Sunday, January 31, 2010

Offline For A While

My cable company was working on my building last night, and they managed to knock me offline. The next tech can only come out Friday, so I guess I am taking an enforced vacation from the Internet and WoW.

Still my cable company is pretty competent, and this is the only issue I've had in three or four years, so maybe a vacation will be good for me.

I've managed to make a post a day for the month of January, and didn't want to see that streak get broken on the very last day.

For the record, you can really see the difference in traffic that a regular posting schedule makes.

See ya in a week (or maybe sooner if I'm lucky).

Saturday, January 30, 2010

Top WoW Videos - #3 - Tales of the Past III

The #3 video on my countdown is Martin Falch's magnum opus, Tales of the Past III.

Note that there's no Youtube link. In a world where most machinima are shorts, rarely exceeding the 10 minute mark, Martin Falch made a full-length, one-and-a-half hours-long, movie. The movie is 2.3 GB in size, so I recommend using the torrent option to download it.

The movie is pretty epic. It features many of the big names in Warcraft, including Thrall, Jaina, Arthas and the Ashbringer. Also, there's a paladin in full Judgment, which automatically makes every movie better. Tales of the Past III is over the top, but gloriously so.

It is not a perfect movie. In a lot of technical aspects, it is weaker than the other movies lower on the list. But I admire boldness and daring. Martin Falch dreamed a greater dream, one that most other machinima creators would hesitate to raise their eyes to. And he had the fortitude to see his vision to completion.

Top Video List (so far):
  1. Tales of the Past III
  2. The Craft of War: BLIND
  3. Big Blue Dress

Friday, January 29, 2010

Measuring Absorbs

Ah, Discipline Priests. According to some meters, they are the worst healers. According to other meters, they are the best healers because their shields prevent so much damage. What's the truth? Who really knows? Probably somewhere in the middle, would be my guess.

All of this is because the combat log does not report absorbs correctly.

The lesson here is, again, one of feedback. You need to feedback to be able to correctly evaluate abilities.

How does Sacred Shield compare to the other abilities? Is it weak or strong? I have no idea. I take it on faith that it's doing good, so I keep it up on the tanks always. But I really don't know.

That could actually be another point in favor of the FoL paladin healing style. Focusing on Spellpower gives you a really strong Sacred Shield. Maybe that's the element which makes the build truly viable. We don't know, though. We can theorycraft, but theorycraft really needs data to confirm the math.

If you have a game with a combat log, all effects need to reported and properly attributable. Blizzard really needs to get a proper reporting system for absorbs up and running.

Thursday, January 28, 2010

Gevlon Is Right

I may disagree with his methods, but at his heart, Gevlon is right. Most people do not play anywhere near their full potential.

Why then is it verboten to point this out? Maybe all the people who advise you not to make a fuss are just cowards. They're afraid of being called "mean" or "elitist". Who cares about about other people, so long as you get your badges. Much easier to take the safe route and avoid confrontation; to laugh about a poor player in guild chat behind her back.

Surely the better path, the harder path, is not to turn a blind eye to poor players, but to help them become better ones.

The difference between Gevlon and I is what we believe are the motives. To Gevlon, bad players are "morons and slackers", who are bad players because they choose to be bad. In my view most people want to be decent players, but a lot of them don't really know how to be good. They don't see the path from what they are to what they could be. Or worse, they ascribe the difference entirely to gear, which is pretty much the worst mistake you can make if you want to be a better player.1

Pointing out that a player has poor DPS is not mean. It's a fact. If you can shame or push them into seeking out external information at Elitist Jerks or the WoW Forums, that's pretty much the only thing which will make them better players.

Of course, though, it's all about style. Calling them terrible and trying to vote-kick them from PuGs unless they do better isn't going to work. It would work if they were actively slacking, but not if they don't know how to do better.

1. See the Why Are DPS So Bad? articles on the sidebar for more on this.

Wednesday, January 27, 2010

Kicking Underperforming Players

Gevlon is turning into a real social type of player these days, on a crusade to improve the level of play in instances. Whatever happened to goblinish profit-above-all motives?

Still, I can't help but think his methods are backfiring. Attempting to kick the lowest DPS resulted in the group kicking him instead. To be honest, that exchange was probably the funniest thing I've seen all week.

People generally aren't going to kick someone for underperforming, so long as the performance does not affect success. If success is a bit slower than it could be, that's okay for most people. However, causing the group to wipe--as Gevlon did--will get you kicked.

In my experience, people are very lenient about trying. As long as the player is not just /afk leaching, people are tolerant of performance, regardless of how good they are actually doing. It's only when wipes happen, where the lower performance results in actual failure, that the tolerance disappears.

I think the carrot might be a better weapon than the stick, at least in random PuGs. Post the meters, say that X could do better, and then point her to Elitist Jerks. That probably will have a better success rate than attempting to play power games, even if you are the healer or the tank.

Tuesday, January 26, 2010

Thoughts on FoL vs HL Builds

There are are currently two styles of paladin healing builds at endgame: one focusing on Flash of Light, and one focusing on Holy Light.

The Holy Light build is the pretty standard gem-everything-with-Intellect build that we all know and love. It relies on massive amounts of Intellect so that mana regeneration from Replenishment and Divine Plea is maximized.

The Flash of Light build focuses on Spellpower, and relies on the fact that FoL is very efficient to handle mana concerns. The idea is to get enough Haste so that your FoLs cast in 1.0 seconds (about 650 Haste rating), resulting in steady stream of very fast, moderately strong heals. It's sort of being like a living HoT. You generally need T9-level gear before this style becomes viable.

One thing to note is that neither build uses one spell exclusively. HL builds can use Flash of Light when damage is low to save mana, while FoL builds can still drop Holy Light when the going gets rough.

