Throwdown: 4-shamans and a tank versus Mixed group

[Bigfish]

Given that we've derailed 2 threads on this topic so far, I figure we may as well have a thread actually discussing the issue than spinning off from existing threads.

First things first, I think its best to describe where we're coming from. I have 1 of each class, and I've extensively experimented with them to determine optimal output. The only roles I haven't specifically written macros for are feral druids, holy pally, ret pally, enhancement shaman, arcane mage, and prot warrior, and of those, I don't expect any to be any more complicated than what I've already done for the other classes.

The conclusion I reached through all this observation and experimentation is that the more dynamic your group gets, the more unwieldy it becomes. However, this also comes from my perspective of optimizing every class, with the ultimate goal of making every class play as close to as it would if an individual were controlling it.

Example: You can toss a fire or frost mage in to a group with the intent of gaining the +10% crit chance. This works on the level that you can get the buff fairly easily, but it becomes a good deal more complex as you dig deeper in to maximizing their potential. When classes are balanced by Blizz, they are done so with the perspective that said class is being played by one person who is capable of managing the various cool downs, DoTs, and procs. Their ideal damage is thus based on the difficulty of the rotation as it would be run by a normal player. Now, you can toss a fire mage in to a group and get your two scorches up fairly quickly, but ideal damage suggests that you should also have living bomb and be taking advantage of hot streak procs. Now you CAN spam Frostfire Bolt and keep these other spells out of rotation, but you won't be getting optimal damage.

Some may view this as a challenge, some may view it as an unnecesary overcomplication.

Expound on this situation with multiple DPS classes. If you are looking for a quick way to up your damage, macro-fiddling tends to work fairly well, and requires little initial investment. However, when you're talking 5 different classes, that investment of time and research starts stacking up. In the case of the 4 shaman groups, Fiddling with your dps macro affects all 4 characters, providing a larger buff to group DPS than you get by improving just one character.

Now in the case of everything being pre-set up and if your macros are running as efficiently as you need them, this becomes a moot point. If you're clearing content and achieving goals, optimization and external opinions be damned. If you hit a wall though, there are a larger number of potential factors to be considered in how to breach said wall if you have a multi-class group than a simple tank-4 shamans.

My position is that running a multi-class group is more difficult, and only slightly more rewarding than the simpler but largely just as effective set up of shamans. That said, when people ask for opinions on the subject, I feel it necesary to put forth both the advantages and disadvantages of the respective set ups. The answer isn't as simple as one or the other, but should be tailored to what the inquisitor's goals, aspirations, and values are.

Its not simply a matter of Synergy versus Simplicity, but how much you value either. Cross class groups gain synergy at the cost of simplicity, and core class groups do the opposite. The two aspects are not mutually exclusive, but vary depending on a number of factors.

[shaeman]

I hope I'm not derailing the thread here, but your initial post got me thinking.
Your key issue is that it is difficult to maximise the damage, to work in procs on different characters.
I can understand this - it's how i've always felt about things when i first considered multiboxing 3 70's I already had when I first started.

A lot of that is related to everything being driven from a main and not being able to micromanage watching the alt screens for these procs,
whilst healing and tanking and spamming the dps key.

What if we could come up with a way to make the handling of these damage enhancing procs much easier.
I'm thinking of perhaps the creation of an addon that you can configure on your main that alerts you to a specific proc on a specific alt.

Using your own example - hot streak procs on your mage - a message appears on your main alerting you to this, perhaps even with a distinct audio) and you hit a key that is only mapped for the mage to trigger the correct spell cast (perhaps even configurable so that it tells you what key to hit).

Is this even possible with the blizzard api - I dont know, My gut feel is that it may be possible. Perhaps I could even research how to do this.

It would require knowing when the proc occurs (may be possible as some things alert you to buffs running out)
It would require relaying that from the alt to the main
It would require the main to display that to the user.

Sorry - bit of a brain dump, something that's been sitting in the back of my mind for ages.

