Editorial Gamasutra: DPS and the Decline of Complexity in RPGs

[Raapys]

It's simple really. There's nothing wrong with a DPS number by itself, it's what it's used for that makes it bad. Not by the player, but by the developers. Having a DPS number available at all times easily lets them balance every single skill, weapon and spell in the game. Add in the fact that those things are usually 'leveled' these days, and you have an utterly boring, predictable and linear increase in power over time, which means you're effectively standing still.

Half the fun in games of old was to find 'stuff' that in some ways were significantly better than the alternatives in at least some situations, discover spells that were arguably completely overpowered and generally having your power increase in huge jumps at varying intervals instead of a constant and linear increase.

 

[felipepepe]

By now I'd think the only thing that matters to people is GEAR SCORE.

Exactly... is something that streamlines even more than DPS; no need to even consider the stats in your equipoment, if the gear score is higher, use it.

 

[Slow James]

I think the problem the article made was saying "DPS" is the problem, when really that is not the case at all. Its the idea that every class is a combat class, or (tying into his previous article about cooldowns), every skill is reusable/spammable.

Let's poke our head's back in time to D&D tabletop, say second edition. A party consisting of a wizard, a priest, a fighter and a thief. Your thief and your wizard are squishy, you'd have to stash them away from the fight so they can either deal out ranged damage or even hide from combat altogether. A thief would not be able to move all around the battlefield, performing back-stab, but actually had to stay in the shadows so they would not lose their "hidden" status. They would then either perform ranged sneak attacks or be able to perform one back-stab attack, which would expose them from cover. Your thief could do SIGNIFICANT damage with this, but only if the right conditions were met. In addition, you had to weigh out the benefits of losing said cover vs. inflicting the damage on your target.

Similarly, with Vancian casting, your mage could unleash a spell powerful enough to mop the floor with your enemies, but it was a one-shot deal. Do you use that "ace in the hole" spell, or do you use a lower level spell? Do you use a more enemy specific spell (for instance, using grasping vines on a Slime would be pretty pointless) or go for a more generic type of damage, since this is the first time you have encountered this exact enemy type? Do you memorize ten versions of magic missile, to make your wizard combat ready, or do you instead stock up on a number of more utilitarian, non-combat skills that may help more in the actual dungeon exploration?



Turning combat into a "this is the best skill for this character to use, so let's just spam it ad nauseum" and completely divorcing the character build from anything else BESIDES combat is a big failure of RPGs in recent years.

Please note, I am not an advocate for Vancian casting or any particular D&D model of stealth/healing/combat/casting, I was merely using it as a model to show that the concept "Here are four of my most powerful combat moves, so these are what my character should use all the time, in every fight" is extremely limiting. Note this also applies to buffs, debuffs, etc. If the perfect strategy for my character is to have X buff, Y debuff and Z attack skill going at all times during every fight, then that is a problem. Or, if it is the best strategy, but there is absolutely zero penalty to the non-combat portion of the game, or for future combat scenarios, to run that "ultimate" combat strategy at all times, then that is extremely limited, linear and, in my opinion boring. A la WoW.

 

[Norfleet]

The argument that the use of DPS has as one of its "goals" the abolition of to hit chance I found specially retarded.

This is utter bollocks, because DPS is simply a statistic, one that hit chance is a factor in. If your raw, 100% hit chance DPS is, say, 100, then if your hit chance is 50%, your DPS is 50. DPS still exists. DPS has always existed, and always will. Back before it was made explicitly visible, people were still calculating DPS as far back as the text age.

You can't blame a measurement for the system.

 

[sea]

Perhaps I should have stressed more clearly that DPS does not directly result in the things I said, but rather it creates a paradigm which strongly encourages/inspires other design choices. DPS does not automatically equal simpler gameplay, however, when developers (and players) start treating it as the main property to pay attention to with gear, and then also base other elements of character development on it (such as skills), the entire system gravitates towards treating DPS as a universal constant and the single most important statistic in the game.

Seriously? DPS is bad "mechanic"? DPS is nothing more than an informative number calculated from a couple of other numbers and should be displayed in games to save you time when comparing items/skills/whatever.