Holy Light's advantage is that it has huge raw throughput. Flash of Light's advantage is that it is "good enough" most of the time, is very steady healing, and you don't have to worry about using Divine Plea or the -50% healing Divine Plea entails.

My personal preference is Holy Light, because the Holy Light build is a bit more "forgiving" than FoL. Massive amounts of healing makes up for a lot of mistakes. If you use HL when you could have used FoL, oh well, you just wasted some mana. On the other hand, if you use FoL (even a strong FoL) when you should have used HL, someone might die.

Monday, January 25, 2010

Exponential Soft Enrages

One new element in Icecrown Citadel are exponential soft enrages. These are mechanics which start out slow and "speed up" as time goes on. There are currently three fights that use exponential soft enrages: Deathbringer Saurfang; Professor Putricide; and Blood Queen Lana'thel.

In the Deathbringer encounter, Saurfang gains Blood Points when he deals damage to someone. When he reaches 100 Blood Points, he puts a Mark on someone and deals damage to them whenever he hits the tank. That means he generates the next Mark faster, and so on. If a person with a Mark dies, Saurfang heals for 5%.

In Putricide Phase 3, the tank gets a debuff which causes raidwide damage. Each stack increases the raidwide damage by a factor of 3. If the debuff expires for any reason, including tank death, Putricide is healed for a large amount (10% maybe, but P3 starts at 35%, so it's effectively a third or more of his remaining health.)

In Blood Queen Lana'thel, one person is turned into a vampire allowing them to do more damage. After a short time, they have to bite someone else and turn them into a vampire as well. After another interval, two new vampires have to be created, and so on. If you aren't able to bite someone, you get Mind Controlled.

One of the most interesting things about these setups is how vulnerable they are to "fiddling" in the early part of the sequence.

For example, in Deathbringer Saurfang, you can let the first Mark die. Saurfang heals for 5% and you lose a DPS, but you effectively reset the soft enrage timer. Let's say the first Mark comes at 2 min with the second Mark coming at 3 min. Letting the first Mark die means that the first active Mark comes at 4 min, with the second mark at 5 min. You gain a huge amount of time because you were able to reset the sequence, and the fight becomes trivial. The penalty of 5% health and a DPS is not enough to prevent the tactic from working.

In Putricide, the penalty is very high. It's high enough to make resetting the stack non-viable. However, you can buy a bit of extra time by inserting more tanks into the tank rotation. Each extra tank would give you an extra 40s in P3 while keeping the raidwide damage low. It's possible that a guild which is having trouble with P3 could drop a healer and put in an extra tank.

For Blood Queen Lana'thel, originally the sequence of 1-2-4-8-16 was less than the hard enrage timer. So people started doing a sequence like 1-2-(4-3)-6-12-24. When you hit four vampires, one vamp would deliberately die to shift the sequence slightly. A small change early in the sequence netted more time, enough that the hard enrage became the limiter. Again, there was no penalty for fiddling with the sequence. Blizzard eventually increased the timer, so the normal sequence would hit the hard enrage timer as well.

So exponential soft enrage timers are a very interesting mechanic introduced in Icecrown. However, they need strong penalties to avoid people tampering with them. An early reset or delay can push back the soft enrage by a large amount. In two of the three fights that use it so far, the penalty wasn't strong enough.

Sunday, January 24, 2010

AoE and Crowd Control

Often the debate on Crowd Control sounds like it is a choice between AoE or Crowd Control. If tanks don't have good AoE threat, then obviously you can't cast AoE spells, but you might use Crowd Control more.

Are we doomed to have choose between Blizzard and Polymorph? If you could only cast one of those spells in instances, which would you pick?

Perhaps the solution doesn't lie in player abilities, but in how mobs are created. For example, you can't really AoE in Faction Champs because the mobs have a -75% AoE buff.

What if we extended that buff to all elite creatures? The major difference between Elite and normal creatures would be that elites would be very resistant to AoE and normals are not. At the same time, rather than only having elites in an instance, Blizzard could put in more normal creatures, or convert some elites in a pack to normals, with higher health and and attacks.

Right now, pretty much every mob in an instance is elite, which has made that designation lose a lot of its meaning.

So now you get a choice, depending on what mobs you face. If you face elites, you might have to use Crowd Control. If you face normals, you can use AoE. Fights that currently use Crowd Control can have the adds tagged as normal, but with the same stats as now.

It might even allow Blizzard to make more interesting trash packs. Imagine a pack with 5 elites and 10 normals. You might have to separate out the elites so that they can be controlled while the normals are AoE'd down.

The big advantage of this solution is that you don't have to nerf tanks. They still have good AoE threat for when it is necessary, or when AoE would make a fight better. But they can't rely solely on AoE threat, and they have to single-target at appropriate times. In a multi-type fight, a warrior would use Thunderclap to pick up the normals, and Shield Slam to hold the elite.

In many ways, I think this solution, creating a difference between elites and normals regarding AoE, could give us the best of both worlds.

Saturday, January 23, 2010

Top WoW Videos - #4 - The Craft of War: BLIND

The #4 video on my countdown is The Craft of War: BLIND, by Percula.

Like 90% of WoW's population, this video inspired me to make a rogue.

Though most of elements such as style, music, plot, and pacing are well done, the real strength here is the animation. Percula used the WoW models and scenery, but did his own animation. He did a superb job. Everyone, from the soldiers to the rogues to the trumpeter, is extremely well done.

In particular, I love how he used the animation to convey the essence of his main characters. The rogues are all style and flair, with lots of movement. On the other hand, Lady Prestor, who is the dragon Onyxia masquerading as a human noble, is very contained, with minimal motions. It is a beautiful contrast.

