Gamasutra: The Art & Business of Making Gamesspacer
View All     RSS
October 20, 2014
arrowPress Releases
October 20, 2014
PR Newswire
View All
View All     Submit Event

If you enjoy reading this site, you might also want to check out these UBM Tech sites:

DPS and the Decline of Complexity in RPGs
by Eric Schwarz on 12/10/12 10:45:00 am   Expert Blogs   Featured Blogs

The following blog post, unless otherwise noted, was written by a member of Gamasutra’s community.
The thoughts and opinions expressed are those of the writer and not Gamasutra or its parent company.


Modern role-playing games really haven't gripped me in the same way that the older ones have.  The reasons for that are manifold, but one of the biggest is in the way that developers have begun to adapt new mechanics and ways of presenting information which is at odds with the complexity that I expect from a good RPG.  The influence of titles like World of Warcraft is felt far and wide - sometimes for the better, as in the case things like interface design, but also, in many cases, for worse as well.

In this article I'd like to discuss what I think is the single biggest issue with modern RPGs compared to their predecessors from the 80s and 90s - the prevalence of Damage Per Second as a gameplay concept.

Why DPS?

In order to get into why DPS, in my opinion, isn't a good thing for RPGs, we need to take a closer look and examine what sorts of goals DPS tries to accomplish.  Some of these are obvious, others, less so, as are the reasons and consequences:

  1. Unifies multiple damage types under one number for easy comparison.  Many modern RPGs, especially those built using the MMO model, tend to be very loot-driven compared to older titles.  The motivation for the player isn't just uncovering the story, it's "look at all this awesome stuff I got that looks so cool and epic."  DPS streamlines the time and effort spent comparing loot significantly - no more damage ranges to work out, no more calculating average damage, no more figuring out if you want X or Y variation on your damage output (i.e. higher max damage or smaller damage range).
  2. Is more consistent in actual gameplay, which makes things more predictable for the player.  Similar to the above point, most modern RPGs are not built around the idea of getting critical strikes, knocking down enemy defenses, and so on.  Rather, they tend to be focused more on threat and aggro management between enemies (which forces the player to attack certain enemies in a certain order to distribute incoming damage more effectively).  Combat is less about finding the most efficient ways to defeat your enemies, and more about being able to survive for X amount of time.  This allows for far more controllable encounter design - now you know why for some reason, when you're playing an MMO, enemy groups always seem to take exactly 30ish seconds to defeat.  It has nothing to do with combat quality and everything to do with making the player spend a pre-determined amount of time on the activity for pacing reasons (and in the case of MMOs, to artificially inflate the gameplay time spent with the hopes of getting additional subscription or cash shop dollars out of the player).
  3. Allows for character builds to be balanced and standardized more easily.  In recent years, the idea of game balance in RPGs has become something of a holy grail, generally as a result of the rise of MMOs.  In the MMO environment, which is predominantly social, interactive, competitive and collaboratory, players both want to feel like they are useful to the team and in the world by themselves, but also don't want to feel like any one players is significantly more effective than them.  The standardization of damage output through DPS makes it easier for developers to create that sense of every player being equally effective.  In a single-player context it means that you will likely know how much damage output the player has at almost any stage in the game, based on their equipment and level.
  4. Removes chance-to-hit mechanic.  In older RPGs, it was possible for characters to miss attacks completely, usually contingent on a die roll and a to-hit chance.  In modern RPGs, chance-to-hit has almost been completely removed, for two reasons.  The first comes down to animation - all animations are created with the expectation that the player will actually connect with an attack, and creating an entirely separate set of animations that correctly depict misses in a way that looks good is fairly difficult.  The second is more a matter of simple player ego-stroking.  Nobody likes to miss attacks, as it adds a degree of unpredictability to gameplay and makes players feel less effective than they are.  DPS, by giving a constant damage output the player can rely upon, "solves" these "problems."  It also allows developers to make missed attacks a function of certain skills and spells, like a "Miasma" spell that makes the player miss attacks 50% of the time, but otherwise never any other time. 
  5. Gives a reference point for additional systems to base themselves off of (i.e. DPS determining spell damage, to give magic users a reason to use certain gear). One of the biggest challenges of developing an RPG system is that often fighters and wizards had completely different ways of dealing with problems.  Fighters would have a fairly stable, consistent damage output that could be more or less relied upon, and it was usually simple "physical" damage if a game did have distinct damage types.  Wizards, by contrast, usually have extremely low damage output with standard attacks, to the point of being almost totally ineffectual, but made up for it both in huge "burst" damage output (i.e. an area-of-effect fireball), as well as in the secondary effects applied (lingering damage-over-time, fire damage type bypassing armor, etc.).  This worked well in party-based games where it wasn't imperative for every character to be the same level of effectiveness all the time.  In MMOs and single-character RPGs, however, which have become the norm these days, developers must rely on the player having a very consistent base level of effectiveness which is not compatible with this system.  By effectively making fighters and wizards cosmetic variations of each other and basing each around the DPS mechanic, suddenly, the classes are much easier to balance.

 All in all, DPS represents a far more drastic paradigm shift in RPG design than some may realize.  It's not simply about communicating a number in a more player-friendly way, it's about completely overhauling the base mechanics of games in order to accomplish fundamentally very different goals from modern RPGs.

The DPS Problem

As an exercise in game design, the development of DPS is actually quite fantastic.  Make no mistake, when I think about game design, I think about problem-solving, and DPS, if your project goals are in line with what it accomplishes, is possibly one of the best systems design decisions you can possibly make for an RPG.  But, I'm also of the opinion that DPS has fundamentally damaged much of the uniqueness of RPGs, the depth of their mechanics, and has in general led to far more boring games.

The first major issue I have with DPS is the most fundamental: standardizing damage types.  While some games using DPS do tend to also maintain multiple damage types, the majority do not, instead treating DPS as a be-all, end-all number.  Even those games that do have those different damage types tend to have only cosmetic effects.  By contrast, the Infinity Engine games made famous by Black Isle were built around the Dungeons & Dragons rules, and as such multiple damage types were in play in any given battle.  For example, instead of "physical" damage, there was slashing damage, crushing damage and piercing damage, all which affected different types of armor and different creatures differently (i.e. quipping staves or maces was critical when fighting skeletons, as slashing and piercing damage were significantly less effective), which promoted diversity in the party and made min-maxing less effective, and meant that even a "simple" fighter had wide utility value and some tactics to consider.

