Avian Gamers Network

9 Classes * 3 major talent specializations per class = 27 Specialized talent variations (assuming one talent tree is most heavily emphasized)

4 classes capable of healing (Priest, Druid, Paladin, Shaman)

4 Talent trees specializing in healing (Resto, Holy, Holy, Holy?)

Assuming random assortment yielding even distribution of classes and talent specializations, 4 out of 27 players will be a healing class specializing in healing.

Smallest dungeon size = 5 man content
1st group = 1 healer + 4 others => 5 players
2nd group = 1 healer + 4 others => 10 players
3rd group = 1 healer + 4 others => 15 players
4th group = 1 healer + 4 others => 20 players
5th group = 1 non-healer + 4 others => 25 players
6th group = 2 people sitting in LFG queue => 27 players

But what about tanking?

Again, assuming random assortment yielding an even distribution of classes and talent specializations, with 3 classes capable of primary tanking (accoring to class description) (Warrior, Druid, Paladin), and each having one talent tree directed toward specializing in mitigation and avoidance, 3 out of 27 players will be specialized in tanking. Those that are not tanking or healing will be specializing in dealing damage or whatever else is availble to them.

Our new figures for 5 person content becomes:
1st group = 1 healer + 1 tank + 3 dps => 5 players
2nd group = 1 healer + 1 tank + 3 dsp => 10 players
3rd group = 1 healer + 1 tank + 3 dps => 15 players
4th group = 1 healer + 1 non-specialized tank + 3 dps => 20 players
5th group = 1 non-specialized healer + 1 non-specialized tank + 3 dps => 25 players
6th group = 2 dps sitting in LFG queue => 27 players

This means that based upon class design and composition there should be on average 25 of 27 players capable of being in an instance at any given time (not counting level variances). These 25 players will make 5 groups.

25 of 27 players able to be in an instance. 92.6%
15 of 27 players able to be in an instance with an optimal group. 55.5%
10 of 27 players able to be in an instance, but with a sub-optimal group. 37%
2 of 27 players unable to group at all in an instance. 7.4%

Due to class composition, even with a perfectly even distribution of classes and specializations, slightly more than half of the player base is able to run instances with an optimized group. Over a third of the player base must run instances with a group that is not well-suited for the task. Seven percent of paying customers are unable to run an instance.

Again, these numbers assume that over 8 million customers, the players average out to equally choose each class and talent specialization, i.e. there is a 1:1 ratio of holy priests to BM hunters. This of course, is not likely to hold true as there are probably fewer protection tanks and holy priests in comparison to other classes. This means that there will be fewer than the projected 55% of the player base capable of running 5-man instances with an optimized group.

What about 10 man instances? Assuming that the desired optimal class make-up of a 10 person instance is equivalent to two 5 person instances and assuming the same criteria as above we find these numbers:

1st group = 2 healers + 2 tanks + 6 dps => 10 players
2nd group = 2 healers + 1 tank + 1 non-specialized tank + 6 dps =>20 players
3rd group = 7 dps unable to group => 27 players

In this scenario, we have some slightly different numbers

10 of 27 players in an optimized group - 37%
10 of 27 players in a slightly less optimized group - 37%
7 of 27 players unable to group - 26%

You may be thinking that of those 7 people, perhaps 5 of them could go and run a 5 person instance, however, these 5 will be using a non-specialized healer and a non-specialized tank.

Now, what about the big 25 person raids? Many people have been posting for a long time asking about what class composition should be, and many of these people speculate that quite a few healers should be required.

Let's assume that a 25 person raid is equivalent to five of the 5 person groups.

1st group = 4 healers + 1 non-optimized healer + 3 tanks + 2 non-optimized tanks + 15 dps => 25 players
2nd group = there is none.

25 of 27 players able to be in the 25 person raid - 92.6%
2 of 27 unable to group - 7.4%

So for a 25 person raid, a raid leader should not expect any more than 4 healing specialized players and 3 tanking specialized players. However, she should expect an additional healing capable class player, and 2 tanking capable class players.

So a 25 person raid composition would be well optimized (again assuming even class/talent spec distribution) with:

Healing (4 slots) = 1 holy priest, 1 resto druid, 1 resto shaman, 1 holy paladin
Tanking (3 slots) = 1 Prot warrior, 1 feral druid, 1 prot paladin
Non-optimized for healing/tanking (3 slots) = 1 arms warrior, 1 ret paladin, 1 elemental shaman
DPS = 1 shadow priest, 1 balance druid, 1 enhancement shaman, 1 fury warrior, 3 mages, 3 warlocks, 3 rogues,
2 hunters

leaving out a hunter and a discipline priest (or any 2 who dont make the cut for whatever reason)

Although, in all honesty the discipine priest is likely to get in the raid for healing as opposed to one of the non-optimized class slots.

What really concerns me about this math is that due to class design and the assumption that all talent specializations will be equally represented by the player base, Blizzard has created a game where only about half of its customers will have full access to the easier group content on a regular basis. Another third will have access, but with increased difficulty.

Of the more difficult 10 person content, only a third of paying customers get the opportunity to have the chance of good success on a regular basis. An additional third will be able to experience this content, but with increased difficulty. A full fourth of the player base will not be able to experience 10 person content on a regular basis.

Oddly, nearly 93 percent of the player base will be able to experience the most difficult 25 person content. However, since many of those are not seeing the 10 person content on a regular basis, there will be some degree of gear discrepancy which may hamper game play experience.

Wow.....just......wow......

2 things come to mind:

1. Just how much spare time do you have to sit and play with statistics?
2. You're information looks very accurate, but to what end are you hoping for?

Hrmm, right after lunch, getting sleepy, must not think.....hehe....

Too long to read, at least not without putting what the purpose of the post is at the top. That's my 2 cents.

_________________
"I built me a bomb and dropped it on every MFer who got in my way!"

yeah, i'm a little shaky on the purpose of it as well. I think in the end, it's an analysis of general game design as directed toward player populace. I definately need to put some thought into that. This may come from a few years back when some friends of mine and I were working on our own MMORPG, and as it turns out, I think WoW could be better designed by talentspec/class balance to allow more feasible 5-person groups.

Oddly, i put a slightly better formatted version up on the raids and dungeons forum, and it got deleted with a minute

LOL... they deleted it!?!?

WoW is a great game... and Blizzard isn't fond of sharing any of there secret recipe (can't says that I blame them). I have always wanted to know the tech details of what they run there servers on, in terms of hardware, softare, bandwidth, etc... but I have a feeling I may never know.

_________________
***************************************
Stats are for losers, the final score is for winners.

Rift/Gnarlwood: Nostro - Mage (Pyro/Necro/Chrolo - mancer) lvl 29
Rift/Gnarlwood: Stilgar - Cleric lvl 8

WoW days - Aegnor/Stilgar