One thing I enjoy in movies is a good introduction of a villain.1 The first impression you get of the villain is often the most important one. Percula's presentation of Onyxia in this video is awesome. He spends the video building up the blood elf rogue, from the tavern fight, to the two rogues fight, to the group of guards before Lady Prestor. Then the rogue, which Percula has taken such pains to establish as exceptionally competent, gets to Lady Prestor. And Lady Prestor just beats down the rogue without breaking a sweat; then has the audacity to mock the rogue about Lady Prestor's secret identity, thus using the rogue's own Blind against herself. It is a superb introduction that utterly establishes Onyxia's credentials as a genuine bad-ass.

There are also a lot of small touches in this video. For example, the scratches in the Blood Elf's sword at the end, and the way the guard flinches when Onyxia deflects the dagger at him. All these small touches add to the finished product, making The Craft of War: BLIND a great, great video.

Top Video List (so far):
  1. The Craft of War: BLIND
  2. Big Blue Dress

1. My favorite villain introduction is in the movie The Illusionist. That introduction is just sublime in its efficiency.

Friday, January 22, 2010

Difficulty and WoW

Various people are once again discussing the difficulty (or lack thereof) of WoW. I don't really understand the fascination with this topic.

Here's a some cold, hard truths:

1. If an activity is hard, most people who attempt it will be never be successful.

2. Fifty percent of all people are below average.

So what does this mean? Let's say you made questing in an MMO to be of average difficulty, whatever that may be. Congratulations, you just eliminated half of your potential customer base! I'm sure your shareholders are thrilled with the purity of your Vision™.

The basic activity in your game must be fairly easy. Questing should be easy enough that 90% of all potential players can do it. (And judging by the fact that people seem to need addons like Questhelper, it isn't there yet.)

After that, you can scale difficulties upwards. 75% of players can do regular dungeons, 50% of players can do heroics, 30% of players can do normal raids, 10% of players can do hard modes, and 0.1% can do the single hardest fight in the game.

To my mind WoW has done a very good job of providing content of appropriate difficulty for everyone. There is something for people of every skill level including those people who are not that skilled. Providing basic gameplay that low-skilled players can do is just good business sense.

If you think WoW is too easy, maybe you should step up to the next level of challenge.

Thursday, January 21, 2010

Will You Play Star Trek Online?

I'm not in the Beta for Star Trek Online. The shenanigans to get a beta key seemed like more work than I was willing to put in. But I've been watching a lot of the news and reactions to the game.

I'm really on the fence about this game. The ship combat looks kind of neat, but I'm not sure about the rest of the game.

I've been watching the EJ thread on Star Trek (amusingly entitled "Elitist Kirks"), and one thing that strikes me is that the game seems to get very complicated very fast. And I'm not sure I'm up for a very complicated MMO.

I'm also not sure about the IP. I like Star Trek, but it's always struck me as something I watch. To see a story unfold. I never really thought of it as something active, something you play. I think I've only ever played one Star Trek game (Star Trek:25th Anniversary on the NES, if I recall correctly). I don't actually remember too much about it, but Star Trek has always been firmly in the "not-game" category in my mind.

I guess we'll see how bored I am when STO actually launches.

Wednesday, January 20, 2010

Beacon of Light and Range Tip

One fact that's often overlooked is the fact that Beacon of Light has *60* yard range, even though paladin heals have a 40 yard range.

So if you're assigned to heal a target that occasionally goes out of range, keep Beacon up on them, and heal people closer to you. This gives you a little more leeway in moving. You don't have to move right away to keep pace with your target, you can move a bit later. Or sometimes you don't have to move at all, if they will come back in range.

This is a particularly useful tactic for fights where someone has to go pickup adds and bring them back to the group, as they'll often only be out of range for a few seconds. Beacon of Light can even be cast when the target is up to 60 yards away, so you don't even need to refresh Beacon early.

This is a small tip if you are having range issues as a paladin healer. Choose your Beacon target wisely. Sometimes the main tank is not the best choice.

Tuesday, January 19, 2010

Unifying Paladin Mana Mechanics

Each paladin spec has it's own mechanic to regenerate mana. Holy pops Divine Plea once a minute to get 25% of total mana back. Protection has Guarded by the Light to get 100% of total mana back per minute. Retribution has Judgements of the Wise which returns roughly 190% of base mana back per minute.

The interesting thing is that the amount of mana returned is not that far apart. If we say Holy has a total mana pool about 5 times base mana, then Holy gets 125% base mana back per minute. Ret is a bit of an outlier, but not that much higher.

Rather than having three separate mechanics, maybe it would be better to have one mechanic that all three specs use. I think the best mechanic out of these three is Retribution's JotW. Judgments are an iconic part of the paladin class. Plus, Judging is fun. A giant golden hammer drops out of the air and smacks your enemy in the head. Never gets old. Hitting Divine Plea once a minute is not particularly fun. As well, most tankadins seem less than enamored with the way Guarded by the Light works.

Suppose JotW's functionality was built into the base Judgement spell, so that each Judgement returned a net of 15% base mana. JotW could increase the mana returned by another 5% or 10%, to keep Ret the same.

That way, each spec would be encouraged to Judge a lot, and Judgement would be the primary method of regaining mana. We could then remove Divine Plea from the game.

Monday, January 18, 2010

Crowd Control and Tank Threat

In a comment to the last post, Jeremy offers another suggestion for promoting Crowd Control:
Another alternative is to make it difficult or impossible for a single tank to hold threat on all mobs simultaneously.

Back in the good (bad?) old days of vanilla WoW, when warriors were basically the only tanks and had very few AoE threat-generating abilities, you *had* to CC because the tank simply couldn't hold all of those mobs at once.

In BC, when tankadins became viable, their AoE tanking abilities became the gold standard: groups forgot about CC, and other tanks suffered because groups expected them to just tank everything. (As a warrior tank at the end of BC, I routinely felt like I just didn't have the tools to tank effectively.) In WotLK, this imbalance was fixed by giving *every* tank reasonable AoE capabilities, which made CC irrelevant.