This leads into the second major problem, which is almost a direct consequence of the first: standardizing character classes.  In games which feature DPS, that DPS tends to completely remove any uniqueness in gameplay from different types of characters.  It doesn't matter whatever permutations of a class you are playing - in virtually every game of this sort I have found that almost all of the differences between characters were not in gameplay, but in aesthetics.  The fantasy of playing as a barbarian wielding a two-handed sword, versus the one of a svelte assassin backstabbing foes, is definitely a compelling one, but ultimately the only real difference in gameplay tends to come down to ranged vs. melee, and tank vs. damager - distinctions which already existed in other systems and, by virtue of the inclusion of DPS, have less depth to them than they would otherwise.  For all their "diverse" character classes, most MMOs I've played have all classes feel pretty much identical, with the only major exception being Diablo III, which still pales next to earlier games in the series.

The end result of this are games that feel significantly less replayable and also tend to be far more boring, both in single-player and multiplayer contexts.  When I'm playing a single-player game and the only difference between my character is what the loot I use looks like, that is not conducive to a varied gameplay experience - it means that if I ever go back to playing the game a second time, I'm not going to play it significantly differently, and even during the course of a single play-through, if DPS is all that matters, chances are I'm also going to get very bored very quickly.  In an MMO, DPS has the effect of making distinctions between different character classes contingent largely on support abilities, i.e. aggro management vs. healing vs. buffing, and this tends to make different characters interchangeable with one another - one fighter is the same as any other, one cleric is the same as any other, and so on.  This becomes overly formulaic and reduces the uniqueness of your given character significantly, even though creating a unique character and growing him/her in power according to your wishes is supposed to be one of the biggest draws of an RPG.

DPS also tends to get in the way of good equipment systems.  Most more traditional RPGs did have their fair share of very standard weapons - a long sword +1 is obviously going to be a linear upgrade over a standard long sword.  But, interesting distinctions between exotic weapon types were made possible by the differences between damage ranges and attack speeds.  For example, a sword may do 4-8 slashing damage, while a scimitar might do 2-10 slashing damage - it's a small difference between these two, but it can influence your equipment selection and the way you build your character when combined with other mechanics.  For example, maybe if you have a lower chance-to-hit, your decision to use a weapon will be different than if you have a higher chance-to-hit.  The same applies if we are talking about 1 attack per round vs. 2 attacks per round.  These small distinctions in equipment selection made some RPGs more "fiddly" to play for certain gamers and  added a little more to wrap one's head around, but it's this in-depth tinkering and decision-making that made them interesting to play in the first place.

In other words, DPS is very much responsible for the push towards making RPGs glorified action games instead of a distinctive genre with complex systems-driven interactions.  By making character advancement more linear, character classes less distinct, and gameplay more predictable, it robs RPGs of the more simulation-oriented gameplay they became successful for. 

Closing Thoughts

The reason I bring this up isn't because I hate DPS and I think that DPS is something that shouldn't exist - on the contrary, it has worked very well for certain games.  My problems with it are mostly caused by the way in which it has helped transform RPGs from a unique style of game with their own nuanced rule systems, towards action games with glorified progression systems.  DPS has definitely been a boon as far as marketability and mainstream appeal goes - just like leveling up, it's an easy carrot for players to follow that is basically foolproof in significance - but generally speaking, all the classic RPG fans I know are very much aware of the differences in gameplay that DPS brings.  The ones who have embraced it are primarily not of the same community that originally supported RPGS in the first place.

That's not to say that DPS, is solely responsible for this degradation in RPG gameplay, and many of those effects of DPS are secondary in nature and can be altered.  For example, DPS doesn't force developers to use only one damage type, but it does certainly encourage it, and once you've started to create DPS with all the nuance and depth of the older systems by adding additional modifiers on top of it, well, we've really just come full circle and made DPS obsolete anyway.  DPS is a trend that's critically approaching the same level of saturation as XP progression in shooters, and I'm afraid that it's slowly taking us down the road towards homogenization of genres.

Related Jobs

Digital Extremes
Digital Extremes — LONDON, Ontario, Canada

New York University
New York University — New York, New York, United States

Faculty, Department of Game Design
The College of New Jersey
The College of New Jersey — Ewing, New Jersey, United States

Assistant Professor - Interactive Multi Media - Tenure Track
Next Games
Next Games — Helsinki, Finland

Senior Level Designer


Ulf Hartelius
profile image
I personally don't see DPS as something new; developers have been doing that since long before the term was invented. The main difference is showing it explicitly to the players instead of requiring them to calculate it themselves.
Go back to any classic hack-n-slash role-playing game, like Dungeons & Dragons, and players are more or less forced to calculate 3D6 vs 2D10; something that immediately gets some people alienated (since they have a mysterious adversity towards arithmetics) and others just using whichever weapon has the coolest name.

That said, it may very well help influence modern, streamlined game design, along with several other factors, but DPS in itself doesn't inherently have anything against mechanics such as damage types (you'd just have one DPS value for each damage type). The reason mainstream developers have foregone, or at least simplified, something like damage types is because of the ambiguities involved; apart from obvious elemental enemies it's seldom easy to understand why this or that type works well or not. And because in the end you're just walking around with an arsenal that needs more care than the rest of the game combined (see Vagrant Story).
I'm not against having complex systems, I just think that, once in a while, I would've appreciated an Average Damage value or two back in the days.

Eric Schwarz
profile image
Perhaps I should make a distinction: I'm not talking about DPS as a display calculation. I'm talking about DPS as it is used for damage numbers in and of itself. Diablo III, for instance, actually does have damage ranges and attack speeds, but they are completely secondary to DPS because DPS is used to calculate everything (skills) and if players aren't spamming skills constantly, they are doing something wrong.

