View Full Forums : A DPS model


Erianaiel
12-07-2007, 05:34 AM
I did one for healing once and while pointless it was fun so here I present:

A DPS model


First of all we must point out that there are (currently) two types of fights
in the game: trash and named.

The first type of fight typically lasts less than a minute, frequently closer
to half that. The second lasts for several minutes. This is an important
distinction when it comes to caster DPS, and to a much lesser extent to melee DPS.

The second important distinction is between caster and melee DPS. Casters have
their DPS naturally restricted by their mana pool. I.e. they can do a finite
number of attacks within a given time and after that they have to wait for at
least a minute before they can contribute again (in some zones it is over 5
minutes or even longer). Melee DPS is not similarly restricted, only by the
number of swings per second. Certain skills do cost a limited resource for them
as well, but running out does not negate all their DPS.

In general, for the health of the game, it must not make a difference if DPS
is contributed through melee, casting or a combination of those two (e.g. Mages)
I.e. over a longer period of time the DPS of a specialised melee class must be
more or less the same as that of caster classes, otherwise one of the two DPS
modes is made obsolete.
This naturally means that caster DPS is burst in nature compared to melee. This
may be alleviated by the way spells are set up, but that enforced minute of down
time must be calculated into the overall DPS over a, say, 10 minute period.
We can also say that if a DPS caster goes for a sustained mode where the mana
used and recovered naturally are in balance their DPS should be roughly the
same as that of pure melee DPS classes.

Within both modes the -methods- of dealing damage may lean towards burst or
towards sustained. E.g. nukes versus dots. There is no clear similarity for
the melee DPS, but if they were made to be that way poisons would have been
a good candidate to provide sustained melee DPS. As it is the method of their
application is too cumbersome to consider them a viable DPS method.

The fact that caster DPS is burst in nature means it tends to be overpowered
in short fights. Casters can front load their DPS, and use the between time
to recover. A melee can only provide DPS when there is an enemy to fight and
unless there is one at all time then there will be downtime that reduces their
DPS, but this will not have an effect on casters since this downtime is already
compensated by their higher base damage. So this means that either the melee
DPS classes must be able to produce burst damage as well, or the sustained DPS
of casters should be a little below that of melee classes.
As the numbers get bigger it also leads to aggro issues since the burst damage
is so high that a warrior can not overcome its effect. This leads then to the
situation where casters must intentionally hold back for the first part of a
fight to allow the warrior to grab aggro before starting their contribution.
This means they must deliberately lower their DPS, but that has the unwanted
side effect of either lowering their DPS (and thus meaning a melee class would
be more effective) or their burst damage must be even higher (but that would
just make the problem even bigger).

The only clear way that would have mostly avoided (or at least postponed) this
problem is if rather than giving higher level spells and enemies simply bigger
numbers they would have mostly gotten better defenses and resists and access
to more varied ways to mess up the players' tactics. It would have kept buffing
and debuffing a more important part of the game, it would have made the utility
classes more viable and it would have slowed down the numbers inflation a lot,
giving the developers a slower development path and less problems with percentages
that as the base numbers increase tend to become really big differences.


DPS is basically controlled by 3 variables: damage, cast time and recast time.
Actually there is a fourth variable of duration time.

Damage times duration gives the maximum points of damage dealt. For nukes and
melee attacks duration is 1, and the damage is dealt all at once.
Cast time (plus the fizzle time where all attacks are blocked) is how quickly
the damage is applied obviously and for melee attacksthis is instantly (i.e.
there is no time involved in the swing or stab action that the game can use
to control DPS, unlike casters who can be interupted during the casting time).
Recast is how quickly after the initial attack it becomes available again.
Cast plus recast gives the total cycle time, but there is an important distinction
here in that a short cast time allows to chain more attacks together within a
given amount of time. Shorter cast times turn attacks into burst damage. 15
attacks with a 2 second cast time can be chained together in 30 seconds. Only
2 attacks with a 15 second cast time can be chained in that same time period.
This is assuming though that the recast times in both cases are 28 seconds for
the short cast and 15 seconds for the long cast time spells (i.e. both have a
cycle time of 30 seconds).
This is not necessarily a bad thing if the 15 second attack is made 7.5 times
as powerful as the 2 second one. The problems here again are first the fact
that a long cast time takes a DPS class out of the action for a long time where
a short cast time provides, optional, flexibility. The example given here is
obviously silly since there is no way to have 15 different attacks but it does
illustrate the principle. In the example of the longer cast though two potential
flaws show up. First the amount of damage in these attacks is staggering and
they would generate an immense amount of aggro. Second, if the fight lasts not
30 but 2 seconds, the DPS in the long cast time case would be halved while the
short cast time would only lose about 7pct.