Thus the "quick" fix for the lack of CC is to nerf every tank's AoE threat generation abilities to the point where they can't reasonably expect to hold more than three active mobs at once.

I don't agree with this idea. The thing is that there is an alternate strategy to Crowd Control if the tank's AoE threat is poor. The DPS simply focuses on a single target, and nukes them down one by one.

Now, admittedly, many DPS seem to find the concept of "focus fire" and "assisting" to be advanced techniques, but there's nothing really preventing them from learning. Kind of honestly, they seemed to have the same problem with "not breaking sheep".

A tank just has to generate enough AoE threat to get past the healers. Strictly speaking, good AoE threat is not absolutely necessary, it's just necessary if people want to use AoE spells.

Tankadins really came into prominence in TBC because it was "easier" for a paladin to gather everything and the raid AoE things down. I remember seeing Hyjal tanked by Warriors, and they were fine, you just had to be a little more careful.

You could have such weak AoE threat that the healers would pull off a tank, and that would make Crowd Control much more useful. But it would also make running dungeons a bit more dicey. Sheep breaks, and the mob runs straight to the healer. That's pretty much a recipe for a wipe.

Tanking is already an unpopular role. Making it harder for tanks probably isn't the best of ideas.

In a way, that may have been the bargain: Crowd Control was sacrificed to make tanking easier, making more people willing to try tanking. Overall, I'd say it was probably a good bargain. I would rather have more tanks than more use of Crowd Control.

Sunday, January 17, 2010

Crowd Control

A lot of people feel that Crowd Control should be more prominent. You should have to Polymorph more, or Banish, or even Sap. I am not entirely sure I agree with this. I'm not sure I disagree with it either. I remember the days of Sapping or Sheeping, and they were kind of cool.

However, when I think of Crowd Control, I think of those demons in Tempest Keep. The ones with the Buzzsaws of Doom. You may remember them. If you failed to CC them, they started launching buzzsaws at the raid, doing massive damage before you could get them locked down again. Even the gap between Banishes was dangerous. Kind of honestly, I would rather not see trash like that again.

The thing about Crowd Control is that, in order for it to be used, the mob must be too dangerous to leave active while other mobs are alive. That means if you make a slight mistake, it often leads to a wipe. If the mob is any less dangerous, then it will just be tanked down.

Another problem is that mob must eventually be tanked and killed. You can't really make an untankable mob and expect people to control that. So the controlled mob exists in this zone between "can be tanked" and "cannot be tanked". And the better the gear you get, the more the mobs move to the "can be tanked" category.

I suppose if you really wanted to enforce Crowd Control, you have to create dependencies between mobs in the same pack. For example, Mob A applies a debuff to the tank increasing Fire Damage taken by 500%. Mob B does Fire damage. The optimal solution here is to fight the two mobs seperately, either by using two tanks or crowd controlling one until the other is dead.

I think that would be a fine mechanic to use once or twice. But if every trash pack started having these dependencies, just to force people to use Crowd Control, it would soon seem contrived.

Saturday, January 16, 2010

Top WoW Videos - #5 - Big Blue Dress

Since Saturday is a pretty slow day, I thought I'd take a look at what I think the best WoW machinima videos are over the next few weeks, and in true Blessing of Kings style, over-analyze them to death.

We'll start with #5, Big Blue Dress, by Cranius.

Now, I'm not precisely sure that Big Blue Dress is the fifth-best WoW video, but I couldn't think of anything else I would put above it. Plus, Cranius does have a very good track record with his other videos as well.

Big Blue Dress is also nominally a PvP video. PoM-Pyro has always been a classic PvP mage tactic, and it's amusing to see a video devoted to it.

Finally, the kicker is the absolutely hilarious gnome chorus, which just fits the video perfectly.

Friday, January 15, 2010

Limited Attempts on Putricide

I really like the limited attempts model on Professor Putricide. For those who are unaware, Putricide is the last boss in the Plague Wing of Icecrown Citadel. You get 10 attempts on Putricide each week, after which you cannot engage him in combat any more.

I like the 10-attempt limit. It's a good number, large enough so that it gives you room to maneuver, but small enough that it really restricts you and causes a good stratification among the guilds. The 50-attempt limit in Trial of the Grand Crusader was way too many. Many weeks, we never even used up all 50. In contrast, only 10 attempts makes it feel like you need to make each attempt count, while still giving you time to talk things over.

(We haven't beaten Putricide yet. We've pushed into Phase 3 a couple times, and have 5 attempts left for our next raid day.)

I'm not sure how the rest of ICC's raid model will work. From what I understand, the attempt limit will increase, but each end boss will share attempt counts. I.e. Maybe next week you will get 15 attempts to defeat both Putricide and Blood Queen Lana'thel. The number of attempts will keep increasing. As well, eventually the faction leaders will provide a raid buff to help guilds complete ICC.

Personally, I think that seems a little bit complicated. I think a better model would be to have attempts on the bosses be independent of each other. Then, the number of attempts should remain at 10 always. For example, you get 10 attempts at Putricide, and then 10 attempts at Blood Queen Lana'thel. Only the faction leader buff will "nerf" the instance as time progresses.

I think this model would be easier to understand, and easier to make decisions about. You don't need to juggle number of attempts and bosses. Additionally, the 10 attempts provides a real limit for all guilds, and will produce the nice stratification that the Putricide limit has created.

Thursday, January 14, 2010

Lowbie Tank Tales: Authority

One of the interesting things about tanking is how the rest of the group automatically defers to you.

I've ran Gnomeregan a fair bit, and now know the easiest way to get to the boss with a minimum of fuss.1 When I'm on my tank, the group falls in behind me and we move through the instance smoothly.