Frank DAngelo
profile image

correct me if I'm wrong, but I was pretty sure that DPS numbers in Diablo 3 were not related to the actual damage the skills do. I think this is a common misunderstanding, since most people believe that the damage an ability does in Diablo 3 is based off DPS solely. (for example, if an ability says it does 160% weapon damage, people think that said ability just takes the displayed dps number and it does 160% of that number). As someone that enjoys theorycrafting, I was under the impression that abilities actually scale off damage range instead. For example, if your damage range is 100-200, each ability cast will scale off whatever damage number from 100-200 the game rolls when the ability is cast. Like I said, I could be wrong, but maybe I just misunderstood what you are saying. All and all, it makes you wonder what the point of the DPS number even is in Diablo 3. It also leads to lots of confusion, since depending on your build, weapons that have lower dps but higher damage ranges may be better than a weapon with a much higher DPS. An example of this is comparing the slow attacking spears and maces with the fast attack speed daggers and swords. A sword may have 900 dps, and the mace 850 dps, but if you use a critical damage build, you will do much higher crits with the 850 dps mace because it has higher base damage. Likewise, if you have a life on hit build, the sword will be more beneficial because of the increased attack speeds.

Anyway, sorry to go on a little ramble there, but I've always wondered what the point is of that DPS number in Diablo 3 since it doesn't directly affect damage done with any skills. It's kind of just some balancing point that shouldn't even exist in an ARPG

Darren Tomlyn
profile image
Dammit - I was planning (well, I still am) on writing a post on this subject later on, though with far more additional context based on the posts I'll need to write before it.

There are a number of different reasons why this is happening to computer games, almost in general. The main reason, however, is that such systems and mechanics have to be built form scratch for each game that is made, unless there is a previous game that uses them, or, in very limited cases, you've got a proven system to build on (e.g. D20). Since the cost of games is only increasing, adding such complexity is the opposite of what people are looking to do.

Until we get some kind of standard 'gameplay development framework' (since that is what all this is about, and for), linked with any particular engine, everyone having to re-invent the wheel every time they want to make such a type of game, is going to have such an effect.

The foundation of such a framework, is, of course, the ability to represent or link everything necessary as and by a number and then link them with whatever algorithms we require and have them affected by them in a standard way. Since everything in a computer program is a number anyway, you wouldn't have thought that this was that hard to do - we just need a consistent manner of manipulating it, without having to write the code from scratch.

Luis Guimaraes
profile image

What do you have in mind regarding new languages to organize and work gameplay development ('gameplay development framework' as you say)?

Jonathon Walsh
profile image
I think you're bundling a bunch of different things together and blaming it on DPS.

Take DPS in Diablo 3 for example: You may complain it makes things like damage ranges and the like useless, but instead it's the opposite. Damage ranges were useless and displaying DPS exposes the false choice.

For things like attack damage types, some games use them to good effect but more often than not the issue falls one way or the other: a mundane optimization that's boring clicks for some additional dps, or a difficulty spike/punishment you can do nothing about. If you can just swap weapons to optimize the fight, the system is just some inventory baggage. If you can't (or you need the gear or guys specialized with different weapons) then it's just a difficulty spike you're stuck with. Since most RPGs make you invest in decisions upfront when you don't know a lot they don't want to punish you for those choices. Where you see damage types work well are roguelikes that are build on iterative mastery. In that case they can punish you for not having different damage types under the assumption you'll play again and diversify.

All the talk about DnD systems evaluating 2-10 vs 4-8 or chance to hit systems fall under the same thing. They're really all just an optimization problem that exposing DPS trivializes (for the good, the mechanics aren't interesting). You may have some complex goals like managing tolerable variance based on eHP, but it's still just boils down to solvable math.

The more underlying more valid complaint you have is that nothing is filling in the space created by removing all these false choices. Instead of adding in cool meaningful choices, many games just expose DPS, remove the false choices, and leave the game mechanically empty. Still the intention is the same; RPGs are trying to become less about optimization & math (which better UIs and player interactions have exposed as the case) and more about making interesting decisions.

Eric Schwarz
profile image
"Take DPS in Diablo 3 for example: You may complain it makes things like damage ranges and the like useless, but instead it's the opposite. Damage ranges were useless and displaying DPS exposes the false choice."

Arguably then the problem is that the number of attacks you get in a game like Diablo is "too high" to support damage ranges well, because at that point average damage does tend to win out. But I also think that the difference between doing 50 damage and 20-80 damage is a pretty significant one and isn't at all a "false choice" unless the rest of the game systems make it so (i.e. skills deliver fixed damage amounts instead of weapon damage range having much influence).

"Since most RPGs make you invest in decisions upfront when you don't know a lot they don't want to punish you for those choices."

Having forethought in character development defines much of an RPG's gameplay. One might say it is the single most important (and fun) part of the game, systemically speaking. Without consequence there can be no significant choice, etc.

Ricky Bankemper
profile image
"All the talk about DnD systems evaluating 2-10 vs 4-8 or chance to hit systems fall under the same thing. They're really all just an optimization problem that exposing DPS trivializes (for the good, the mechanics aren't interesting)"

I would call this quote right here, making a build for a character. I kind of feel this is what RPGs are suppose to be about, at least to me. The "trivial optimization problem" is part of figuring out what works in the game, part of the fun. I suppose that is subjective fun though.

Joshua McDonald
profile image
I agree with Jonathon.

A little extra detail: Eric kind of addressed this point, but if the fights are longer than a few seconds, mathematically, it will boil down to DPS (30 hits at a flat 50 or at 20-80 damage will come out to pretty similar numbers). Likewise, damage types, as noted by Adam Miller below, are more about tedious inventory management than interesting choices. One of the reasons Baldur's Gate doesn't feel like a DPS game is because most enemies have very few hp, so a spike in damage is extremely noticeable.

In a game like WoW, the choice used to be more of a burst vs. sustained and AoE vs. Single target, but the encounters (particularly boss fights) were designed in such a way that single target sustained was pretty much always best, so rather than changing encounters to make other builds work or even be better, they changed the classes so that their single-target sustained DPS all came out pretty close. Admittedly they might have changed it again, since it's been a long time since I've played top-level content.

