Blog Archives

The Pre-4.3 Numbers

As I did back in June – has it really only been five months? – for posterity’s sake here is a screenshot of WoWProgress’s Firelands numbers as they stood on Tuesday, November 29th, at around 2am:

Since there is no 100% boss (but Shannox gets close), a little reverse-engineering results in a total of 45,839 guilds having killed at least 1 boss this tier. I would do a further breakdown as I did last time, but what’s the point? About 71% of every guild that started Firelands in some fashion finished it. Unlike last time around, Blizzard rolled out the content nerf before the patch hit, which obviously influences the completion rates in this bizarre way.

Speaking of last time, there were 62,405 guilds that downed at least one 1 boss in T11 content. Compared with today, that is a drop in activity of 26.55%, or 16,566 guilds that fell off the grid.

As always, the numbers get a little fuzzy if you want to look at the number of players instead of guilds. If we assume a generous 18 raiders per guild, 825,102 players have killed 1 boss in Firelands, down from 1,123,290 killing 1 boss in T11. Back in June I had what I assumed was a reasonably accurate count of all non-Chinese subs (i.e. all guilds WoWProgress tracks) at 6.5 million, but obviously that has changed in the midterm. Back then, it meant only 17.28% of players raided. Today that would be just 12.69%, but only if the overall population had not decreased as well.

To understand exactly how generous I am being vis-a-vis the 18 people per guild estimate, WoWProgress says that only 4934 guilds killed Shannox on 25m, compared with 39,861 10m kills. In other words, there are over eight (8) times as many 10m kills of Shannox than 25m of the same. That 8x figure is fairly consistent across all bosses until you hit Ragnaros, interestingly enough. In fact:

Boss 10m guilds 25m guilds Difference
Beth’tilac 39,165 4,821 8.12x
Lord Rhyolith 38,122 4,704 8.10x
Alysrazor 37,086 4,467 8.30x
Shannox 39,861 4,934 8.07x
Baleroc 38,320 4,574 8.37x
Majordomo 37,619 4,516 8.33x
Ragnaros 27,595 3,991 6.91x

If those 25m numbers don’t seem jarring to you, perhaps this will illustrate it better:

Boss 10m guilds 25m guilds Difference
The Siege of Ulduar ??? 31,993 n/a
Beasts of Northrend 86,187 58,801 1.46x
Anub’arak 84,044 52,903 1.58x
Lord Marrowgar 84,136 59,356 1.41x
Lich King 48,523 11,567 4.19x
Magmaw 60,390 4,395 13.74x
Nefarion 39,390 4,580 8.60x

I was not actually aware of the Magmaw discrepancy until just now, but… wow. Assuming that Blizzard making it difficult to differentiate between 10m and 25m kills achievement-wise doesn’t impact the accuracy of WoWProgress, this seems an armor-piercing argument that the merging of lock-outs (and possibly of gear) is not just killing 25m raiding, but driving it before us, while we hear the lamentation of its women.

While I understand the LFR system may address the casual PuG content gap, these numbers cannot bode well for the future of 25m raiding. Less than 5k guilds running normal 25m content means all that content is being made/balanced/tuned for the entertainment of less than 90,000 150,000 people. There will likely be three two times that number of players engaging in Pet Battles at any given time of day, let alone overall.

Hmm, perhaps the decision to include that as a major feature is not so incongruous after all.

OT: The pre-4.2 Numbers

I think it is a bit early for a more formal “postmortem” on Cataclysm’s first tier of content, but for posterity here is a screenshot of raiding progression as it stood at nearly 4am Tuesday morning, before the numbers could be “sullied” by the 4.2 nerfs.

Since there is no 100% boss, some reverse engineering of WoWProgress’s numbers shows that there was a total of 62,405 guilds that killed at least 1 boss this tier. A further breakdown estimate goes something like this:

1/12 – 62,405 – 100%
9/12 – 44,107 – 70.68%
12/12 – 23,122 – 37.05%
13/13 – 812 – 1.3%*

Depending on how many raiders you associate with a raiding guild (15-30), this means roughly between 589,245 to 1,178,490 players who started this tier did not finish it on Normal. WoWProgress pulls its data from NA, EU, TW, and KR servers, which comprise roughly ~6.5 million subscriptions per MMOData. This means that at the upper end (30) the raiding pool this tier is about ~28.8% of all accounts. Or, 71.2% of all subscribers did not raid, and of those who did raid, 62.95% did not kill all 12 normal mode bosses.** In this context, seven bosses in Firelands may almost make sense.