On the other hand, when I'm on my rogue, generally people don't know which way to go, and the group stops for discussion. I can tell them the way to go, but they never seem to listen to me. Half the time we go in a crazy direction, and the group never makes it to the end boss. This happens even I am the nominal Party Leader.

Generally, my tank is perceived to have much more authority than my rogue, even though there's not a lot of difference between the two characters. It's like WoW players are trained to defer to the tanks.

Maybe it's just that the tank goes in first. Maybe being the first person in the battle confers an aura of leadership, in the style of ancient generals and kings.

I find the automatic mantle of authority to be a very interesting part of tanking, even if I'm not completely sure why the tank is assumed to be in charge.

1. Generally, always go left, and don't jump down. Jumping down with an inexperienced group inevitably causes more problems than it's worth.

Wednesday, January 13, 2010

Trauma and Paladins

I got a new mace from ICC tonight: [Trauma]. It replaced my mace from Ulduar.

It's rather underwhelming so far. The proc does like 0.6% of my healing.

According to Ghostcrawler:
Estimates of Bryntroll at 8 to 10% of dps seem exaggerated to us. Something at the 3 to 3.5% seems more typical. It should be eminently possible for healers to get 3 to 3.5% healing out of Trauma.

I'm not sure what data Ghostcrawler is seeing, but I get nothing close to this. And I'm usually healing tanks, so it should be hitting the melee DPS as well.

The worst part is that [Trauma] is pretty much the only upgrade for a healing paladin in ICC-25. The only other alternative is a sword with hit rating.

I hope Blizzard takes a look at this weapon and upgrades the proc effect for paladins.

As well, they really need to drop an alternative weapon type that paladins can use. The other healing classes get daggers and staves in addition to maces. Paladins can use caster swords, but Blizzard always seems to put hit rating on the swords, restricting the swords to the DPS casters.

Edit: I forgot to mention that in addition to being a lackluster proc, effect-wise, it also lacks flash, which is a very common complaint for healing effects. In comparison, [Nibelung] spawns Val'kyrs to smite your enemies for you! Why does healing never get the cool stylish effects?

Tuesday, January 12, 2010

More on Tanking and Healing

Gordon responds to the previous post:
Secondly, I'd like defined "harder" in this context. I think leveling a healer or a tank (with that spec) is tougher and more time consuming than leveling a DPS class. Yes, tanks could level as a DPS spec but then would suffer the consequences in their grouping ability and/or be forced to pay the 1,000g to allow for dual speccing. DPS classes don't face this obstacole.

I disagree with this. First, you could always tank and heal normal instances regardless of what spec you have. The base class always includes the necessary tools to heal and tank. Back in 2008, I wrote a post about Retribution Tanking for WotLK normal dungeons. It's really only when you hit end-game heroics and raids that you need to spec for your role.

Second, the new Dungeon Finder makes leveling as a tank or healer significantly easier than it used to be. I've taken a warrior from 15 to 32, and it's been blazingly fast. I would rank it as faster than questing. Not to mention that character has much superior gear for her level. She's almost all blues, without really trying.

With the Dungeon Finder, it is quite possible that tank and healer leveling has become easier than leveling as DPS.

Also, I think there's a fair argument to suggest that both tanking and healing are more stressful roles than DPSing because you are really carrying the weight of your group or raid on your shoulders. I know people will vehemently disagree with me on that point but so be it.

In my opinion as a long-time healer, the idea that "you are really carrying the weight of your group or raid on your shoulders" is an illusion. It is an illusion born of unfamiliarity with the role and the fact that you are an obvious point of failure.

It's true that if the tank or healer makes a major mistake, the group wipes. If you're new to healing, this idea can seem intimidating. This is especially true when you're faced with a new role. DPS is familiar, it's what you did all through your leveling experience.

So when faced with an unfamiliar role, where failure is obvious, it's not really unexpected that the roles of tanking and healing would be invested with more weight than they deserve.

However, once you tank or heal for a while, the unfamiliarity goes away. Once you get as comfortable with healing as you were with killing mobs, it doesn't seem harder or easier, just different.

The first time I tanked on my warrior at 15, I was very nervous. Now, at 30, I'm jaded. I know my role, I know my purpose, and I can take care of my end. That's how fast the acclimation process goes.

You don't have the weight of the raid on your shoulders. It just seems like you do at first. And practice cures you of that illusion.

Monday, January 11, 2010

Should Tanks and Healers Get the Biggest Rewards?

Gordon, from We Fly Spitfires, recently wrote a controversial guest post on World of Matticus, Tanks and Healers Should Get The Biggest Rewards.

I think a lot of the controversy comes there are several issues entangled together in this piece. If we unravel the issues, I think a clearer picture emerges.

Gordon says:
Tanks and healers are the most important classes for any group...These are the two most important classes that exist in any MMORPG. But the DPS? They’re just meat in the room.

This statement is true for certain levels of play, but completely and utterly false at different levels. At the lowest levels of skill--for example, normal dungeons, easy heroics, Naxxramas--yes, the tanks and healers are the most important. A solid tank and healer can carry the DPS to victory. A bad tank or healer will probably wipe the raid.

But at the higher levels of play, the DPS takes on more and more importance. Good DPS becomes supremely valuable, much more so than tanks and healers. Excellent DPS can carry decent tanks and healers on difficult content, while the reverse cannot happen. In my view, the single greatest difference between Royalty guilds and the rest of us is that their DPS players are significantly better than most of the DPS in lower tier guilds.

I mean, it’s in our culture to reward those that do the most and work the hardest, right?

This line goes to the heart of the issue. Gordon assumes that tanks and healers do the most and work the hardest. But he never actually proves that they do.

In my experience, good DPS works just as hard as tanks and healers do. They theorycraft, get the best gear, and hone their rotations on the target dummies. Kind of honestly, the DPS in my guild do more work than I do.