The only way to provide meaningful choices is through deeper encounters. That doesn't mean more complex script to follow in order to win, but varying situations with multiple solutions that are rarely repeated exactly.

While it's not exactly an RPG, a game like League of Legends can be really instructive here. For example, the deeper your tanky character rushes in, the more likely the enemies will be to focus him (which is what you want), but the longer it will take for the rest of the team to catch up and start helping. The decision is going to based on the positioning of both friends and enemies, the status of health and mana bars, proximity to towers, and many other circumstances that will never be exactly identical in any two situations. Even max-damage rotations will often be deviated from because your circumstances leave you with a superior option.

Overall, I think it's easier to design interesting short fights (Baldur's Gate/League of Legends) than longer fights (WoW/Diablo III bosses) since in the latter, things like burst or varied damage just become part of an average DPS calculation, regardless of whether the game displays that stat to you.

Adam Bishop
profile image
Good stuff. I don't like the growing tendency for RPGs to try to play more like twitch action games. To me, some of the most important decisions in an RPG ought to be made outside of battle. Preparing for battle by figuring out what equipment and abilities are most useful in any given situation is a really important part of those games for me.

Take, for example the "All" materia in Final Fantasy 7. Because of the limited number of materia slots available you can't just add the "All" materia to every ability, so you have to choose: do I want to add it to my healing spell so that I can quickly heal the whole team? Do I want to add it to my haste spell so I can give the whole team increased opportunities to attack? Do I want to add it to my lightning spell to save MP because I'm in an area full of robots and I can more quickly dispatch them by hitting them all at once with a lightning blast? That kind of decision changes what you're able to do in battle and how you approach it, but it also requires the player to try to understand the game's systems to figure out what works well together, what doesn't, and when an ability is most useful. I find that far more appealing than the route they've taken with Diablo III, which is to try to make everything useful in almost every scenario through the simplicity of DPS.

Jonathon Walsh
profile image
D3's simplicity comes more from the lack of difficulty & early on gear restrictions than the DPS simplification. That and the monsters being largely the same. When you reach the 'end game' and are pumping up the difficulty of inferno AND have access to a wide array of gear, then the stat & DPS system becomes more complex. It's at that point where how attacks interact with attack speed vs critical hit chance, the value of life on hit vs life leech as a %, and how you can combine abilities together for synergy, have a large influence on builds & resulting effectiveness. When you're going from 1-60 the game is just click to make things die (which is a major and totally fair complaint of the game, but not one that feeds into this article).

profile image
The problem is not really DPS but the mindset of "game as progression" instead of "game as management". The problem is that old RPG was often poorly balanced as you had to do upfront decision without knowing what's coming else, either you have limited inventory and it is a lottery with blind guess, or you have significant inventory and then it became simple pattern matching. Poor "game as management" became "game as progression" when optimization kicks in as knowledge negates choices.

However one game pull it greatly ... pokemon, region allow you to anticipate type weakness, there is limited inventory that allow for plenty way to deal with the same situation and attacks go beyond simple DPS and involved planning and careful use. The main critical path is basically a game of progression with optional pocket where management is mandatory.

However the problem is that weakness is just not enough, investment mechanics is the spice of RPG. For example let say you have a quick attack that get better each combo you made? How many turn before this attacks ROI get bigger than a stronger attack? It becames obvious that long terms and short terms decision can change the tide of the battle, such an attacks would be great on long battle but useless on short one ... Basically adding feedback loop on top of dps add depth and anticipation.

There is also positioning, not just spatial positioning but any mechanics that change the balance of advantage and weakness relative to opponent. long range VS short range is just as good swapping type weakness or aggressive vs defensive stance relative to terrain bonus.

The last aspect is trick, consider the "stake space" (how close you are to losing/winning forming a 2D space, in rpg it's your health vs enemy health), direct manipulation of that space is a trick, for example swapping your health to the enemy health is a powerful swing in the stake, ability to prevent or promote it are also trick. Stun, option lock, etc... are trick, they give advantage without actually advancing "inherently" toward winning or losing, for example the swap health mechanics would be a bad move if you are leading in health. It promote outsmarting your enemy by setting traps. Yugi oh card game is mostly a game of tricks.

Eric Schwarz
profile image
"The problem is that old RPG was often poorly balanced as you had to do upfront decision without knowing what's coming else, either you have limited inventory and it is a lottery with blind guess,"

So? You learn from your mistakes, you plan ahead properly. In fact the lack of any significant planning and preparation required in modern RPGs (as opposed to "rush blindly into epic battle!!11") is one of the major problems I have with the MMO-inspired games these days.

"There is also positioning, not just spatial positioning but any mechanics that change the balance of advantage and weakness relative to opponent. long range VS short range is just as good swapping type weakness or aggressive vs defensive stance relative to terrain bonus."

This is very true, but there are very few games where spatial positioning is systematized in any significant way. In fact, the RPGs that do tend to do this are tactics RPGs, not MMOs.

profile image
I do not see trials and error as good design or even good experience, I also assume that you have in mind Save scumming as a gameplay mechanics to balance that. In fact I do love being surprised and toss into a hopeless situation, except I have less patience as my time get more precious than in my young days. And there is a genre just for that, rogue like, and their main draw is that level get reset so it does not became a tedious rot learning.

However the problem is more the lack of preparation and planning, it can be adressed without having flaws from old or new rpg. I did mention pokemon as doing it well, it has depth thanks to the right limitation that force decision, if you take it seriously.

Another point, I read your post (all of them) since the beginning and enjoy them a lot, they have a very clear POV that really help to understand they aesthetics you promote. One thing that I note and it is tangentially link to this issue, is "game as manipulation". A lot of people would say that having fake complexity (tons of useless items, obscure stat relation) is bad design, I bet you see them as good design simply because it engage the player mind just like a level, the "mental sorting" of what's good and bad (items, skills, stat increase) is stimulating. Having the junk too easily discernible from the good bit is killing that fun. The "simulation" excuse is just about giving relevant value that ties it back to mimicry. That aesthetics is something lost into the "mathematics" of modern game design. Maybe it's time to differentiate between "game design" (mechanics) and "play design" (experience).

