarantees that the next melee attack will be blocked but is consumed by that event.  Again, probabilities give you a natural DR mechanism and automatic anti-cap control, and a proc triggering off of damage dealt gives value to DPS stats (hit, exp, haste).

Or the trigger could be defensive instead of offensive.  For example, it would be neat if every successful block had a chance to proc a heal or short-duration absorb bubble.  Mastery would increase the proc chance and the absorb/heal amount.  This is an interesting proposition for two reasons: it procs defensively, much like the reactive damage of the old (BC-era) Holy Shield, and it’s ex-post-facto.  The second part is the important one – it means that the heal will rarely be wasted (unlike offensively-procced heals like Mending and Souldrinker), and that the effect isn’t as powerful as a straight-up increase in block value, because you need to survive the larger-sized hit before the absorb/shield bubble is granted.

A more controllable alternative might be for a successful block to grant a bonus to Shield of the Righteous, Holy Shield, and Word of Glory.  The bonus could increase damage/healing/absorb done and duration, and mastery would increase the potency and proc chance.  This has the advantage of tying the mastery into the Holy Power system and giving the player some choice in how that benefit gets used.

You could imagine any number of other, less block-centric proc-based benefits.  For example, SotR could proc an absorb bubble or mini-heal that gets consumed on the next unblocked attack, as a way of smoothing out damage.  That’s getting closer to what DKs and bears have, though, and veers outside of our self-imposed “must involve blocking” constraint.

BCF, Mastery Converts Blocked Damage to Absorb/Healing

Taking an idea from the previous section, and removing the proc mechanics, mastery could simply convert blocked damage to healing or damage absorption.  Block chance could remain fixed, while mastery gave a small increase to block value and increased the conversion factor.  So when you block, you might reduce the damage by 35% instead of 30%, and get half of the amount blocked as a free heal.  Absorb bubbles would give more damage smoothing than the heal, if that sort of mechanic is desired.

This has the advantage of being RNG-less outside of the inherent combat table rolls.  It’s not particularly easy to math out because it’s boss-dependent, but it’s something you can easily identify and analyze in parses.  The disadvantage is that it’s similar to the druid and DK mastery effects, though it’s triggered and calculated in a different way.  Unfortunately, it’s also a primarily passive effect, in that nothing you do can increase the trigger rate, unlike Savage Defense.  But so is the current incarnation of block chance mastery.

BCF, Mastery Grants Rollover Minutes

Using another idea I touched upon earlier, what if our mastery gave us “rollover minutes?”  Blocking a melee attack could grant credit towards a buff which causes a heal or absorb bubble on the next unblocked attack.  The idea is that it’s a Lifebloom-like effect that triggers off of a guaranteed damage spike.  Mastery could determine the conversion rate, just like above, or could adjust the proc chance if a more unreliable mechanic is desired.  And again, the concept here is that every block continues to contribute to the buff (hence “rollover minutes”), so that multiple blocks in a row will result in a bigger heal when the effect triggers.

This wouldn’t be that hard to implement, and there are more than enough parameters to tweak to make it an interesting mechanic.  It does share some similarities with druids and DKs, but the proc and usage mechanics should be enough to make it feel unique.  But again, it’s a passive effect rather than something the player can control or influence.

Get Rid of Tank Plate

On the more radical end of the spectrum, we could consider getting rid of tank plate entirely.  After all, druids already use the same DPS leather gear that rogues do, and monks will be doing the same in MoP.  At that point, dodge and parry rating go out the window too, and get the same “fixed amount” treatment that block does in some of the earlier suggestions.  By choosing a tank spec, a warrior or paladin automatically gets 20% dodge, 20% parry, and 30%-40% block.  Our mastery would then have to trigger off of damage dealt, making hit, exp, crit, and haste more attractive.

There are several ways one could go with this idea, many very similar to the BCF solutions.

• Mastery could grant a stacking block amount buff from successful melee attacks (maybe ~5% amount per stack, up to 6 stacks, consumed upon a block event), with mastery rating increasing the amount per stack.
• Alternatively, you could grant block chance, again with a per-stack amount that’s consumed upon a successful block.
• Or with either of those methods, the stack properties could be dependent on damage done, tying crit into the mix as a useful stat.
• Or the stack properties could be fixed, and mastery simply provides the proc chance per successful attack, possibly double for crits.

However, this idea isn’t without it’s host of problems.  If all tank plate gets eliminated, then plate DPS gear becomes far and away the most highly-represented gear selection in raids.  There are already 2 warrior specs, 2 DK specs, and 1 paladin spec fighting over DPS plate; adding 3 more tanking specs to the mix will require it to drop even more often, which isn’t good for drop tables (or, for that matter, for other classes!).  In addition, this would require a re-working or replacing of all tanking gear in the game, from level 1 to 85.  That’s a lot of developer time, and I’m not sure it’s worth it when there are other, arguably better solutions on the table.

Closing Thoughts

In this post, I’ve provided a number of possibilities for the paladin (and warrior) mastery implementations in MoP.  Some of them were a little far-fetched, others were very down-to-earth.  I think that the “BCF,” (block chance fixed) solutions hold the most potential for a reasonable implementation.  They’re all fairly straightforward, and easy for a new player to understand.  They work within the current combat table system, and with current gear, and thus don’t need heavy restructuring of the game’s fundamental mechanics.  Most of them grant hit, expertise, and haste survivability benefits.  And finally, they completely eliminate the problem of block-capping by removing the mechanism that gets us there, while giving us other neat effects to compensate.