I think a lot of the problem here is that Gordon is conflating "scarcity" with "moral value". Specifically, it is undeniable that tanks and healers are less popular roles than DPS. But that does not mean that DPS players players choose DPS because they are lazy. Or that tanks choose tanking because they are noble souls devoted to the betterment of humanity.

There are many other reasons. DPS is more active. You are actively killing the enemy, not just holding the line the way the tanks and healers do. The DPS archetypes are often more popular and may resonate in the imagination to a greater degree. Hunters, mages, and rogues in particular are very iconic archetypes. Retribution is much closer to the archetypal paladin image than Holy is.

So I don't think that tanks and healers "deserve" the biggest rewards. "Deserve" implies connotations about moral value that I do not agree with, especially given how important superb DPS is in the harder levels of the game.

However, it might still be a good idea to give tanks and healers higher rewards. A higher reward might nudge more people into trying tanking or healing. It's like the new rewards for Oculus. If you complete Oculus, you don't "deserve" better rewards. It simply isn't that hard, especially after it was nerfed. But the rewards do encourage people to stick with Oculus, leading to a better game experience.

Similarly, a higher reward might convince some DPS to switch to tanking or healing. More tanks and healers mean faster dungeon queues for everyone, making everyone a winner. Extra badges would be a bad idea, because that would gear tanks and healers up faster, encouraging them to stop earlier. However, extra gold might be an interesting incentive.

But to reiterate, extra incentives to play a certain role does not mean that role is morally superior to the others.

Sunday, January 10, 2010

Resource Theory

There are a few main types of resources that drive class abilities in most MMOs. This is a categorization of the ones I've seen.

1. Net-Loss Resource

This is a resource which, at the end of the encounter, is lower than it starts. The classic example of this is mana. Classes with mana start at full, and spend it as the fight progresses.

One of the advantages and disadvantages of mana is that all abilities are available right at the start. As well, there is no innate restriction to using any sequence of abilities, until the entire resource is exhausted. This can cause burst damage in PvP, but it does grant the character maximum flexibility to deal with whatever situation may arise.

The major problem with net-loss resources is that they are very dependent on encounter length. If the encounter is longer than the resource can sustain, the character cannot do anything. If the encounter is shorter, the resource does not constrain the character in any significant way.

Net-Loss resources work best when their rate of consumption depends on something in addition to time.

In my view, Net-Loss is the best resource type for healing, because healing depends on damage done. Additionally, having the most flexibility to deal with unusual situations is very valuable when healing.

2. Cyclical Resource

Cyclical resource are used up and renewed multiple times in an encounter. The cycle length is much smaller than encounter length.

The main advantage here is that total time does not really constrain the class. The class is constrained by ability costs and the rate at which the resource regenerates.

There are two main subtypes here: resources which start at full, and resources which start at empty. There are also systems which use a resource that starts somewhere in between, but those system tend to act like whichever extreme is closer.

Full-phase resources (such as Rogue Energy in WoW) allow characters to take immediate actions when an encounter starts. This gives the player something to do as the fight starts, but offers the possibility of burst damage in PvP.

Zero-phase resources (such as Warrior Rage in WoW) require characters to wait before they can take actions. This usually guarantees that the other side has a chance to do something, but is often not as much fun as taking actions right off the bat.

In my view, zero-phase has the potential to be more strategic, but it depends on how fast the resource is acquired, and how important the trade-off between building or expending the resource is. One example of a strategic zero-phase resource is power points in Wizard 101.

One interesting pattern often used is Two Linked Opposite-Phase Cyclical Resources. One resource starts at full, and as it is used, it generates the other resource which powers different actions. Examples of this in WoW are Rogues, where Energy generates Combo Points, and Death Knights, where Runes generate Runic Power. This patter is used because it is relatively simple, but deep enough to be interesting, and provides gameplay where several abilities are used at different times.

I think that cyclical resources are the best for both tanks and DPS characters. Generally, threat and damage depend more on time than anything else, and the other resource systems tend to be too good, or not good enough, depending on encounter length.

3. Net-Gain Resource

This, I suppose, is more a theoretical resource than anything. I'm including it mostly for completeness purposes. This is a resource which increases over time.

The only resource I can think of that does this is Land (potential mana available for that turn) in Magic: the Gathering. On turn 1, you have access to 1 mana. On turn 2, you have access to 2 mana, and so on.

This resource depends heavily on encounter length. If the encounter is long, the resource ceases to constrain the player at all.

This might actually be an interesting resource type for a PvP game. It would make the early stages of the game very strategic, while the final stages would end in a flurry, ensuring the game does not drag out.

Saturday, January 09, 2010

Pally DPS Song

Courtesy of TheGreyFoo, this one is for all the old-school Classic paladins.

His Spacebar Clown is also very good.

Friday, January 08, 2010

Actions and Consequences

In a comment to the previous post, Kring asks:
If the game is designed to have non-revertible changes, isn't it kind of cheating if you return to an old save game to change a non-revertible event?

Isn't it a failed design if you revert to cheating because the game isn't fun if played in the way it was designed to be played?

Well, I don't really know. On the one hand, if you only ever follow one decision path, you'll miss a lot of content. There's a lot of dialog options which you might want to choose just to see what happens. Generally, the snarky options are great examples here. I might want to lip off to the king just to see what the writers did, but I don't want it to be permanent. The thing about non-reversible changes is that they often make you play very conservatively, instead of taking chances.

The freedom to go back and revisit old decisions is also the freedom to see all of the game's content. Depending on how "wide" the game's decision tree is, if you never revisit decisions, you may only see a small fraction of the total content. And that seems like a waste of the creator's hard work.

On the other hand, consequences don't really mean anything if you can wipe them out easily. And then you get into the question of is it worth getting a consequence if you don't like the actions you took to get there. If you feel the actions you took were right and just, should you not then accept the consequences of those actions?