Eric Schwarz
profile image
So what you're saying is that the player should not be punished with death (failure state) if the player makes a serious mistake (fails)?

Most RPGs are not especially difficult to get a grip on. Almost every RPG I have ever played (at least good ones) have made it clear that there are alternatives to rushing into encounters. Maybe you have a class that excels at scouting that you can play as. Maybe you can sneak around and bypass combat entirely. Maybe you can solve a quest to make the enemies less of a thread. When it does come time to do battle, usually basic rules like "water beats fire" apply.

Pretty much every RPG explains its mechanics directly to the player - it's just that it may require *gasp* a little reading to figure out, or perhaps you can try something out for yourself. Usually failure in RPGs results in death only if combat is inescapable, which is not true in the majority of RPGs I have played. In turn-based games you also usually have the ability to run away, or to adapt (because combat is slow and gives you time to think).

Sorry, but if a player cannot figure out that maybe rushing headlong at a minigun-toting Super Mutant while using just a 10mm pistol and leather armor isn't such a good idea, well, actually I'd be surprised if that player exists, because I don't think you could get that far in Fallout without learning at least a little bit.

All you're doing is perpetuating the myth that RPGs are extremely complex, impenetrable, unintuitive games that are simply not accessible to outsiders. They're not in the least; the complexity they sport is actually not much greater than your average shooter - the difference primarily comes down to how information is presented (stats instead of visual and spatial relations). That aesthetic barrier represents a significant hurdle if you are targeting certain markets, but it does not mean that, say, Dragon Age is an especially deeper, more thoughtful game than Call of Duty.

Regarding "fake complexity" - there is no such thing unless we are talking about mechanics which truly have no influence on the game at all, which is pretty rare. However, I am definitely one of those people who thinks "more complexity = better" in almost any situation where it is called for, and that the situation is almost always called for. That doesn't mean all mechanics need to be introduced from the beginning or that they must be presented in awkward, archaic ways, it just means that a game needs to be stimulating from a mechanical standpoint, and simple mechanics often get boring fast.

I remember Harvey Smith once said that one of Deus Ex: Invisible War's biggest failings was the fact that it over-simplified gameplay by removing redundancy. The rebreather item was redundant with the aqualung augmentation, and those were also redundant with the swimming skill... however, despite the redundancy, the player fantasy of being able to be "aquaman", as well as the option to open up a new character build from a mechanics perspective - *even if it was completely useless* - brought an extra dimension to character building. Even if only 1% of players actually did this, or fewer, it also brought meaning to the wider systems simply because it existed in the first place, and contrasted other elements in the system.

Naturally, now we have a whole generation of games that try to make up for this with more cutscenes, more set pieces, more flashy lights, etc. - but if the core gameplay is boring and does not evolve significantly during the game's length then it's no wonder it becomes very hard to sustain interest.

profile image
I think you completely miss my point, by trials and errors I'm targeting the randomness of encounter that limit "decision" from the part of the player. I'm not advocating for equip fire to go in ice cave, that's not anticipation but optimization.

What I'm in favor is that the next situation is (roughly) known BUT present many dilemma from a gameplay perspective. Ie should I prepare an agile build that would stealth out the area and keep my resource for the next trials or should I prepare for huge spending on recovering resource to explore nook and crany of the area? It has nothing to do with optimization but actual choice.

Plus anticipation favor immersion as you know what the player expect and can break expectation: "what does a fire monster in a ice cavern?" should hint at plot twist and also offer a spike of difficulty to optimized build (fire vs fire is weak), notice that the player who already know the twist still has to go through the dungeon with an unoptimal ice build (mixed) to lower the difficulty of the fire boss by increasing the difficulty of small encounter before it. It's not blind runs but effectively choice and playstyles that is favored.

Not knowing what we will meet favor average "jack of all trade" build until you know the game by rot learning and optimize, that's not choice and pretty much against what you say is good, faillure is good when I make a mistake, not when I don't know ahead what i'm getting into. It has nothing to do about complexity, it's about good gameplay and variety of choice.

BTW I don't understand the greatness of useless skill if it actually penalize my overall experience. I'm for it in case where it is harmless mimicry, but my mimicry is hurt if there is no way for me to actually experience my choice. At the end i'm against wall of random choice like in morrorwind because it only make sense if you play the game at least once, keep the choice but also offer prebuild class for newcomer that is tailored to the experience or let me build it over time as i get comfortable with the game and the interplay of mechanics.

Robert Boyd
profile image
This seems much more like a rant against MMORPG-style design in RPGs than it is about DPS.

Just about every single RPG out there has several damage types. At the very least, you almost always see a Physical/Magical split and then various elemental damage types. Splitting physical out into multiple types is generally complexity for the sake of complexity and is just one more thing for the player to have to keep track of. If you still want that level of D&D-esque flavor, you can do that quite easily by just adding modifiers to different weapons. For example, Maces could deal more damage to undead, daggers could be more likely to deal critical damage, spears could pierce defenses, and so on.

Eric Schwarz
profile image
You're partially right about that.

Still, show me an MMO where DPS, when used as the main indicator of combat damage:

a) DPS is divided into damage types and those damage types significantly affect combat (straight damage and skills)

b) DPS doesn't tend to negate the differences between weapons

c) DPS is used but does not injure differences between character classes

I certainly cannot think of one. That's not to say that DPS is in itself responsible for all evils in RPGs, just that when it is used in the MMO sense it also tends to promote certain systems design decisions which, in my opinion, are not conducive to complex interactions and therefore gameplay suffers as a result. I went into a lot of these in the first section.

Now, obviously we can get into "why does an RPG have to be complex?" but that's kind of another question for another day.

Robert Boyd
profile image
But even in a game like Diablo 3 where DPS is king, you're going to notice a big drop in your killing speed if your attacks are all an element that the enemy is resistant against.

Eric Schwarz
profile image
I never noticed this in Diablo III - when I played through normal, nightmare and hell, elemental damage and resistance was a rarity on everything except my own gear. Maybe it only comes into play on inferno difficulty, which I did not play significantly. Either way, cutting out the actual interesting systemic interactions until the endgame - which most players will not stick around for - is not a good idea.