The other thing I want to mention briefly is that I expect Blizzard’s Q2 investor call to either look absolutely amazing, or completely terrible depending on timing. As you may or may not have heard, Blizzard sent out emails to existing accounts which essentially contains a free copy of the original WoW game, 30 days of game time included. Secondly, Blizzard is poised to release Cataclysm in China July 12th. Finally, and perhaps more earth-shatteringly from a subscription standpoint, Blizzard increased the Recruit-A-Friend XP bonus from 1-60 to 1-80. If you are an alt person as I am (or was, considering I have a full 10 character slots on Auch), this is about as close as Blizzard seems willing to get to letting you buy a Cataclysm character. Back when RAF originally came out, I had two instances of WoW running and essentially spent $5-10 to get a level 60 rogue, priest, and hunter (with the gifted levels) in about two weeks of leisurely play. And for that month, for all intents and purposes I was two subscriptions to Blizzard. Of course, Blizzard recouped $25 or whatever it was when I decided to transfer the RAF priest to my primary account before shutting the RAF account down.

So, basically, depending on when the Q2 sub numbers are compiled Blizzard will likely be seeing huge growth (due to 4.2 being released, dual-boxing RAF accounts, free copies of games going out, new expansion in China) or further drops depending on when the numbers are locked in for the report.

*Apparently Heroic Ascendant Council is more difficult than Sinestra based on number of guilds having killed it: 812 vs 926 (Sinestra). It might be that people were racing for Sinestra kills before the patch, but it is interesting nonetheless.

**By contrast, only 42.32% of raiders who downed Marrowgar did not also kill the Lich King. It is entirely possible we will see more 12/12 after an equivalent amount of time has passed, of course.

OT: The Final Tanking Solution

There has been a lot of discussion surrounding the systemic problems with the LFD feature when it comes to tanks, and quite frankly, a lot of truly bizarre “fixes.”

Rohan from Blessing of Kings thinks that maybe we should move from the Holy Trinity (e.g. tank, healer, 3 DPS) to something like 2 tanks, 2 healers, 2 DPS. While that seems out of the blue, apparently it is more similar to what Age of Conan was (is?) doing. Green Armadillo at Player Vs Developer thinks that since there are more DPS than the other roles, why not simply have a tank, healer, and 7 DPS? Gordon from We Fly Spitfires thinks that simply no one wants to play those roles, and so Blizzard should either make those things more fun to do or perhaps give up on the Holy Trinity model altogether. Adam from The Noisy Rogue asks why Blizzard is bribing tanks when they could simply make running heroics required for raiding by adding a (stacking) buff for completing them. Incidentally, Adam appears to hate non-pure DPS classes and wishes them pain. Gevlon from The Greedy Goblin thinks it is an “education” issue that should be fixed by somehow teaching people how to play better (or how to tank), and that no amount of bribery will work.

And to round it out, here is actually Blizzard’s official take on the situation:

We don’t feel the tanking and healing roles have any inherent issues that are causing the representation disparity, except that fulfilling them carries more responsibility. Understandably, players prefer to take on that responsibility in more organized situations than what the Dungeon Finder offers, but perhaps we can bribe them a little.

With Great DPS-Power, Comes… No Responsibility

All of these “solutions” flirt around the two problems, outlining what is really going on in the vaguest of ways. Tanking (and healing) carry more responsibility… why? Easy: tanks and healers cannot be carried. When a tank fails, by definition, the group wipes. When a healer fails, the group wipes. There can be sloppy play, for sure, like a tank not positioning the boss correctly, or not reacting to a certain ability the correct way. However, if the group is still able to rally and defeat the boss anyway, the “failure” really is not a failure. Conversely, DPS failing generally means that the DPS died, which is not that big a deal…

…until now.

Problem 1: Cataclysm dungeons are (still) too difficult for the LFD tool.

The first time I zoned into the revamped Shadowfang Keep and saw that there were actually three separate spells that a DPS could (and has had to in two cases) interrupt on Baron Ashbury, I knew I would despise this expansion. Commander Springvale, Lady Naz’jar, Rom’ogg Bonecrusher, Corla, Herald of Twilight, Ammunae, Setesh, Rajh, Corborus, (post-patch) Asaad, Foe Reaper 5000, Admiral Ripsnarl, “Captain” Cookie… all of these bosses from nearly every single heroic have mechanics that the DPS has to take care of or else the group wipes. Blizzard actually thought they could add responsibility onto the historically least responsible role and have everything work out? Nevermind the endless mob packs that need CCing along the way.

It might come as a shock, but bad players play this game. I do not subscribe to Gevlon’s “M&S” sociopathy, but there is a full spectrum of player abilities out there, and not every one is tall enough to ride this roller coaster, so to speak, especially after jacking up the scale a few more inches. You cannot keep these players out of the LFD tool though, because there are not enough “good” players to support the function – as I mentioned back in February, there has to be a critical mass of success going on for the tool to operate at a self-sustaining level. No matter what you think about the LFD tool, it is still entirely better than trying to make your own non-guild group, as otherwise you would be doing that instead of sitting in a DPS queue for 45 minutes.

I Ain’t Got Time to Bleed

Speaking of sitting around for 45 minutes…

Problem 2: Cataclysm dungeons (still) take too long.