The way melee attacks are set up makes them very limited. They have no cast time
so they always are maximum burst, their damage is balanced by their recast time
giving the developers fine control over the amount of damage done, but it also
rules out any tactical choices. Basically it is a matter of press the attack
button and the whole thing goes on autofire. That is their DPS contribution so
to give melee DPS something to do they get skills that act like spells a little
in that they have to be activated manually. Very few of them have a tactical
use (though of course in the case of casters the number of tactical choices
tends to be rather limited as well, but they still have more choice in what
spell to use to best respond to a situation because they have more methods of
dealing damage: direct and through different resists, over time, area of effect,
multiple targets, point blank, attacks with secondary effects and the variety of
utility spells that are not directly damaging).


* * *

Now, after all this rambling the actual model for DPS

Caster DPS should focus on burst over sustained. To avoid problems with aggro and
to increase flexibility they should generally have shorter cast times on their
spells, the damage per cast should go down a little to compensate so the overall
DPS does not go up. Some classes could still maintain a couple of 'big booms', i.e.
slow casting high damage spells, but those should be the exception rather than the
norm. Spells to use to punch through self-heals.
Within that archetype focus can be on burst or sustained DPS. Burst classes get a
greater variety of fast casting direct damage spells that they can chain together.
Sustained classes focus on more efficient lower damage spells, but to maintain their
overall archetype this means they primarily work with dots. (the problem here is the
ramp up time and the duration of the dots. If they are expected to stack 4 dots and
the average fight is 42 seconds then the duration of some of those dots would approach
that of direct damage, so instead they need a wide variety of duration and damage that
allows them to pick the combination that is most optional for the situation).
In addition to the burst versus sustained distinction we also have the single target
versus group distinction. I.e. a burst caster class can focus on area of effect and
multiple target direct damage spells, or on single target ones, just as a sustained
caster can have single target dots or group versions of that type of spells (that
would be insect and poison clouds I think, maybe infectious diseases as well).

Finally caster DPS can be distinguished by the amount of debuffing they can do for
their own spells (this affects to an extent the degree by which they can solo). However
if the game design is to have more specialised buff and debuff classes then pure DPS
casters should have very little of this aspect. It also makes teaming up much more
mandatory, so this is a very fine line to thread.


Melee DPS needs much more work because of its current somewhat one-dimensional nature.
The only way it can be expanded right now is along the 'more damage' axis as there is
very little tactical choice involved and thus almost no other way for the developers
to affect gameplay and develop new strategies. This leads to the problems mentioned
above with ever bigger numbers and burst damage casters, amongst others.