[Ken]

The conclusion I reached through all this observation and experimentation is that the more dynamic your group gets, the more unwieldy it becomes. However, this also comes from my perspective of optimizing every class, with the ultimate goal of making every class play as close to as it would if an individual were controlling it.

I agree that a group gets more complex to manage when adding more different classes, but this is not a difficult problem to overcome and it basically just requires some effort to solve.
So yes, it could be a problem, but it's not a permanent one. Especially because abilities can easily be grouped into categories like: long-term buffs, short-term buffs, single target dps, multi-target dps(perhaps with separate ones for PvP/PvE), etc.
When your keybinds are grouped like this, it becomes relatively easy to fit any character into your setup. The only thing left is to find the correct cast sequences, but they can be easily found online or figured out by yourself.

Example: You can toss a fire or frost mage in to a group with the intent of gaining the +10% crit chance. This works on the level that you can get the buff fairly easily, but it becomes a good deal more complex as you dig deeper in to maximizing their potential.

Yes, it requires more work to maximize the potential of your group('cause you have to do it for each class), but this is only a 1-time issue and depending on whether you like to macro or not, it might just as be a positive argument as a negative one.

When classes are balanced by Blizz, they are done so with the perspective that said class is being played by one person who is capable of managing the various cool downs, DoTs, and procs. Their ideal damage is thus based on the difficulty of the rotation as it would be run by a normal player. Now, you can toss a fire mage in to a group and get your two scorches up fairly quickly, but ideal damage suggests that you should also have living bomb and be taking advantage of hot streak procs. Now you CAN spam Frostfire Bolt and keep these other spells out of rotation, but you won't be getting optimal damage.

I agree that you can't get optimal damage with cast sequences, but that goes for all classes, not only for multiclass usage.
Even shamans with lavaburst have this problem: when constantly casting onto 1 and the same target it's not an issue, but when you have a situation with cast interruptions and multiple targets, lava burst and flame shock will have to be micromanaged just as well.

Some may view this as a challenge, some may view it as an unnecesary overcomplication.

Expound on this situation with multiple DPS classes. If you are looking for a quick way to up your damage, macro-fiddling tends to work fairly well, and requires little initial investment. However, when you're talking 5 different classes, that investment of time and research starts stacking up. In the case of the 4 shaman groups, Fiddling with your dps macro affects all 4 characters, providing a larger buff to group DPS than you get by improving just one character.

At one side you're right, but at the other side: you only gain a minimal set of new abilities each time you gain a level, so it's rocket science to figure out an enhancement for your macros. It is indeed more work, but it's also just a temporary issue.

Now in the case of everything being pre-set up and if your macros are running as efficiently as you need them, this becomes a moot point. If you're clearing content and achieving goals, optimization and external opinions be damned. If you hit a wall though, there are a larger number of potential factors to be considered in how to breach said wall if you have a multi-class group than a simple tank-4 shamans.

My position is that running a multi-class group is more difficult, and only slightly more rewarding than the simpler but largely just as effective set up of shamans.

Agreed.

That said, when people ask for opinions on the subject, I feel it necesary to put forth both the advantages and disadvantages of the respective set ups. The answer isn't as simple as one or the other, but should be tailored to what the inquisitor's goals, aspirations, and values are.

Its not simply a matter of Synergy versus Simplicity, but how much you value either. Cross class groups gain synergy at the cost of simplicity, and core class groups do the opposite. The two aspects are not mutually exclusive, but vary depending on a number of factors.

So, what I think should be considered is:

Multiclass benefits:
- more buffs (direct or indirect through procs and/or talent tree)
- easier to gear when classes are chosen carefully
(- more macro work (can also be disadvantage))

Single class benefits:
- easier to configure
- less work to put into macros

So overall if you are just learning to multibox, it's probably better to start with less different classes. However, I think the additional buffs that a diversified group brings can be a benefit overall.

Maybe we should make a list of benefits/disadvantages in the topic start?