As I already said: DPS is a mechanic when it was treated as a mechanic. How is spell/skill/etc. damage scaling off of DPS not a mechanic, for instance?

1. DPS has nothing to do with damage types.

Almost every game I have played which uses DPS as a mechanic almost entirely downplays multiple damage types of any sort. This is not a direct causal relationship but the tendency is that since DPS represents a constant number, damage types need to be reduced to reflect that.

2. This is completely untrue, you can implement shitloads of other mechanics while still providing some basic DPS info.

Not what I'm talking about.

3. How exactly is this bad? As long as you don't sacrifice diversity for balance, then it's a good thing.

The problem is that the way modern RPGs are designed (single character, action-based, soloing) necessitates a level of consistency between characters that did not work in other RPGs where there are more extreme differences (D&D-based games are one example, Fallout is a classless example). By virtue of having combat be the main focus of the game and having every character class be equally effective at combat, you end up with character classes that are often more or less identical between one another, with only the aesthetics differing significantly. Again, not a direct causal relationship, but that is certainly the tendency in pretty much every MMO I have played.

4. Completely untrue. When calculating DPS values you should always consider mechanics like chance to hit, crit chance, crit damage, etc. Therefore I can't see how this mechanic is in any conflict with DPS.

Maybe some developers do take these into account when calculating DPS, but that would be strange because DPS needs to be an objective and consistent standard between players for gear comparison. Maybe some devs implement it as separate "base DPS" and "character DPS" values. Either way, the only significant deviations of this in modern RPGs tend to be criticals. I have seen very few modern RPGs where to-hit chance is significant outside of status effects, and I have read it straight from the mouths of developers that "missing isn't fun, so we got rid of it."

5. This has more to do with stable DPS vs. burst damage and I don't really understand what problem are you trying to point out here. In PvP environment you can't have too much burst without means of countering it as it results in quick player deaths and not much of a PvP is left. In PvE games (singleplayer, cooperative ones) bursts still lead to quick deaths and in games with limited pool of characters this will most likely lead to high frequency of saving/loading which is retarded design. On the other hand having a very stable DPS (fighters in IE games) and having encounters designed around HP bloat and attrition is retarded design as well.

The point is that in more traditional systems, class distinctions were bigger, in part as a result of less homogenized damage output (which is strongly encouraged by using DPS as the sole metric of effectiveness). Wizards would die almost immediately in melee combat but could unleash powerful status effects, area damage, could charm enemies, and so on. These exist in some modern MMOs, however, because of the need for every character to be solo-capable in PvE play, what you end up with is wizards that have more or less the same DPS as fighters, the main distinction being that their gameplay is based around kiting rather than tanking. Many modern RPGs give fighters powerful active area skills, summons, and other abilities which usually overlap with other character classes precisely because the PvE play has been created with it in mind and now every class needs those abilities in order to survive. The end result is that you have classes that play extremely similarly to one another and the tactical considerations you have to make between them are pretty much identical. Once again, not a direct causal relationship, however, what I am discussing are the reasons for and implications of the adoption of DPS as the primary expression of damage potential - and therefore overall effectiveness.

 

[Infinitron]

It's like the Sapir-Whorf hypothesis of gaming, bros. The terminology we choose to use influences and constrains the ways in which our games are designed.

 

[Johannes]

Once again, not a direct causal relationship, however, what I am discussing are the reasons for and implications of the adoption of DPS as the primary expression of damage potential - and therefore overall effectiveness.

The reason we use DPS is because it's the most sensible way to show and compare damage output. I really don't see what good would come out of using a less descriptive meter.

 

[felipepepe]

The argument that the use of DPS has as one of its "goals" the abolition of to hit chance I found specially retarded.

This is utter bollocks, because DPS is simply a statistic, one that hit chance is a factor in. If your raw, 100% hit chance DPS is, say, 100, then if your hit chance is 50%, your DPS is 50. DPS still exists. DPS has always existed, and always will. Back before it was made explicitly visible, people were still calculating DPS as far back as the text age.

You can't blame a measurement for the system.

False, there was no point in measuring it up before. Think a D&D PnP game, the roles where completly different, DPS would never make sense. What point there is in comparing the damage a barbarian and a cleric do? Versatility was the important thing, a cleric did less melee damage, but could heal and support, not to mention non-combat skills.