First of all melee combat needs to get more dimension. To keep things simple it would
be along the same line as casters: burst versus sustained and single target versus group.
As an archetype the melee DPS would still shift towards the sustained DPS end of the
spectrum, meaning they can do less to temporarily increase their DPS output at the
expense of a longer downtime, but they can continue pretty much indefinitely if needed.
Still, a certain degree of downtime is necessary to prevent the game being split into
caster-only and melee-only spheres.
First of all the game must get rid of the auto-attack. Unless a players -does- something
his or her character is only defending itself (using shield and weapon to block incoming
attacks but not attempting to counterattack).
Second, melee attacks need to get more methods. Right now the stab/slash/bash distinction
is not a method but a specific resistance to overcome. One way to increase the tactical
choices by melee fighters is that different types of attack have a lingering effect on
both attacker and defender, in creating localised defenses and weaknesses. I.e. bringing
up a shield to block an overhead attack also means a vulnerability on the lower body and
prevents any 'high' counter attack (and may reduce general defense because of the reduced
perception as the shield is in the way). The attacker at the same time is vulnerable to
attacks from the side because of the position of the sword. While this detailed example
is likely too complicated for everquest, the principle of it can still be applied. By
making attacks positional to a limited extent each of them can create positional defenses
and weaknesses both in attacker and defender. Melee fighters get a range of attack skills
that they must manually choose to direct the battle, and to respond to the enemy attacks.
Like for casters these attacks cost varying degrees of endurance, but melee fighters will
only use this up very slowly, but for prolongued fighting (either single target or by
chain pulling) they can get too tired to continue (or to use the more fancy counters).

Within the archetype then there are fighters who focus on slower, sustained, fighting
techniques. The exact method can vary with each class, but typically this would be using
poisons (and a less clumsy method of applying them!) and causing small wounds, followed
by fighting defensively to wait out their enemy to weaken.

There are several types of attacks that could be applicable to multiple enemies at once.
E.g. a broad sweeping stroke could hit more than one enemy in front of the fighter, as
would a whirling swing. And throwing a handfull of darts at several enemies in front of
the fighter also could be interpreted as a single attack (i.e. not requiring separate
targetting).

Similarly to support casters there would also be room for support melee fighters. These
would do less damage themselves, but would weaken their opponents (through misdirection,
blinding, traps, tangling or otherwise hampering weapons and so on).
Both support roles could be cross archetype (i.e. casters could create certain melee
vulnerabilities, just as fighters could create weaknesses to certain types of magic. E.g.
a fighter dousing an enemy in oil ...


* * *

That is about it that I can think of right now. I am aware that this is just a 'would
be nice' type of post, but I would still like to hear what others think about it.


Eri

Madie of Wind Riders
12-17-2007, 03:46 AM
Interesting perspective. I would have to say I think the Melee classes would like the suggestions you have made - especially being able to attack more than one target at a time.

As for your suggestions regarding caster classes - I think "basically" that is how the game is set up now. With Wizards and Mages in the burst DPS category and Necro's and Druids in the sustained category. I think I would need more specifics to really understand how you would change the current system for casters to make a real difference.

Good post!! Thanks Eri :)

Erianaiel
12-17-2007, 03:18 PM
Interesting perspective. I would have to say I think the Melee classes would like the suggestions you have made - especially being able to attack more than one target at a time.

As for your suggestions regarding caster classes - I think "basically" that is how the game is set up now. With Wizards and Mages in the burst DPS category and Necro's and Druids in the sustained category. I think I would need more specifics to really understand how you would change the current system for casters to make a real difference.

Good post!! Thanks Eri :)

*smiles* Thank you.


I have not been thinking a great deal about fitting the current classes in Everquest into this model since it is theoretical and only for fun. It would change the existing game in too many ways to be viable (and it would have to be changed from the ground up, not just tacked on in the next 5 levels).

For casters we have 4 classes: Wizards, Mage, Necro and Enchanter.
We also have 4 'roles' within the archetype. The problem of course is that Enchanters can barely be considered DPS currently. (technically they are debuffers, providing a 100pct damage mitigation on any mezzed monster).
Also, Druids traditionally were capable, but not specialised, DPSers, a role they have been losing a lot of ground on lately.

All casters must, by the nature of using mana to power their spells be burst, but within that we have classes that specialise in burst and those that specialise in more sustained burst (i.e. one will do well on short fights but run out of steam on long ones, the other will not be optimal in short fights but can be very effective on extended fights).

