Wednesday, July 20, 2011

Eliminating Reforging, Hit Rating, and Expertise

I hate reforging. It's just so inelegant. It's complex and time-consuming. You basically have to use an external application to figure out how to reforge properly. All in all, it just seems so unnecessary.

So how can we get rid of reforging?

The primary reason that reforging exists is because of Hit Rating and Expertise Rating. In particular, Hit Rating has a very steep value curve that looks like a step function. When you are below the Hit cap, Hit Rating is usually the most valuable stat. Then past the Hit cap, Hit rating is a complete waste. Expertise displays a similar value curve.

So the best path to eliminating reforging would be to eliminate Hit Rating and Expertise Rating. If those two stats were gone, reforging would not be needed. It still might be nice to have, but the extra complexity would outweigh the value it adds to the game.

The easiest way to get rid of Hit and Expertise would be to say that special attacks and spells cannot miss, or be dodged or parried. White melee auto-attacks could still be affected by those stats.

If this was done, the game would not change greatly. Once you hit raiding, pretty much everyone has enough hit to make this the default state anyways. Indeed, Blizzard has started moving towards this state in small steps. First, taunts were changed to always succeed. Then interrupts were changed to always succeed. A lot of threat dumps used to miss, but now always succeed. Extending this to all specials just removes a little randomness.

Then Blizzard just needs to introduce a new stat for melee dps and ranged dps. Perhaps something like changing Expertise to increase autoattack damage. Maybe casters could get a stat like "Spell Charge" which increases damage done by spells with cast or channel times. Then each role would have a unique secondary stat. Healers would have Spirit, tanks have Dodge and Parry, melee DPS would have Expertise, and caster DPS would have Spell Charge.

But none of the remaining secondary stats would display the steep step change of value. They wouldn't go from amazing to useless at a single point. There would still be breakpoints where the value changes slightly, and that would keep the theorycrafters happy. But adding more of stat would always improve your character.

The need for reforging would disappear. Blizzard could eliminate reforging from the game. Mylune would save all the wee forest animals in Hyjal. And we would all live happily ever after.

Monday, July 18, 2011

Dev Interaction with the Community

Blizzard recently ended their "Ask the Devs" series of community interaction. In Ask the Devs, the community posted questions and the questions were voted on. The 10 or so questions with the top ratings for each topic were resented to the devs and the devs gave answers.

It was an interesting exercise in community management. However, Ask the Devs is generally regarded as unsucessful by both the Blizzard CMs and the community.

Let's take a look at why Ask the Devs failed. It was a worthwhile experiment, but I think that there are two major reasons why it failed.

First, the vast majority of questions were thinly-veiled versions of, "My class sucks. When will you buff my class?"

That was not conducive to interesting answers.

The second reason is that asking good questions is just plain hard. A lot of the time, it is difficult to ask a good question unless you already know the answer.

For example, let's take a question like, "Is Blizzard planning to change the paladin healing model?" If the answer is No, this was a very pointless question. But if the answer is Yes (as in pre-4.0), all of a sudden the question is extremely interesting.

But the playerbase doesn't really have enough information to know which are the interesting questions and which are the boring questions.

This was the biggest strength of the previous interaction between Ghostcrawler and the forums. Because Ghostcrawler knew the answers, he also knew which questions were worth answering.

Of course, that model was not sustainable, and Ghostcrawler's presence on the forums tended to warp them, as everyone started trolling for blue responses.

The new form of Dev interaction, the development blog posts, also have weaknesses. In particular, they have a tendency to be at "too high a level". Ten thousand foot overviews are generally not interesting. Specifics are interesting. The best Dev blog was probably Ghostcrawler's line-by-line explanation of the patch notes, which was about as specific as you can get. Even the ten thousand foot overviews can be greatly improved with concrete examples.

But dev blogs are probably an art, and something that needs to be learned. The gold standard of gaming dev blogs, in my opinion, is Mark Rosewater's columns for Magic: the Gathering. A dev looking to communicate with her game's community would probably be well served by studying how Rosewater did things.

Thursday, July 14, 2011

Valor Point Capping

Over the last couple of weeks, when I've sat down to game on the computer, here's a list of activities that I wanted to do:
  1. Raid Firelands
  2. Work on the new Molten Front daily quests and achievements.
  3. Play Mass Effect
Notice what's not on the list? Running Troll Heroics for Valor Points.

I liked the troll heroics. I don't mind doing them occasionally. But honestly, being "forced" to do them when there's all this cool new content to explore is very annoying.