As Inifintron perfectly placed, is like the Sapir-Whorf hypothesis, people only think in DPS-related terms now; they are unable to appreciate situational skills or support abilities is they don't affect DPS. Mages are only for fireballs now, not for teleport, light, detect magic, protection from evil, invisibility, identify, knock or the likes anymore...

 

[Johannes]

As Inifintron perfectly placed, is like the Sapir-Whorf hypothesis, people only think in DPS-related terms now; they are unable to appreciate situational skills or support abilities is they don't affect DPS. Mages are only for fireballs now, not for teleport, light, detect magic, protection from evil, invisibility, identify, knock or the likes anymore...

Except that I can. And I think you do also?


Edit: and the more you actually need those interesting special abilities, the more people will appreciate them. If teleporting is more of a situational gimmick, and just nuking everything (or shooting them in the eye, or whatever) is the simplest way to go, why would people bother with more complex tactics?

 

[Juggie]

Perhaps I should have stressed more clearly that DPS does not directly result in the things I said, but rather it creates a paradigm which strongly encourages/inspires other design choices. DPS does not automatically equal simpler gameplay, however, when developers (and players) start treating it as the main property to pay attention to with gear, and then also base other elements of character development on it (such as skills), the entire system gravitates towards treating DPS as a universal constant and the single most important statistic in the game.

Ok, I agree with this. Simpletons need a single property by which everything is measured and they don't want to have many situational advantages to consider. The developers are making games for such people so we are getting simple games mostly. This still doesn't mean that there is anything wrong with DPS. Even if you aren't mathematically impaired and can compute these stats yourself it saves you time and effort if the UI displays it for you.

As I already said: DPS is a mechanic when it was treated as a mechanic. How is spell/skill/etc. damage scaling off of DPS not a mechanic, for instance?

I'm not sure I understand what you mean here or how is this a problem. If you mean Diablo 3's spells scaling with weapon's DPS, then yea... That was impossibly retared move there, otherwise I would say that physical skill scaling with weapon damage or (normalized) DPS makes perfect sense.

Almost every game I have played which uses DPS as a mechanic almost entirely downplays multiple damage types of any sort. This is not a direct causal relationship but the tendency is that since DPS represents a constant number, damage types need to be reduced to reflect that.

Well, personally I'm not a big fan of damage types. Actually I don't know if I've ever played game where damage types weren't just simple gimmick. It's usually based around finding out or guessing resistances or vulnerabilities of your opponents and using skills that deal extra damage to them. I prefer when skills differ in mechanics and not just flavor. Also, I still don't see any connection between damage types and DPS.

Not what I'm talking about.

I guess I misunderstood. If you meant that the games are designed around time it will take for you to defeat your opponents then sure, that's stupid. But I cannot see how anyone would do this in non-MMO as it would be not only pointless but also annoying.

The problem is that the way modern RPGs are designed (single character, action-based, soloing) necessitates a level of consistency between characters that did not work in other RPGs where there are more extreme differences (D&D-based games are one example, Fallout is a classless example). By virtue of having combat be the main focus of the game and having every character class be equally effective at combat, you end up with character classes that are often more or less identical between one another, with only the aesthetics differing significantly. Again, not a direct causal relationship, but that is certainly the tendency in pretty much every MMO I have played.

Well, D&D is designed as P&P game therefore it's designed with DM in mind. You cannot really implement non-combat skills in a PC game with no one to handle every case, therefore non-combat skills usually feel inferior to combat skills and are often even more trivialized. This also means that classes are often balanced around combat effectiveness only. This has more to do with what approaches you allow in your game and it's usually either combat only or combat as main game mechanic. What Fallout did was ok, but it wasn't groundbreaking. Out of all those non-combat skills how many were truly useful? Speech and lockpicking and maybe traps, science and repair in a couple of situations. Sure you could use them to avoid encounters sometimes offering you a "unique" style of play, but I fail to see how clicking on a special reply in dialog is superior to well designed combat encounter.