It is bad enough that having one (or two) bad players from the LFDisaster lottery in your group can torpedo an entire heroic run. What drives the situation into an irredeemable farce that it is today is asking raid-geared tanks to slog through perfect no-wipe, no-death runs… in 50+ minutes. All for 70 Valor Points. Every other non-raid aspect of this game is designed around being completed in 30 minutes or less. TB dailies? 30 minutes. Battlegrounds? 30 minutes. Arena games? They can technically last for 47 minutes or whatever, but most games last no more than ~3 minutes * 10 games (assuming 50% win percentage), so ~30 minutes. What happened to 30 minute heroics? As I have openly opined in this space before, where are all the players asking for 5-man raids on their off-nights?

Blizzard seems to be addressing this particular problem, e.g. 70 VP/hour not being good enough, without actually addressing the equally (if not moreso) pressing Problem 1. This is why their bribing “solution” will be a short-term improvement at best: I actually do plan on running LFD more often, but only because the Reins of the Raven Lord is something I have been farming off-and-on again since TBC was relevant content. So I will either get the mount rather quickly, or I will become so frustrated (again) at the prospect of running these heroics that I would rather be farming Sethekk Halls the old fashioned way, and getting done in a fraction of the time.

Always the Last Place You Look

So what is the actual final solution to this seemingly tanking crisis? Does it involve fundamentally changing the way the game is played, perhaps with 8-9 person dungeons with two tanks or whatever? Nope. How about rating systems, or game tutorials? Bzzt!

Solution: Tune Cataclysm dungeons like Wrath dungeons were tuned.

Problem solved! Did you, or anyone you know complain about DPS queues in Wrath? Not a chance. DPS queues were 11 minutes back then. Eleven minute DPS queues. Can you imagine? I can! Despite the low queue times, I still tanked on three separate toons because it was faster and I enjoyed tanking more. How could I enjoy tanking and it’s responsibility? Because I had control over the outcome. You could be the worst player in the world and I would carry you kicking and screaming to a 15-20 minute LFD success. And then everyone could move on with their lives.

People did complain Wrath heroics were too easy. And you know what? Those people complain that Cataclysm dungeons are too easy. There is no satisfying those people, because they have such warped perceptions as to what daily group quests should consist of that placating them is a waste of time. These dungeons were new, relevant content the first few times you beat them. After that? They are farm content. Does anyone legitimately enjoy “challenge” on farm content? That seems like a contradiction in terms, does it not? If you are in heroic raiding content, does it please you to wipe on normal Magmaw? I find that unlikely. Challenge does not consist of RNG wipefests because you happened to be grouped with a stoned college student, a stay-at-home mom changing a diaper with one hand on the keyboard, a Fan of Knives bot, and/or the social bully studying to get his minor in Sadism.

Tanking and healing will always have more responsibility than DPSing. The goal should not be to “fix” that by adding more responsibility to DPS. All that will do is make tanking/healing more frustrating, because on top of what you already have to worry about, you have to worry about whether you will wipe through no fault of your own because Worgenlol of Random Server 316 did not interrupt the one-shot mechanic of a dungeon boss.

Challenging content does have a place in World of Warcraft and that place is in organized content. Grabbing five random players and sticking them into a group that will never exist again is not “organized content.” You can even still have challenging heroic dungeons for people that want it… just don’t stick those dungeons in the Dungeon Finder. The way the revamped ZA/ZG are rolling out is a bastardized version of this, but Blizzard could do it better. Have them exist as brutal 5m content for 1-2 months or whatever, and NOT be in LFD. Then, nerf them appropriately when you finally do add them to the tool. This solution actually appeases the “save the community (that we never interacted back in vanilla/TBC)!” camp, as you could get your local e-Street Cred up for successful ZA/ZG runs while anyone else who couldn’t be bothered could still get their daily 7/week VP the (now) old-fashioned way. Between the 353 gear and the doubled (!) VP gains from ZA/ZG, I do not think there would be any danger of there not being enough people running those dungeons. And when 4.2 rolls around at the end of August when it’s ready? Glide ZA/ZG right into LFD.

No matter what you thought about Wrath, the one thing it indisputably got correct was how to make a vibrant, healthy LFD community. Player activity only went down after nearly a year without new content, and the prospect of another gear-reset expansion. Cataclysm player activity is down four months after launch. No, seriously. Blizzard talks big game about the design iteration process, but as we know from Wrath they kept the same difficulty model the entire expansion. I actually have little hope that they will turn Cataclysm around before plowing into that iceberg, and what we are likely to see is another brutal raid tier in Firelands and less reasons to feel the need to log on every day. Design decisions like “let’s require endboss kills before they can get a 4pc bonus!” sound good on paper (and quells the forum trolls), but in practice it makes people less interested in ANY tier pieces. Cataclysm difficulty? Same kind of deal.