Sometimes though, it's hard to tell what the consequence of your action will be. This is especially true in games, where the information given to you can be limited. There's an example of in Dragon Age in the Dwarven section. You're supposed to pick between two potential leaders. The problem is that you really can't tell what the significant difference between the leaders is. I started helping one leader, then accidentally did something for the other leader that got me stuck with him. And quite frankly, they were both pretty much the same, so I didn't really care and just went with it.

If a game makes some consequences seem arbitrary, shouldn't the player have the freedom to revisit those actions?

So I'm not really sure which way it goes. Personally, I think you should give extensive saves, and let the player decide for themselves. If someone wants to keep to the narrow path, they can easily do so themselves.

Finally, I'll conclude with a quote I've always liked from Lois McMaster Bujold from one of her Vorkosigan books:
His mother had often said, When you choose an action, you choose the consequences of that action. She had emphasized the corollary of this axiom even more vehemently: when you desired a consequence you had damned well better take the action that would create it.
- Lois McMaster Bujold, "Memory", 1996

Thursday, January 07, 2010

Archival Saves

I have to confess that I haven't actually beaten Dragon Age yet. I am really close to the ending. However, I made a choice much earlier that will mean I will not get the ending I want. In addition, I have a problem in that I did some quest that required me to turn in 20 health potions, and now I don't have enough potions remaining to beat the fight currently blocking me.

My first thought was to go back to an earlier saved game, but I realized that I hadn't made a backup save in the timeframe I wanted. I'd have to redo a large section of the game.

It occurred to me that Saved Game systems are rather primitive. Most games have a couple of autosaves, a quicksave, and some manual saves. I think saving progress could be done much better.

For example, you could take a cue from various archival schemes. Consider a scheme where you saved the last 5 quicksaves, then retained every fifth quicksave after that. That would give you a solid history of saved games, while not wasting that much space.

Or since the "important events" tend to be known in a game, you could make a save after each section of the game is completed. This would make it much easier to go back to an earlier point and continue on.

I think saving games could be more automated. Relying on the player to keep correct saves is brittle. A game developer should look at saves more like the creation of a history of that playthrough. Of course, the difficulty part would be accounting for forks in that history.

Perhaps representing the saved game timeline as a tree would be interesting. A player could move through the tree to pick the point she wants to start a new branch. Older, or stubby, branches which did not go anywhere could be pruned away to save space.

Or perhaps just use a linear time model, with dense amount of saves closer to the present, and a sparser amount of saves in past. After all, there's only one instance of an individual playthrough running at any one point in time.

Whichever way, I think save game schemes could be better than the current standards.

Wednesday, January 06, 2010

How To Do Top DPS Video

Here's a very nice video about how to improve DPS after you have the basics (gear, gemming, rotation) down. It's by a hunter, Kripparrian, from the guild Exodus:

Direct Link

I think the best point is to think 2 or 3 abilities ahead. So you aren't reacting, but executing a plan.

Tuesday, January 05, 2010

Lowbie Tank Tales: Floating Rage, Hand of Reckoning

Floating Rage

Thanks to the people who suggested floating Rage from encounter to encounter. That makes a huge difference. At these levels, a Sunder and Revenge are enough to hold a mob until it dies, leaving lots of Rage for the next encounter. I only needed to use Bloodrage when the healer needed to drink.

The Best Laid Plans of Blizzard Designers...

I was in this one Wailing Caverns run with a 18 Ret Paladin, who seemed to be a newish player, or at least new to paladins. He was a pretty good guy, but he seemed to be putting out a ton of threat. I'd Sunder and Revenge, and then move on the next mob. But then I'd notice that the mob was attacking the paladin. I checked for Righteous Fury, but he didn't have it on.

This kept happening until I had a suspicion and checked Recount. The paladin was using Hand of Reckoning as part of his DPS rotation!

When I asked him about it, he responded that it did damage to the mobs, which I couldn't refute.

In a way, I can't really blame him for that (at least, not that much). All the poor guy has for DPS abilities is Judgement every 8s. When he got HoR at level 16, he must have been excited that he got another button to press. And HoR does pretty good damage for Ret and Prot specs.

I couldn't really get him to stop doing that, so I just let him take some hits, and Sundered to grab the mob back. Still, it is an amusing side-effect of having a damaging Taunt.

Monday, January 04, 2010

Low Level Warrior Tanking

With the new Dungeon Finder, I've found that I really enjoy running the old, low-level instances. There's something pure about a group of lowbie characters facing old-school challenges like Deadmines and Shadowfang Keep in greens and blues. Before talents really make a major difference, where people use the old meat-and-potatoes abilities instead of the fancy-shmancy expansion abilities. Where drops are actual upgrades instead of pure disenchant fodder. Where you enter Wailing Caverns, get lost, and finally finish several hours later, defending a tauren healing a night elf from a giant murloc.

After a couple of runs with dual-wielding or two-handed tanks, I decided to make a warrior and level and gear strictly through the Dungeon Finder. A warrior that, further more, would tank properly with: 1) a shield; and 2) Defensive Stance. The DPS warrior tanks did a lot of damage and didn't really die, but they couldn't hold aggro at all. My rogue kept pulling off them and she doesn't even have any heirloom gear.

So far, my warrior has made it level 19, and I've tanked Ragefire Chasm, Wailing Caverns, and Deadmines.

The thing I've found with low level warrior tanks is that they are really dependent on getting hit once at the start of the fight. Once I get enough Rage to Sunder or Revenge even just one time, the mob is stuck to me, and no one can touch me in threat for the rest of the fight.

However, many DPS like to hit the mob before it hits the tank. That makes it much harder for the warrior to recover and you have to start burning taunts. So far, the main culprits have been paladins. I think it's because Judgement has a slight range on it, and because they have mail armor so they're not scared of the mobs. It's terribly enraging to see a mob running towards me, my finger hovering over the Sunder button, and then see a golden hammer hit the mob's head and "Changed Target" flash on my screen right before the mob reaches me.