Maybe some developers do take these into account when calculating DPS, but that would be strange because DPS needs to be an objective and consistent standard between players for gear comparison. Maybe some devs implement it as separate "base DPS" and "character DPS" values. Either way, the only significant deviations of this in modern RPGs tend to be criticals. I have seen very few modern RPGs where to-hit chance is significant outside of status effects, and I have read it straight from the mouths of developers that "missing isn't fun, so we got rid of it."

If you look at World of Warcraft for example: hit chance is a major factor and at the same DPS is probably the main topic for the players (or right behind loot). Even if you want to balance your game around DPS, hit chance is extremely easy to factor in (just multiply your overall DPS by it obviously). The "missing isn't fun, so we got rid of it" design is pretty stupid, but on the other hand with high crit and miss chances the damage output becomes too unpredictable leading to highly random outcomes of combat and this is not a good thing unless you prefer gambling to employing tactics. I prefer systems where you can affect the accuracy of your attacks (think swing vs. thrust in BaK or extra APs for aiming in JA2).

The point is that in more traditional systems, class distinctions were bigger, in part as a result of less homogenized damage output (which is strongly encouraged by using DPS as the sole metric of effectiveness). Wizards would die almost immediately in melee combat but could unleash powerful status effects, area damage, could charm enemies, and so on. These exist in some modern MMOs, however, because of the need for every character to be solo-capable in PvE play, what you end up with is wizards that have more or less the same DPS as fighters, the main distinction being that their gameplay is based around kiting rather than tanking. Many modern RPGs give fighters powerful active area skills, summons, and other abilities which usually overlap with other character classes precisely because the PvE play has been created with it in mind and now every class needs those abilities in order to survive. The end result is that you have classes that play extremely similarly to one another and the tactical considerations you have to make between them are pretty much identical. Once again, not a direct causal relationship, however, what I am discussing are the reasons for and implications of the adoption of DPS as the primary expression of damage potential - and therefore overall effectiveness.

I can't say I liked D&D's design of mages being glass cannons and literally getting one-shotted quite often (at least in PC games) and fighters having practically no tactical options apart from positioning and choosing a target. On the other hand I agree that if you want to make all classes equally useful in all aspects of the game you'll be sacrificing diversity. Balance and diversity are always in conflict and in most competitive and popular games balance is considered priority. That's part of the reason why I couldn't stand Diablo 3. On the other hand some games have extremely imbalanced character builds and there is usually no way of telling which skills will be useful without playing the game first as their usefulness depends entirely on the game's design. This was major flaw of Fallouts (Gambling and Outdoorsman skills as extreme examples). Including such "trap" skills in your game is just plain stupid because it only leads to some characters having major disadvantage and possibly no realistic chance of completing the game. The developers need to think about how much balance is really needed for the game. But this still has nothing to do with DPS, which just allows you to compare things more quickly.

 

[Infinitron]

As Inifintron perfectly placed, is like the Sapir-Whorf hypothesis, people only think in DPS-related terms now; they are unable to appreciate situational skills or support abilities is they don't affect DPS. Mages are only for fireballs now, not for teleport, light, detect magic, protection from evil, invisibility, identify, knock or the likes anymore...

Except that I can. And I think you do also?

Except this isn't about you, or me, or felipepepe. It's about game developers. Gamasutra is a site for game developers.

and the more you actually need those interesting special abilities, the more people will appreciate them. If teleporting is more of a situational gimmick, and just nuking everything (or shooting them in the eye, or whatever) is the simplest way to go, why would people bother with more complex tactics?

Which is why we need more game developers to create games that are not DPS-centric and require these tactics. Get it?

 

[Johannes]

and the more you actually need those interesting special abilities, the more people will appreciate them. If teleporting is more of a situational gimmick, and just nuking everything (or shooting them in the eye, or whatever) is the simplest way to go, why would people bother with more complex tactics?

Which is why we need more game developers to create games that are not DPS-centric and require these tactics. Get it?

DPS-centric is still a silly term for this. Just call simple games, I don't know, simple? Or damage-centric if you must, so you get to include TB games like Fallout.

I don't agree that this has really anything to do with displaying DPS to the player.

 

[Moribund]

Except that no one uses DPS to design single player games outside of this guy's head.

 

[felipepepe]

I don't agree that this has really anything to do with displaying DPS to the player.