Justin Speer
profile image
Very interesting article, I share a lot of your feelings here. I think you've correctly identified a design crutch that's contributing to homogenization in game design. I think the way the argument is framed may need a bit of refining, considering the back and forth here in the comments, but by all means continue with this line of thought. Again, good stuff here.

Gian Dominguez
profile image
I dont think dps is a problem. I just cant help but think of WoW with regards to some of your issue and how WoW has dealt with it over the years.

"In games which feature DPS, that DPS tends to completely remove any uniqueness in gameplay from different types of characters."

WoW has managed, in my opinion, to make not just classes but specs to act different. An elemental shaman has a different play style than a mage. Different priorities in stats/different rotations and different tools sets to use for different encounters. While I think the devs. are in a constanr battle to fight homogenization I think they have done quite well, considering WoW has all

"Similar to the above point, most modern RPGs are not built around the idea of getting critical strikes, knocking down enemy defenses, and so on. Rather, they tend to be focused more on threat and aggro management between enemies "

Threat is no longer a significant mechanic. Heck as long as a tank hit the enemy first they all but guranteed to stick to him. Most dps can unload almost immediately.

Bisse Mayrakoira
profile image
"An elemental shaman has a different play style than a mage. Different priorities in stats/different rotations and different tools sets to use for different encounters."

Stat priorities just change which spreadsheet you should use to prioritize your gear, it has zero effect on playstyle. I have a feeling the rest of your comment applies mainly to high end bosses, while 99% of PvE fights in the game are straightforward "keep as many enemies crowd controlled as possible, burst down the rest one by one" and as long as you have someone to tank for you, all DPS works mostly the same. Different canned rotations are not a meaningful playstyle difference either; classes are played fundamentally differently only if their significant decision points during fights are different.

Gian Dominguez
profile image
"Stat priorities just change which spreadsheet you should use to prioritize your gear, it has zero effect on playstyle."

My point there is how you "build" your character is fundamentally different between classes. That there is no "one perfect build" that everyone follows.

"I have a feeling the rest of your comment applies mainly to high end bosses, while 99% of PvE fights in the game are straightforward "keep as many enemies crowd controlled as possible, burst down the rest one by one" and as long as you have someone to tank for you, all DPS works mostly the same. "

WoW, if nothing else has evolved with regards to there PvE encounters. There is no simple "tank and spank boss" or "keep enemies cced while you deal with the boss".

"Different canned rotations are not a meaningful playstyle difference either; classes are played fundamentally differently only if their significant decision points during fights are different."

Its not really just rotations. Its the whole tool set of abilities people use to deal with different encounters. In an encounter with heavy movement X class would be better, while in a fight where you dont normally move Y class might be better.

Bisse Mayrakoira
profile image
I think you missed my point. What you are talking about is boss fights, and more specifically, boss fights at high levels. 99% of the fights happening in the game are NOT boss fights, but fights against a bunch of dumb mobs. In the vast majority of those fights you can just run through a canned rotation of skills, and that is doubly true if you play a DPS class and are playing with another player or a pet that is capable of tanking. Ergo, most of the time in the game, all DPS classes work largely the same, with few interesting decisions to make. Anyone even slightly competent at the game can be taught to play a new DPS class passably in minutes - you just outline their main damage rotation/abilities, explain what they have in the way of crowd control ability (if the class has any), and set them loose.

Are there differences between the classes? Sure, and I think WoW has some very entertaining fighting mechanics and managing to master the subtleties is occasionally very rewarding. But those things pretty much only come into play in PvP. Not only are the average PvE enemies very simple and one-dimensional (they have no AI at all, and I think most of them have just one, or at most two special abilities which they use at random?), but also the PvE -specific game system rules have been tuned to further dumb down and homogenize fights, steering players towards the most straightforward approach and equalizing the classes. For instance, almost all of the subtle and complicated tactics as well as demands placed on the player which make the rogue class interesting to play in PvP are completely removed or shut down by the game system in PvE. Different rules are applied in PvE to determine the "in combat" status of the player than in PvP; mobs gain a full status reset if the player enters stealth during a fight; and since enemies are skill-less, AI-less dummies, there is no cooldown strategy, no kiting, etc. After seeing the depth potential of the mechanics under PvP play, slogging through enemies in PvE is just mindnumbing.

Robert Williamson
profile image
Interesting article. What you said about balancing stood out to me though. I think balance issues begs the question of whether classes in modern RPGs are unique enough. By that I mean if DPS is a standardizing measure of the output or usefulness of a particular character, doesn't that mean that all modern classes are overly dependent on damage as a measure of utility?

I could be totally wrong, but it seems to me that the rise of DPS indicates a lack of differentiation between classes. Why aren't we designing classes with skills that include a majority of debuffs (non-damage dealing) or classes with complex unlocking skills. For that matter why aren't we designing bosses and dungeons that require more specialized classes?

Adam Miller
profile image
D&D style damage types (slash/stab etc.) or elemental damage don't add interesting complexity to a game. Generally adding such things introduces the tedious process of changing equipment for different foes, or as you mention the need for party variety--but without really giving the player interesting reasons to create balance (how is it fun to have a fire guy or a bashing guy because that's what kills undead?)

For RPGs to truly evolve they need to include varied problem-solving possibilities ala the original Fallout games (you can talk your way through situations, you can rob your way through them, etc.), or more verisimilitude/less stats. A simple example would be finding a sword that LOOKS sharp/heavy/glowing, without being able to see associated numbers. Rely on graphics and physics, in other words, to less precisely (but more realistically) communicate information. RPGs today lack mystery because everything is reduced to numbers, and those numbers are shoved in your face.

Or imagine if fire could, you know, burn wood. In real D&D players can solve problems in dynamic ways, say burning down the leg supporting an archer's tower. But you can't freeze the wood or poison the wood and accomplish the same thing. Or as some games have done, make fire area of affect, lightning chain, ice slow, etc. At least that varies how the player must play the game. Conversely, making fire affect certain enemies more than others isn't an interesting gameplay dynamic. At all.