[Bigfish]

Multiclass benefits:
- more buffs (direct or indirect through procs and/or talent tree)
- easier to gear when classes are chosen carefully
(- more macro work (can also be disadvantage))

Single class benefits:
- easier to configure
- less work to put into macros

So overall if you are just learning to multibox, it's probably better to start with less different classes. However, I think the additional buffs that a diversified group brings can be a benefit overall.

Maybe we should make a list of benefits/disadvantages in the topic start?

First off, any of the shaman boxers out there have a concise list of their totem farms? I haven't looked at them in some time, and patches being what they are, I'd rather not cite old or innacurate information.

As far as specific advantages and disadvantages go, I'm kind of hesitant to put a list down in stone (er, electrons) as different people value different aspects. Macro writing and gearing may well be an advantage or disadvantage depending on how you approach it. Where I'm curious is what the differentiating factors in between people who run all shamans and multi-class teams.

As a basic and quick run down of marginal benefit, lets assume a 5 shaman group on relevant contents. You get your full range of totem farms, but no tanking power, so the group faces potentially deadly incoming fire. So take out a shaman, and add in a tank. The tank adds their synergy (kings/wisdom, ebon plague, whatever), greatly increases survivablity. At this point, you face the question of 4 shamans DPS/spothealing versus 3 straight DPS shamans and a Resto. You lose some DPS, but maintain the totem farm. Switch the resto out for a dedicated healer, and you get their associated buffs at the cost of their share of the totem farm.

Now at this point, I feel it necesary to point out just how much of the totem farm we're losing: a potential fire elemental on difficult fights, a magma totem on trash, a grounding totem against casters, etc, so I don't think its fair to simply dismiss the totem aspects we're losing to gain buffs. It may still be better to have a healer of another class, but we need to keep in mind what is being lost.

Back to the group, we're now at 3 shamans, a healer, and a tank. At this point, there is a virtual cornucopia of prospective switches you can add for caster DPS buffs, but the totem farm keeps getting thinner and thinner. Is the trade off of various buffs worth it? Yes and no, depending on who you ask.

I'm starting to wonder how much class balance actually figures in to most of this, versus the resources to allocate to make it successful. A question to anyone reading this thread: Do you get your groups to cap and then play with their composition, or decide on a composition before hand and stick with what you level?

[Ken]

So, to prove that a well-chosen multiclass group is easier to gear than a 4shaman+paladin group, this is a set of calculations that check how much chance there is to get a useful drop for an item for the first drop.

Considerations:
Here are some reasonable assumptions to make the calculations easier. Feel free to comment for corrections.

1) Drops in heroics drop this kind of gear:
- tanking plate(has def/dodge/etc)
- tanking leather(has def/dodge)
- pure melee leather (AP-related)
- pure melee mail (AP-related)
- pure melee plate (AP-related)
- melee leather with int (has AP-related stats with int)
- melee mail with int (has AP-related stats with int)
- melee plate with int (has AP-related stats with int)
- cast cloth
- cast leather
- cast mail
- cast plate
We assume that cast gear is useful for both DPS and healing.
Total: 12 types of gear.
2) When an item drops (world drop and boss drop), any item category has equally much chance to drop. So when an item drops there's 8.33% chance (1/12) that it belongs to ANY specific category.
3) Each class is assumed to have 1 specific role and we assume that healing gear can also be used for cast dps gear and vice versa.

Calculations:
1) 4 shamans + 1 paladin tank:
- Shamans can wear cloth, mail and leather, which covers 3 gear types out of the 12 that drop.
- This leads to the chance P for shaman gear: P(shamangear) = 3/12 = 0.25 (25%)
- A paladin tank is assumed to wear plate with defense stats, so his gear only is covered by 1 gear type out of 12
- This leads to the chance P for paladin gear: P(paladingear) = 1/12 = 0.08333 (8.3%)
- shamangear and paladingear don't share elements, so mathematically this becomes: P(shamangear or paladingear) = P(shamangear) + P(paladingear) = 0.3333% =~ 33%
Conclusion: The first item that drops for a 1paladin-4shaman group has 33% chance to be of use to that group.