What's especially aggravating is that I know that in a couple of weeks the novelty of new content will have started wearing off, and it will be enjoyable to mix in a few heroics. But by then my guild will be 6/7 or 7/7 and we'll be capping out Valor Points just from raiding.

Of course, "forced" is a bit of a misnomer. But right now, capping valor points is the most effective use of non-raid time to help your team defeat bosses. And the worse your guild is, the more pressure there is to do more heroics. Someone in a 4/7 guild only has to do 3 heroics, while someone who is 1/7 needs to do 6.

Please note that the single most effective use of time overall is to put in more attempts on the boss. If your guild is thinking of raiding less and instead doing heroics for valor point gear, reconsider! Down that path lies failure, madness, and a shattered raid team!

I am really not fond of mechanics that keep me doing things that I may not be interested in. If I just want to raid, why not just let me raid? Why make the optimal path include a lot of other stuff?

Tuesday, July 12, 2011

Mastery/Crit Build In Practice

I decided to give the Mastery/Crit build a try last night in raid. We were going after Baleroc, and it was mentioned that this build was particularly good for Baleroc since the fight is mostly single-target heals.

The Mastery/Crit build uses the following stat priority: Intellect > Mastery > Crit > Haste > Spirit. The idea is to rely more on Holy Light and less on Divine Light. It was a little nerve-wracking to reforge away so much Spirit.

Performance-wise, the build is solid, I think. The numbers were definitely comparable to the other paladins. Additionally, the single highest source of healing done is actually from mastery shields, Illuminated Healing (something like 27% for me, followed by Beacon at 25%), and damage prevented is always better than damage healed. As well, overheals are not a complete waste, as they still give shields.

For playstyle, initially the build feels very sluggish. High haste means that all your spells are very fast, and it feels easier to react. But you get used to it after a while, and you still have Holy Shock and Word of Glory for speed.

Surprisingly, mana is not really an issue. The thing about Spirit/Haste builds is that even though you regen mana faster, you also spend mana faster. This build seems a lot less "swingy" when it comes to mana. You spend mana at a steadier rate. You still have to Judge on cooldown and use Divine Plea, like normal.

Personally, I rather like this build. It feels very solid and complements the other healers in the raid well. I'd like to see how it performs across more fights, but I think it's the build I will try for the next few weeks.

Monday, July 11, 2011

Holy Paladin Builds

Blizzard's class dev team is probably estatic at how their changes have worked out for Holy Paladins. Holy Paladin theorycraft is all over the map at the moment. As I can see it, there are currently three different stat weights being discussed.

1. Intellect > Spirit > Haste > Crit > Mastery

The standard build from 4.1. It's still the best build if you have to raid heal, and is also the most versatile build.

2. Mastery > Intellect > Spirit > Haste > Crit

The ultimate single-tank healing build. Insanely large mastery shields buffer the tank's health. Probably only viable in 25-man raiding with good raid healers to cover for you.

3. Intellect > Mastery > Crit > Haste > Spirit

Eloderung of Eternal Reign is touting this build. He says that it is a very strong two-tank Beacon healing build. It relies a bit more on Holy Light, and takes advantage of all three new mechanics: 100% HL transfer through Beacon, 200% crits, and a stronger mastery. Unlike build 2, this is likely a viable build for 10-mans.

This is probably the most interesting time to be a Holy Paladin in a long while. Excellent work, Blizzard!

Monday, July 04, 2011

Zaroua on Holy Paladin Mastery

Zaroua of Premonition made a really interesting post on the official forums discussing Holy Paladin Mastery today:
I'm creating this thread hoping to greatly reduce the influx of PMs and tells I get regarding our mastery. Keep in mind that is pretty much only for 25 man raids; I see some potential uses for this in 10 man, but for the most part it probably won't be as useful in a 10 man scenario.

The Firelands fights come in two categories: sustained AoE damage stages and no AoE damage/AoE damage that Holy Radiance isn't suited to heal. Beth'tilac, Lord Ryolith, Domo and the ground phase of Alysrazor are all mostly based around AoE healing for sustained amounts of time while while Alysrazor air phase, Shannox, Baleroc and Ragnaros are fights where Holy Radiance isn't even worth casting.

For the AoE fights, all we really have going for us is Holy Radiance and Light of Dawn, while all of the other healers are able to pump out AoE heals nearly non-stop. Now the way I look at it is that someone still needs to heal the tank and that since Paladins are pretty horrible at sustained AoE healing, a 25 man raid may as well put one or two Holy Paladins on full time tank healing and just have them Holy Radiance on cooldown.