Now I truly understand the fuss warriors made over Rage and priest shields. It must have been extremely annoying to them to be shielded right before the start of the fight and not be able to get any Rage at all.

I wonder if the warrior experience would have been slightly better if the "rest level" of Rage had been twenty instead of zero. (It seems appropriate that Warriors get to be slightly angry all the time.) If out of combat, Rage rose or fell to 20 rage. This would give warriors enough rage to use one or two moves at the very start of combat and make them not so dependent on getting hit right off the bat.

That's one of the advantages of Paladins and Death Knights when tanking. Their resource starts at full, so they aren't dependent on getting that initial hit.

Sunday, January 03, 2010

Elegant Design: Glyph of Life Tap (WoW)

Since a lot of the time I focus on things that I dislike in games, I thought it would be interesting to take a look at some small elements that I consider examples of elegant design.

First up, [Glyph of Life Tap].

In the beginning, there was Life Tap. And the warlocks looked at it and saw that it was good.

So the following scenario unfolded in dungeon after dungeon:

1. Warlock pew-pews until she runs dry.
2. Warlock Life Taps all the way back to full mana.
3. Healer has a heart attack when they see that someone just dropped to 10% health unexpectedly.

Efforts to get the warlocks to Life Tap more often met with failure. It's not that they meant to make life excessively exciting for the healers, it's just that you don't really think about mana until you run low.

But Glyph of Life Tap changes that. It encourages the warlock to Tap every 30 to 40s, keeping their damage buff up (which is what the warlock really cares about) while simultaneously keeping the warlock's mana levels up and keeping the health hit manageable for the healer.

I like Glyph of Life Tap because it nudges the warlock to play in a group-friendly manner, while still preserving the full use of Life Tap. A more heavy-handed solution might have been to impose a 20 second cooldown on Life Tap. But this would make Life Tap more rigid and less versatile. Because Life Tap affects your health, there are many situations where you need to postpone using Life Tap--or use it early--and make up for it later by Tapping multiple times.

Saturday, January 02, 2010

Disenchanting Option

The new disenchanting option in the default loot systems is interesting. Coriel is an enchanter, so it's always available whenever I do a random heroic.

In general, I like the option when doing random heroics. It's very convenient, and it saves me the trouble of handing out shards at the end. I'm somewhat sympathetic to those enchanters who feel it has taken away the profitability of enchanting, as they're probably correct. But I don't really look at professions as a means to make money, so more convenience for me is more important.

It's an interesting change, as it was a game system change that was driven by a social behavior. Enchanters were gathering unwanted items and handing out shards at the end of runs. Blizzard didn't really make this change for gameplay reasons, but to preserve a emergent social behavior that the playerbase had agreed to. This seems like a rare move for them, and I wonder if there are any other examples of social behavior that was later codified into game law.

However, I don't think the DE option is perfect. In fact, I think that the Disenchant Option should be disabled in the Group Loot system.

The DE Option is perfect for random heroics because of the forced loot system and transitory nature of the groups. You can't really trust the enchanter to stick around and hand out shards. As well, since you can't trade permanent items cross-server, the downside to accidentally DE'ing an item is low. After all, you can just run more heroics.

However, in other situations, I think the chance of making an irreversible mistake with the DE button is too high. Let's say you're in a raid that wants to use Group Loot with a "Need for main-spec, Greed for Off-spec, pass otherwise" rule. Right away you can see that someone accidentally hitting DE is going to cause problems.

It's especially annoying because Blizzard allowed Bind-on-Pickup items to be temporarily trade-able to other people in the raid. That made loot distribution a lot easier, because you could recover from mistakes. Did Deborah the Death Knight accidentally roll need on a caster staff? No problem, just trade it to the right person.

But the DE button cuts against that. It allows people to make unrecoverable mistakes with loot, which leads to drama and GM tickets. Simply disabling the DE button in Group Loot would eliminate all those problems. After all, you can still disenchant items manually if necessary.

To sum up, the Disenchanting Option is a good fit for the Need Before Greed system used in Random Dungeons, but it is not a good fit for the Group Loot system used elsewhere in the game.

Friday, January 01, 2010

PuG Tales

Actually, I don't really have a whole lot to say about grouping through the Dungeon Finder. It's pretty much been, in the words of Scott Jennings, "Bam! Loot! Dog!"

In any case, I'm sort of surprised at seeing some of the negative stories about Dungeon Finder groups going around. Pretty much all of mine--knock on wood--have been successful. No comments about low DPS, no snarkiness or elitism, just clean and competent groups.

This includes heroics run on Coriel and Valarin, as well as low level dungeons run on a couple alts. Maybe they've been somewhat quiet, but everyone says "Hi" at the start and "Thank you" at the end, and really, what more can you ask for?

There have really only been two negative happenings. In one Culling of Stratholme, two DPS ran ahead and died, so the tank ditched us in the middle of the waves. I ressed them, the DK switched to tanking gear, and we carried on until we got a new tank from the Dungeon Finder. We even made the Drake Timer (albeit just barely, with literally seconds to spare).

The only other place that regularly has problems is Gnomeregan. And that's mostly because it's Gnomeregan. Low-level, long, a confusing layout, with somewhat rapid respawns: it's a recipe for unsuccessful runs.

Even Oculus isn't that bad. Speaking of Oculus, I've never found it difficult, but I do find it a bit annoying. It's very start-and-stop. You get on drake, fly a few feet, and get off the drake. Rinse and repeat again and again. I wonder if it would have been a better instance if it had been more like Malygos. If you spent the first half running around, and then the last half entirely on dragonback, rather than switching all the time.

Still though, the Dungeon Finder has been very successful and a lot of fun for me so far.