Eric Schwarz
profile image
"D&D style damage types (slash/stab etc.) or elemental damage don't add interesting complexity to a game. "

Oh? How does more important and nuanced party composition enforced by inventory limits and damage types make a game more tedious?

"For RPGs to truly evolve they need to include varied problem-solving possibilities ala the original Fallout games"

Absolutely true, although personally I don't think the use of graphics is necessary for RPGs to "evolve." Other genres already do the visual/spatial relations well enough, that was never the forte of RPGs to begin with.

"Or imagine if fire could, you know, burn wood. In real D&D players can solve problems in dynamic ways, say burning down the leg supporting an archer's tower."

That said, Divinity: Original Sin does exactly this, going back to that old Ultima-style idea of a simulated world. It's one of my most anticipated games for this reason.

[User Banned]
profile image
This user violated Gamasutra’s Comment Guidelines and has been banned.

Michael Stevens
profile image
Slash/pierce/blunt melee damage is only useful in games where classes are locked into one type. If the difference in damage between types is significant, it can be used to to get players out of their standard combat roles (ex: because the white Mage uses staves, he is now the primary damage dealer and the party needs to reorganize to operate without his healing magic.)
It's a less interesting choice when the solution is just "Knight unequips sword, Knight equips stave" and everyone continues playing the same way they always did. Vagrant Story nearly avoids this problem by having those bonuses and deficiencies be gradually curated by the player throughout the game, but plenty of people found that part annoying and I don't think it would translate well to group play.

I feel like the tension in this article is between designing for scenario vs designing for self directed play. Jrpgs tend to do scenario better than wrpgs, and are also more likely to walk the middle path. I'd say Mana Khemia 2 is a good game to look at for scenario battles. It's turn-based but has real sense of time with characters jockeying for turn order, unoccupied turns, skills that boomerang back into effect after a certain amount of time has passed, and characters regenerating in the back ranks. Characters have incredibly idiosyncratic builds and players need to constantly be swapping back and forth mid-battle.

Kyle Jansen
profile image
I don't think such a type system is only useful in games with strict class limitations. I would simply say that the weapon type system could have adverse interactions with the class system.

I mean, there's the obvious "what if my game doesn't have a class system?" argument. A weapon type system could work well there, particularly if you have limited inventory slots (ie. Char A is carrying a sword and spear, Char B is carrying a sword and mace - which ones do you equip?).

Then there's the one-man-party argument. Without multiple characters to handle different enemy types, you would need to be able to swap weapons, or else be completely ineffective.

But even in a game with large parties and a strong class, I would not say multiple damage types are completely useless. You just need to provide a proper incentive for equipping characters "properly". The easiest way would be performance penalties - your barbarian guy does full damage with two-handed swords, half-damage with other two-handed weapons, and quarter-damage with everything else. It keeps the classes significant, but it gives the character some utility in battles where he would otherwise be useless - imagine a "long-range" battle, where only ranged attacks do damage: the barbarian would normally be useless, but could be re-equipped to be marginally useful.

Garrick Williams
profile image
I feel like the article mostly attacks the symptom and not the disease. Shallow games lead to shallow implementations of DPS, not the other way around.

To be deep, a game needs to present the player interesting scenarios and choices. A game fails to do so when they do not deliver circumstances that call the player to make new decisions, or if the game only presents a very limited amount of viable options. When this happens, raw DPS becomes the only important factor. Thus, DPS is the result of a shallow game -- not the cause.

Damage types do not necessarily grant depth. They have to present scenarios that force a player to change their style of play or make meaningful decisions. There needs to be a significant trade off. I played many games with damage types that fail to do this. In Torchlight, weapon switching is easy and there's no difference between my Sword of Lightning and Sword of Poison. There's no interesting choice here. If I'm up against zombies, I'm going to switch to my Poison Sword with no trade off except for the possibility that it's a lower level sword than the other one.

D&D, on the other hand, is different. If I'm a sword fighter and encounter skeletons, there's several options to weigh. Do I sacrifice a turn to switch to a club, or do I stick with the sword for a better chance to hit at the cost of less damage? What about the other monsters in the room; is the skeleton a significant enough threat to justify the cost of switching to an inferior weapon? There's pros and cons to each choice. In fact, even just getting a club is gameplay. Carrying too many weapons hinders you, and preparation and planning are a major parts of the game.

Hence, a damage type system alone does not make depth.

This is why several games do well to implement tactical aspects to different types of weapons. Each weapon in Phantasy Star Online has different attack animations that play a huge role in how its used and different scenarios where it's useful. Then it's not just DPS and circumstancial benefits that influence decisions, but also player preference and skill.

[User Banned]
profile image
This user violated Gamasutra’s Comment Guidelines and has been banned.

Kyle Jansen
profile image
What a strange coincidence - I noticed precisely this sort of problem in my own game, and my solution to it almost mirrored your own ideas.

However, we completely disagreed on the root cause. You seem to posit that the mere existence of (or at least, the emphasis on) DPS is the cause of overly-simplistic gameplay and bland class/equipment options, while I think it the *cause* is the simplified gameplay, and the emphasis on DPS a symptom.

The problem, at least as I see it, is not that games are boiling combat effectiveness down to a single number. The problem is that games have such simple combat that you *can* boil that down to a single number.

My own game did not display DPS anywhere, nor did it really factor into balance testing much. But because I only had two systems (physical and magical), both fundamentally the same, the combat was boring. And equipment builds were similarly boring, as the gear with the best offense usually had the best defense, so the only really viable option was to balance both. The stats in my game, by the way, are purely determined by your equipment - the only stat that "levels up" is the one that determines what, and how much, you can equip.

My solution, which I'm currently executing as part of a major game overhaul, was to further split up the options. The "physical" damage system was split into stab, slash and blunt (almost precisely mirroring the slash/pierce/crush of DnD purely by chance), while the "magical" system was split into fire, ice, geo and pure. I also moved the specialized defense attributes from the weapon to the armor, so one could equip a fighter with good anti-magic armor without turning him into a paladin or battlemage. Different armor will also be better against different sub-types - which is preferable, a Helm of +10 Fire Resistance, or a Helm of +5 Ice and +5 Geo Resistance?