Necros, through their DoTs and theme would be the obvious candidate for sustained DPS and multiple target roles (e.g. their DoTs become poison clouds and they have damage auras)
Mages pets are essentially a controlled and semi-permanent DoT. This puts them in the single target and sustained DPS category/
Wizards already are in the burst DPS category but we can put them in the single target (nukes) or multiple target (pillars etc.) category, depending on which type of spells they will focus on.
This leaves a missing caster DPS class. Enchanter DPS is done through charmed monsters, which could be considered burst and single target (though they need a lot of changes to be considered a DPS class), and that would mean Wizard logically would fall into the burst and multiple target category.
If we want to consider Druid DPS (which from the writeup on the site Sony no longer does), they are heavily split between single target burst and single target sustained. There also is some multiple target burst ability but it does not seem to be the core spell lines of the class.


For the melee DPS role I have postulated a similar distinction of burst and sustained, though both would be below the levels available to casters (to compensate for the fact that melee fighters do not need to rest as frequently and long as casters do). I also suggested a single or multiple target focus in their combat styles.

Problem here is that there are 6 melee DPS classes, though one of them (Bards) is an odd one.
Berserkers are, by their discription burst damage, as are rangers. The choice of weapons suggests that Berserkers could be multiple target (i.e. they get short range sweeping 'cone' attacks) while rangers are more single target (especially the bow, but they could limit their weapon choice to ones that are not suitable to multiple target).
Beastlords should be considered sustained and single target. Much of their DPS comes through their warder which by its nature is single target, and being a pet it is a summon once and it will provide DPS indefinitely source of damage )
Rogues can be imagined as either single and multiple target. With the exception of their backstab they would be sustained DPS (and the flavour text has them slipping through combat, striking as the opportunity arises and using poison and other tricks to weaken their opponent). So if we do away with backstab as a separate mechanism (just make positional damage available to all classes, but rogues better capable of recognising it when it is possible), and increase the ease of use and potency of poisons and other tricks.E.g. they can create poisons that last a given number of procs for a variety of effects like damage, weaknesses, slowing, confusion etc, and they get other things like smoke bombs and other misdirectons to change the situation of the fight. It would still make them sustained and single target i.e. the same niche as the beastlord but a much more varied class to play and use in a group (their secondary role would be to create and use temporary weaknesses in their enemies).
That would leave monks as sustained multiple target DPS. With their shuriken, stunning blows and other skills that is not totally out of character for their class description I think.

Bards, as I said are an odd one since they can do everything. They are force multipliers, which is a usefull class to have in a group but makes for pretty poor solo ability (if there is no force there is nothing to multiply). Their DPS role would naturally fall into the sustained and multiple target niche.


For healers we have three classes, and essentially three different types of healing. Clerics would specialise in reactive healing (i.e. they restore hitpoints after they are lost). Shamen would specialise in preventive healing (they debuff enemies and buff allies to prevent damage). Druids would specialise in pro-active healing (rather than affecting the incoming damage they would boost the outgoing damage, shortening the fight so that little healing is needed. How this would match with their DPS role is difficult to see unless these have a similar force multiplier effect that bards also offer, but that would have negative consequences for the druid's ability to solo).

On the tanking side we have three classes but I know very little of the tanking game so I could not argue if that is meaningful or not (they do seem to follow the same general principle of warriors absorbing damage outright, paladins able to mitigate it or heal it away, and shadowknights able to increase DPS to shorten fights and thus damage taken).

All DPS classes need a secondary role to offset their primary one. Some already have a clear such role (e.g. enchanters offer crowd control) but in many cases this would have to be invented, though the positional combat system I postulated and the necessity to create temporary weaknesses in enemies to be able to defeat them (instead of the slugfest we currently have) this would create ample opportunities to give classes secondary specialisations to augment their primary DPS role.


Well, above is pretty much how I would redesign the game if I was given the opportunity. I would try to make classes more interchangeable while at the same time creating combinations that work well together, and make the game as a whole more challenging by relying less on big numbers. Harder monsters would not so much have more hitpoints but more resistences and be less susceptible to temporary weaknesses, creating the need for bigger and better coordinated groups and opening the game up for more tactical gameplay where both group action and reaction to how the monster responds become much more important)


Eri