2) 1 paladin healer, 1 hunter, 1 mage, 1shaman and 1 druid tank:
Since some classes share gear, I will list the useable gear first:
- pure melee leather (AP-related) -> useable for hunter
- pure melee mail (AP-related) -> useable for hunter
- melee leather with int (has AP-related stats with int) -> useable for hunter
- melee mail with int (has AP-related stats with int) -> useable for hunter
- tanking leather
- cast cloth -> useable for mage/shaman/paladin
- cast leather -> useable for mage/shaman/paladin
- cast mail -> useable for mage/shaman/paladin
- cast plate -> paladin
This totals 9 usable gear drops.
So the chance of a useful drop becomes 9/12 = 75% chance to get a useful item on the first drop.

Conclusion:
The first item drop for a 4shaman+1paladin team has a 33% chance of being useful, while the first item drop for the mixed group has 75% of being useful.

Appendix:
On following drops, the calculation becomes a lot more complex, because you have to consider for each class that the gear slot might be taken and you have to consider how many classes that can use the newly dropped gear.
For the shamans this means that consecutive cast gear drops are useful right until the 4th drop of any specific slot-related drop.
For the multiclass group this becomes a lot more complex to calculate.

I'll add more calculations later in relation to consecutive drops.

[edit] Multiclass listing

[Dorffo]

First off, any of the shaman boxers out there have a concise list of their totem farms? I haven't looked at them in some time, and patches being what they are, I'd rather not cite old or innacurate information.

Default (staggered in this order usually):
heal heal heal heal (not glyphed = 187 x 4 a pulse)
wrath of air, windfury, grounding, grounding
str of earth, armor, earthbind, earthbind
tot of wrath, searing searing searing (or ToW + 3 magma)

Gundrak -
adds poison cleansing totem and nature resist

Culling of Strath -
more groundings less windfury
Magmas to deal with wee skellys

Pvp:
heal heal cure poison cure disease
tremor tremor tremor grounding
Wrath of Air, grounding, grounding, grounding
Tot of Wrath, searing, searing, searing (unless dropping magmas in a clump of folks)


Mostly use the default farm with modifications being - tremors for Dred in DTK, resists as needed (fire/cold resist in the mage fight for nexus etc).

[Hachoo]

I think most people here know that I am very "for" same-class groups, as far as maximizing realistic potential. IMO, part of this is because cast sequences are just a drain on DPS, it will net you a much higher DPS to manually cast spells than using castrandom or cast sequence. Doing this for even 2 separate DPS classes is nearly impossible, for 4 there is no way it could be done efficiently - you'd have to rely on castsequences or castrandoms for sure, which will lower your potential DPS.

Personally I think most people can agree that you should multibox whatever is the most fun. That being said I will say this: Regardless of the fact that having 4 different "synergistic" classes seems to have higher potential, realistically I doubt anyone will ever get as much real performance out of a group of separate classes than people are getting with 4 shamans or 4 druids, etc. Proving this is very difficult, but there are a lot of shaman or druid boxers out there, and a lot of mixed class boxers out there, and if you look at self performance people are quoting, I mostly see DPS for mixed class groups in the 1500-2200 range, and there are a lot of all shaman and all druid dps groups doing >4000 dps (or even if you remove bloodlust > 2900 dps).

Part of the reason for this is that, to get the numbers above with an all shaman group at least, you really have to do your casts manually. A castsequence or castrandom will just not net those numbers compared to manually hitting FS, manually hitting LVB, spamming LB for awhile, then manually LvB, manual FS, etc, repeat. Doing something like this for 2, 3, or 4 different classes would be impossible - for one there wouldn't even be enough buttons on a keyboard but even if there were you couldn't hit that many at a time in the right sequence without breaking something or messing up

So, I still say for min-maxing, 4 shamans or 4 druids is the way to go (I'm partial to shamans), but again, play whatever is fun.