I'm also working on ways to make the attacks more distinguished. There's already the distinction between damage types (unlike your suggestion, I put the damage system on the attack, not the weapon - while a daikatana might be far better at slashing attacks than stabbing attacks, it will still be capable of them).

But that won't be enough, I think. Since combat is still one-on-one, the obvious solution of single-target and multi-target attacks won't work. I'm currently leaning towards what I'm calling a "two-cost" system. I dropped the MP/AP cost of attacks, replacing them with a "time" cost (as it was proving too hard to make the AP/MP cost actually matter). The two-cost system would refine that further into a "warm-up" and "cool-down" cost - first there's the delay before attacking, followed by a wait before picking the next attack. This gives the speed stats a further emphasis. This should also combine well with the block/counter/shield/reflect systems - a high-delay high-damage spell, like a summon, might not be worth it if the enemy can squeeze off a few quick attacks and still have time to use a block attack.

However, given this new complexity, I think it is best if I add some sort of DPS-like number to the UI in places. I will probably display it as multiple numbers (one for slashing, stabbing or blunt attacks, for example), simply bolding or otherwise emphasizing the most powerful. That way players can see "my current sword does 2500 slashing damage. this new one does 2800 stabbing damage. But since I have no good stabbing attacks, I'll just hold onto it until I find one, or just sell it. then again, i've noticed all the enemies around here are weak against stabbing. let me try it out a bit with my old stab attacks and see how it works".

I suppose that's the real test. You can show whatever DPS numbers you want, but if you have to factor in too much other information, it becomes less meaningful. Meanwhile, even if you don't show a DPS number, if your system is too simple, people will just do the math themselves, and all you accomplished was make your players do some multiplication.

Craig Stern
profile image
I agree that DPS is lamentable insofar as it reduces strategic complexity. However, this piece goes off the rails once you start decrying DPS for reducing randomness--randomness is not coextensive with strategic complexity. To wit:

Matt Wilson
profile image
Curious as to your take on Star Trek Online, which uses damage types reasonably strongly, but still is an MMO that revolves around DPS. Perhaps that's a bit too granular for your statement, which I largely agree with.

Tony Wang
profile image
Hi Eric.

Great post. It got me thinking a great deal about just where we should approach game design from a combat perspective. Since my train of thought ended up being far longer than I planned, I wrote the article on my personal blog 'A pretentious sandwich'.

I would be extremely interested in hearing feedback from you regarding my analysis of your post. I mostly draw examples from World of Warcraft, so you may also have arguments stemming from gameplay in other games that I would love to hear about.

Hope to read more of your blog!

Eric Schwarz
profile image
Copy-pasting my response to that article:

Thanks very much for your extensive and detailed response, I'm flattered.

The short of it is that I'm approaching DPS sort of along the lines of the Sapir-Whorf hypothesis - the way that information with respect to mechanics and gameplay systems changes the way that we think about them, and ultimately games and genres overall.

You're quite correct in suggesting there is not a 1:1 causal relationship between DPS and other trends in modern RPGs, but I never actually said there was a causal relationship in the original article. My point is that by thinking about RPGs in terms of DPS, DPS becomes the central factor in RPG design. Everything begins to relate to and revolve around DPS, because that is now the metric we are using to understand game balance and indeed gameplay overall.

The same trend can also be seen in the heavy emphasis on combat gameplay and "phat loot" that most RPGs, especially MMOs, tend to sport, after combat began to become a huge marketing point for RPGs (faster, more responsive, more action-like) and Diablo II popularized random loot with prefixes and suffixes. It's common to see the focus on the combat and loot grind in games that don't even necessarily benefit from them or support them, simply because somewhere along the line developers started thinking about RPGs primarily in those terms. Even "traditional" RPGs like Dragon Age are susceptible to such things.

"RPGs where there is high variation in the mechanics of each fight (Baldur’s Gate/Neverwinter Nights/Dragon Age/Elder Scrolls, tabletop gaming, well-made MMORPGs), will need to be able to balance actual DPS from each class to ensure players can play the maximum number of viable character/skill combinations. On the other hand, games where travel time becomes insignificant, health recovery is not crucial, and theoretical DPS caps are easily reached, players will experience each class combination as devoid of variation and feeling the same."

Not quite. It's not just DPS that matters, but utility value. Modern RPGs tend to make all classes have very similar if not the outright same sorts of skills with small variations. Everyone has a stun, a summon, a buff, a debuff, an area-of-effect, etc. This is because playing a character without one in a modern RPG, especially an MMO, would be boring. This is because MMO combat by and large IS boring, and only gets its "interest" from the fact that you are mashing different hotkey cycles constantly. Is it any wonder why player expertly make macros to do it all for them?

When every character has more or less the same ability set with minor variations, you are going to end up with stagnant gameplay. No MMORPG I have played in the last several years has remained much fun past level 20 because once you have gained all your core abilities, there is nothing new left to do or see in combat, and when you restart as another class you realize you are doing the exact same thing as every other class. Because all classes have to be measured by the same metric in order to effectively complete the game (DPS, because all the game really has to measure progress is combat), utility value of classes goes out the window.

That's not to say there aren't exceptions, and indeed some games that feature DPS actually do have more interesting combat gameplay (sometimes you have to think about which hotkeys you're mashing and in what order!!1), just that this is the average trend I have found in the majority of MMORPGs and even single-player RPGs I have played in the last several years.

I think designers can and should use DPS as a tool, but only as a tool. By exposing those numbers to the players they contribute to the downward spiral, and will tempt both players and themselves to only think in those terms. Ideally an MMO that is meant to be played as a multiplayer title (not a single player game with a chat lobby, which is actually very common these days) should play very similarly to a traditional party-based RPG, where each character class has a very distinct and clear set of strength and weaknesses, with all classes required to reinforce each other to be successful. This interplay of systems also needs to occur in such a way that the mere presence of X or Y class is enough to affect the dynamic of combat (hotkey spamming!) but smart skill use as well.