For the non AoE healing fights, Paladins have the choice to choose between tank healing or... tank healing. Casting Holy Lights and Lights of Dawn on the raid simply doesn't compare to Chain Heal/Wild Growth/Circle of Healing/Prayer of Healing.

My guild's healing team is flexible enough to allow for a Paladin to have a weaker Holy Radiance in order to have more powerful tank healing. This translates in the AoE healers having to spend less time on the tank and more time doing what they're good at.

How does this relate to mastery? The golden rule of tanking and healing is that when it comes to handling damage, the best to worst order to do it in is this: completely avoiding the damage, mitigating the damage, healing the damage with a very large amount of fast heals, healing the damage with slow and large heals. Mastery helps mitigate damage and in some cases, completely avoid it. The reason why mitigating damage is so good is because it leads to reduced frequency of spikes in the tank's health and when spikes do occur, it makes them less pronounced, which in turn means that other healers don't panic and waste cooldowns or inefficient heals on the tank. Before the 4.2 change to make Mastery shield stack, the stat was mostly useless because such a huge portion of the shield was wasted on any given heal that gearing for Haste for faster reaction times was something pretty much every Holy Paladin agreed to being the better choice. But now we're in a position where Mastery is finally viable for a Paladin who wants to focus on tank healing.

The most important thing to note about healing with a Mastery set is that you sacrifice throughput in order to become a more effective tank healer. [Emphasis mine.] Your Holy Radiance (and mana regen) will be weaker than a Paladin who is going for a more balanced approach to gearing or a Paladin going for Haste. But in turn, you'll be putting a downright overpowered shield on the tank every time you heal him directly. And don't kid yourselves: if you're able to reduce the average hit the tank takes by 10k because of the Mastery shield, what you're doing is very nearly game breaking. The shield simply is that good for keeping tanks alive. What a full set of mastery comes down to is your guild's capacity to support one of its healers focus less on raid healing and more on tank healing.

Even if your guild can't (or won't) support a Paladin will a full mastery set, every Paladin should try to get off pieces with mastery on them so they can use them for Shannox, Baleroc and Ragnaros at the very least.

If you look at his gearset, he's gemming and reforging according to the following priorities:

Mastery > Intellect > Spirit > Haste > Crit

This has a lot of drawbacks. It's hyper-specialized for tank healing, and possibly even single-tank healing (no off-Beacon healing).

Now, you probably shouldn't run out and switch to this right away. But any time someone from a top Royalty guild like Premonition or Paragon says something that contradicts common wisdom, it's worth taking a good long look at the situation.

Sunday, July 03, 2011

Melee vs Ranged

It's pretty common to hear that a fight favors ranged over melee. Relatively few fights favor melee over ranged. So let's take a look at why this happens.

In my experience, most fights don't punish you for taking melee, they punish you for taking more melee than normal. Put another way, it's easier to take extra ranged DPS than it is to take extra melee DPS.

So what are the advantages that ranged has over melee?

1. Any effect that hits one melee usually hits all the melee.

Usually if there's an effect that melee needs to avoid, all melee have to avoid it at the same time. However, you very rarely see an effect target all the ranged simultaneously. Usually the ranged can spread out, so that only a portion is hit at any one time.

This also shows up in effects that "chain" from one target to another. It's easier to organize ranged such that chain lightning hits a minimum number of people. It's much harder to do the same for melee.

2. Ranged can go to melee, but melee cannot go to ranged.

If a fight does favor melee, ranged can often avoid it by running into melee range. Take Magmaw, for example. A common strategy is to have most of the ranged DPS stand with melee, with minimal people standing out at range to be targeted by the specials.

If ranged could not do this, it's arguable that Magmaw would be considered a melee-friendly fight, instead of a draw.

Not saying it would be a completely good idea, but if ranged had a 10 yard minimum on their spells, the balance between melee and ranged would be very different.

3. Any effect that affects the ranged will also affect healers.

The healers are usually at range, so any environmental effects will often affect them as well. That means that there is a maximum to what can be done to ranged without making it impossible to heal.


As you can see, it's not so much that individual melee classes are weak, but more that melee and ranged are affected by mechanics differently. Add to that the ranged can "pretend" to be melee characters, and pretty much every advantage ends up falling to the ranged.

Fixing this might be pretty complicated. Large fixes, like a minimum range for ranged, are not really feasible at this point.

The best way to go is probably identify the ranged classes and have effects target those characters. For example, when Magmaw does flame pillar, rather than targetting "a character at range", he instead targets "a ranged DPS spec" even if that character is standing in melee. That would allow melee specs to avoid some mechanics, while preventing the ranged specs from running into melee range.