Bro, the momment that you give players a number and say "the higher the better", people will only focus on it and every other thing will be compromised to reach higher DPS.

Let me remind you that it wasn't developers that made quest compass what it is today; it was the fucking players! They made a quest compass mod for WoW ages ago, a massive amount of the player base started downloading it and in response Blizzard added it officially to the game, because it was what players actually wanted. DPS is the same shit, look at how a wow player rates a raid:

More damage = better, there is simply no other data to analyse; and thats a fan mod as well, that players loved so much that Blizzard fucking added it to Diablo. And since Blizzard is doing, now it becames a trend... and more DPS games will start appearing, with shit combat and binary combat.

Developers adding DPS is not the source of decline itself, but is the visible side of it, the way you look at a game and see that simply having high DPS is enough to fullfil all it's combat needs.

 

[Excidium]

That has more to do with nowadays developers being part of the retarded crowd that rated performance by simply taking a look at the damage done tab of recount. Then you look at healing received and see that that fucking TOP DPS Mage was putting a massive strain on the healers because he didn't avoid boss abilities so he wouldn't lose DPS. But it was worth it, he did 1% more damage and topped Recount, yeah! I knew so many people like that...

 

[made]

lol that warlock suxx

 

[Excidium]

I'm more suprised there isn't a single rogue, this game sure changed. Who's gonna take all those awful daggers that always drop?

 

[made]

idk, seems like an old parse. doesn't boss hp reach billions nowadays?

 

[Black_Willow]

More damage = better, there is simply no other data to analyse;

Oh yes, because recount is only able to report DPS

 

[Excidium]

His point is not what the tool does, but how it's used. The average wow player just used it to spam on raid chat the TOP DPS list after every encounter.

Just how DPS itself is an useful statistic, but the way game design is being streamlined into MOAR DEEPS at the expense of everything else is bad.

Let me remind you that it wasn't developers that made quest compass what it is today; it was the fucking players! They made a quest compass mod for WoW ages ago, a massive amount of the player base started downloading it and in response Blizzard added it officially to the game, because it was what players actually wanted.

You see, this is also an example of the kind of shit killing video games.

So, quest design in wow was so bad that players had to take their time and code a tool that made it faster. What did game developers learn from that example? To improve quest design? Haha...no, of course not. They learned to offer a quest compass by default.

 

[felipepepe]

Exactly, I know it measures spell interruptions for example, but have you ever seen anyone using it? I only saw once or twice in Blackwing Descent, when my guild raid was wiping a lot and we were searchng for answers...

 

[Excidium]

The CC breaker tab was the best one...Well, when CC was still used in dungeons. Because nobody wanted to accuse himself of being the cause for the wipe, but recount doesn't lie.

 

[godsend1989]

DPS is cheap and lame.

 

[Juggie]

Damage done is usually the only job for the DPS classes in WoW unless they have specific job like kiting, dispelling, interrupting or something similar. Therefore it's quite obvious it would become THE measurement in WoW. Anyway the game is not DPS centric. It's balanced and designed around several derived statistics, DPS being one of them which is quite reasonable for that type of game. This surely doesn't make the game bad. The game has many flaws, but I cannot see any of them being result of using DPS as the main measurement of characters.

It's usually games inspired by WoW that only see the surface of the system and try to reimplement it in their own way making the system look similar at the first glance but being hugely inferior. If any game takes DPS measure and makes it its central measurement without having the complexity of combat mechanics of WoW I can see how this becomes a problem. The thing is you still had many other things to consider besides DPS in WoW, in games inspired by WoW you rarely do. Therefore I don't think that DPS is the real cause of the problem here, instead a design that tries to imitate a popular game without actually understanding its mechanics or trying to design own meaningful game systems.

P.S. I haven't played the game for about year and a half, so some things might have changed, but I was referring to the core design of the game and I don't think they haven yet revamped that part.

 

[Harg Harfardarssen]

Well, this observation could also support the argument of the gamasutra article. If a large chunk of game developers are designing games around the concept of DPS as a central measurement and 2) these games are simplistic trash as a result, then it really doesn't matter that they are doing it because of an erroneous view as to WoW's success. They're still doing it, and its still leading them to make shitty games.