For gear - for me thats been a non issue - there is only maybe 2 pieces of gear I have on my shamans that I got from instance drops, and getting 4 of each piece happened way faster than getting my 60 badges for my t7 gloves. Most "good" gear for multiboxers without raiding is either crafted or from badges, so heroic drops isn't really a big deal at all.

This is just my 2 cents of course.

[Bigfish]

I mostly see DPS for mixed class groups in the 1500-2200 range, and there are a lot of all shaman and all druid dps groups doing >4000 dps (or even if you remove bloodlust > 2900 dps).

Idealized numbers pulled from DPS peaks aside, I can pull 2900 on several of my classes. It's largely a matter of gearing and warm up time. Not to mention I suspect that magma totems play a significant role in long term DPS statistics for 4 shaman groups to the point I question how much they boost dps. A tick every 2 seconds on 3+ mobs will boost your dps significantly, which isn't to say I want to discount them from DPS statistics, but if the shamans themsevles are only putting out 2300-2400 DPS and Totem of broken AOE (seriosuly, almost every other AOE requires channeling or GCD spamming) is picking up the rest, it says something about the validity of your argument that sequencing macros are and always will be shit.

[Ken]

I think for the sake of argument you shouldn't calculate the 1/12 range. I think you should base it on if "x role" gear drops, % of chance its useable"

If you actually follow the calculation and read what I say, then you would have seen that the 1/12 range is only the first step and only concerns the item drop rate, unrelated to the class it can belong to. Further down in the calculation you will see that shaman casters has a chance of (1/12)+(1/12)+(1/12) chance of getting a useable item since they can wear cloth/leather/mail caster gear.
It's in the text!

I suggest this because its a RNG thing on the drops. You can't say 1 out of 12 times you'll get "x" item - because in the real world - I've seen the same damn item drop 10 times in a row on consecutive boss kills.

You're only talking about boss drops. My calculations were based on world drops + boss drops together. I actually stated that.
Further more: The fact that you have had the same item drop for 10 times in a row, does not mean that the drop isn't random. Especially considering that a boss drops only a very small amount of items.

Further more: "because in the real world - I've seen the same damn item drop 10 times in a row on consecutive boss kills" is the same kind of logic as stating that the Invisible Pink Unicorn exists, because you asked for rain and the next day it rained outside.

So your % of chance an item can be used is invalid, because the item that drops is based on RNG. However the "x role" can use "y" item is not random - you either can or cannot use it.

I don't think you understood what I wrote down.

The drop % and RNG isn't a factor.

Let me put this simple: No.

[Ken]

[...]
Switch the resto out for a dedicated healer, and you get their associated buffs at the cost of their share of the totem farm.

Now at this point, I feel it necesary to point out just how much of the totem farm we're losing: a potential fire elemental on difficult fights, a magma totem on trash, a grounding totem against casters, etc, so I don't think its fair to simply dismiss the totem aspects we're losing to gain buffs. It may still be better to have a healer of another class, but we need to keep in mind what is being lost.

You don't need to keep in mind what is lost, you need to compare what is lost with what is gained. Big difference.

Back to the group, we're now at 3 shamans, a healer, and a tank. At this point, there is a virtual cornucopia of prospective switches you can add for caster DPS buffs, but the totem farm keeps getting thinner and thinner. Is the trade off of various buffs worth it? Yes and no, depending on who you ask.

Yes or no depending on what the calculations say. In the end, what people 'feel' is irrelevant, but calculating it is difficult, because some talent builds are more effective in specific combat situation (e.g. single target versus multi-target).

I'm starting to wonder how much class balance actually figures in to most of this, versus the resources to allocate to make it successful. A question to anyone reading this thread: Do you get your groups to cap and then play with their composition, or decide on a composition before hand and stick with what you level?

Overall, I'm confident that classes are more or less balanced DPS-wise. Where you lose DPS from losing an elemental, you gain DPS from another skill, perhaps even in another form. (or the same form, cause mages can have a pet too).

When levelling a new group, I chose skills and abilities that I need and then start forming a composition based on what each character adds.