And of course, this isn’t a comprehensive list.  Given more time and thought, other solutions could certainly be found.  I’d be willing to bet that our readership has their own innovative ideas.  If you do, please don’t hesitate to share them in the comments.

In the latest “The Light and How to Swing It,” Matt Walsh (a.k.a. Rhidach from Righteous Defense) made some pretty sharp criticisms of our mastery design in Cataclysm.  I think he’s got some valid points, but I don’t agree with all of them.

For example, he derides the stylistic choice of block chance, and calls it a “returned and renamed stat which was stamped as fit for deleting.”  I don’t think that’s entirely fair.  Certainly, the old block chance was unattractive in Wrath, but that was primarily because block itself was weak.  However, that’s not the fault of block chance.  If block had been very strong, block chance would have been very attractive.  This is exactly the situation we have in Cataclysm – a very strong block mechanic and a high valuation on mastery.

In that light, I think it’s erroneous to call it a stat that was “fit for deleting.”  Far from it, in fact.  The real reason block chance was removed was not weakness, but redundancy.  Since block chance was only useful to two of the four tanking specs, it was a natural choice for conversion into a mastery stat.  So block chance wasn’t removed, really.  It’s still with us, just under a different name.  And it’s actually our most prized stat.

I also disagree with him on the issue of style.  As paladin tanks, we’ve traditionally been the “block tank.”  Blocking is in our blood, and it’s thematically appropriate for the class to have a block-centric mastery.  Block chance seems like it was a reasonable choice on stylistic grounds, as paladins traditionally blocked more than warriors.

However, Rhidach’s post does hit upon an important point.  Block chance differs somewhat from most other stats, in that it can be capped and that cap is fairly readily achievable.  Crit and haste can be capped, but not passively or easily.  Armor can be capped, but again, not with obtainable gear.  Primary stats, health, attack and spell power are all unbounded.  And dodge and parry have diminishing returns (nonlinear scaling) that keep them from reaching an upper limit.  Block chance joins hit and expertise as the only three stats that have this unique combination of linear scaling and an upper bound.

And that’s what makes block mastery broken.  Not in the sense that it’s overpowered (though you could argue that it is, in some cases).  But in the sense that it’s incredibly hard to balance effectively.  It’s either so strong that you want to stack it to cap, or it’s weak enough that you ignore it entirely.  It’s difficult to find a middle ground, where you’d be just as happy with 1 mastery rating as you would with 1 dodge or parry rating.  And it’s especially difficult when only a subset of the tanking classes can do it.

Rhidach is also right that this problem has its roots in the combat system itself.  The single-roll system, where the result is between 0 and 100, creates the boundaries, and the linear scaling of mastery happily expands to fill them.  A bounded system can lead to problems simply because when those bounds are reached, funny things happen.  You can hit discontinuities, get accelerated scaling, and create other unintuitive behavior.

A familiar example to some might be avoidance.  In Burning Crusade, avoidance scaled linearly, so that every X rating gave 1% dodge or parry.  This meant that as you stacked more avoidance, every point of additional rating actually got better.  To see that, consider that adding 1% avoidance when you’re at 50% total avoidance (so, going from 50% to 51%) reduces your damage taken by 1/50 (=dx/(1-x), where x=0.5 is avoidance and dx=1% is the increase), or 2%.  Adding 1% avoidance when you’re at 75% (75%->76%) reduces your damage taken by 1/25, or 4%.  Doing the same at 90% avoidance (90%->91%) reduces your damage taken by 1/10, or 10%.  You can see why this could be a problem for balance – the tank that manages to squeeze out an extra bit of avoidance gets a huge benefit.  So even small discrepancies between classes could lead to large performance gaps.

As of patch 3.0, avoidance sidesteps this behavior with diminishing returns.  No matter how much avoidance you stack, you can’t get 100% of either dodge or parry individually.  And the formula is set up such that it would take a ridiculous amount of raiting to get even 50% of either one.  Even now, in nearly best-in-slot gear in the last patch of the expansion, we can only manage 20-25% of each.  The diminishing returns formula acts as an arbitrary limiter so that dodge and parry rating act like they’re an unbounded system, even though they’re doing so within a bounded structure.  It also makes sure that the impact of avoidance rating gets smaller the more you have of it, so that the 1/(1-x) behavior of Burning Crusade doesn’t happen.

The problem doesn’t generally occur with hit and expertise, despite being a bounded system.  In fact, quite a few DPS classes have a hit cap as their first priority, because it’s their strongest DPS gain.  But damage output works a lot differently from damage taken.  The DPS gains you get in reaching hit cap have a finite upper bound.  You’re increasing the success rate on your way to 100%.  With avoidance, you’re decreasing the boss’s success rate, which gives you the scaling I described above.  That’s why getting 1 more hit out of 100 isn’t game-breaking, but taking one fewer hit out of 100 can be.

So, you may be asking yourself, what can Blizzard do to fix the problem presented by block chance mastery? Or, more generally, what can Blizzard do to redesign our mastery for Mists of Pandaria?  In tomorrow’s post, I’ll be exploring a number of possible ways they can implement mastery for shield tanks and discuss the pros and cons of each.  We will, however, adhere to one constraint: all of the ideas have to involve blocking, somehow.  Sure, they could give us our own version of Savage Defense, but that wouldn’t fit the theme of a shield tank.  There’s no reason to homogenize tank mastery, so we’re going to avoid doing so (apart from keeping both warrior and paladin mastery bonuses shield-themed, of course).

So stay tuned for tomorrow’s installment!