An update on recent technical problems and more from the admin: https://forums.penny-arcade.com/discussion/250292/on-technical-difficulties-mod-coverage-and-other-things/p1?new=1
Options
Defying the Gods: Alternatives and Suppliments to RNG Gear Systems in Diablolikes
Sorenson
Registered User regular
Registered User regular
TL;DR's at the bottom for those who hate walls o' text.
As anyone who's played a Diablolike or any general MMO within the last 12 years can attest to, the random number generators that often form the cornerstones of these games can be a fickle to downright bastardly being. Holding the prospect of great rewards and the once-in-a-lifetime find above our heads, its primary pleasure seems to be dumping upon us heaps upon heaps of laughably useless trash, in some cases taunting us by constructing items in such a way that they could be useful were it not for some glaring critical flaw. Forever egging us on, it can enthrall us for days, weeks, even months before finally, in some half-hearted act of mercy - or pity - it finally beswots upon that mathematical improbability that we have sought in exchange for subjecting ourselves to its every whim and fancy.
Anyone who's played the aforementioned games and been subject to the wrath of the RNG has also asked themselves many a time the following question: why do we put up with this? Why must this be the only way? Isn't there a better system? And with the recent release of Diablo 3 and subsequent discussion on the failings and merits of its current loot system, perhaps this is as good a time as any to finally bring this discussion into the open.
Credit Where Credit's Due: Where the RNG Succeeds
While this post's tone in regards to the RNG gear system as seen in the likes of Diablo certainely comes off as hostile, one cannot discuss how to circumvent or alleviate the ills brought about by said system without first understanding why it's used and the merits that made it such an appealing choice in the first place.
Content Generation: In creating an ARPG (or any kind of game, really) a great many factors must be taken into consideration when designing the tools players will use during the game. There are as many different strategies and playstyles as there are players themselves, doubly so considering the great number of different skill combinations offered to players in more modern Diablolikes, and as such manually creating an item index capable of accounting for all of these would be a Herculean task. By being able to automatically generate items of a great variety, the RNG serves as a developer-end timesaving tool in addition to its ability to provide players with content, albeit of questionable quality in many circumstances.
Player Stimulation: The heart of the Diablolike is, in essence, a slot machine. Its ability to generate items of an extrordinarily vast quality spectrum serves as one of its drawing features as players hope that the next item that drops will be one that fits their particular playstyle. The fact that so many items generated fail to do so serves as a means of enhancing the potency of the one item that does, even if the finding of said item occurs only after many failed attempts and a considerable amount of time invested.
Playtime Padding: Of considerable import back in the heyday of subscription MMOs but of dubious quality in light of their decline, the RNG gear system is an effective means of prolonging the player's completion of the available content by creating larger and larger equipment roadblocks that must be circumnavigated. Due to the gear-centric nature of the player's abilities and the ever-increasing complexity that arises from both difficulty and the availability of more and more abilities as the game progresses, navigating these roadblocks becomes increasingly difficult as the range of acceptable items suitable to use against the blocked content becomes smaller and smaller. The longer the player can be stymied and kept in the game, the more the developer can profit, either by extending the subscription period or by prolonging the player's completion of all available content and thus being forced to develop new content.
A Plague of Vendor Trash: Where the RNG Fails and Damages
Understanding why the RNG can be a good thing (to developers, if not players) we can now explore why the dark side of the RNG undoes much of the good it brings:
Clutter Generation: As the number of generated acceptable items a player can use shrinks, the number of generated items that the player can't use naturally grows. Vendor trash, as said items are commonly known, are considerably damaging to games in which they appear in copious amounts: as many such items can be bought by in-game vendors for various amounts, they thus serve as an additional means by which money may enter the in-game economy, devaluing the current supply and causing the prices of goods offered by players to increase, in some extreme cases to make in-game currency completily useless for bartering between players. Said vendor trash also creates a dulling effect, as players who become accustomed to the increasingly-worthless nature of most items can become increasingly prone to missing items that actually are valuable.
Player Attrition: While the RNG's ability to slow a player's progress through available content can be seen as an advantage, said slowdowns may actually prove to be a disadvantage depending upon the nature of the player in question. While those who are more susceptable to the gamble-based nature may continue to play, others who are more self-aware and strong-willed are much more likely to question the nature of their labors and outright abandon them if they feel the effort invested isn't nearly worth the payoff. In addition to the loss of the player on that current game, said attrition may prevent them from investigating future games produced by a given developer: if the player still bears unfavorable views of the grinding from the last title, they may not even investigate the next.
Farming, de-gaming and dehumanization: As the potential for the RNG to generate a suitable item dwindles, players thus develop techniques by which they are able to maximize said potential, typically by increasing the rate and quality of the items generated to as high a level as possible. To do so, players compartmentalize the game and focus upon exploiting a particular section that meets this criteria, usually built around a specific location or specific enemy that players explore and kill as quickly as possible, repeated in the tens, hundreds, and thousands of times in pursuit of the one lucky break. The farming of particular locals and locales is doubly-damaging, as it turns the game, an item used for pleasure, into a labor in and of itself capable of instead producing frustration; said farming, in being reduced down to a specific set of actions and conditions repeated over and over until a suitable item is found, essentially dehumanizes players that partake in it by virtue of them stripping out all factors not related to the very specific act of farming being undertaken, and indeed the more programmingly-astute of players will create their own programs to do said farming automatically, provided they are able to evade detection and ejection.
Binary Progression: When used in exclusion, the RNG gear system creates a progression system that can result in stop-and-go gameplay: either you find the item needed to progress (and do so) or you stop and are barred by whatever mechanism the game chooses (increased enemy strength, primarily). This creates a rough gameplay experience in which the player is extremily likely to find themselves running into walls as opposed to meeting a more gradually-increasing curve of resistance.
The Payoff: Circumventing and Improving RNG Gear
Knowing the strengths and weaknesses of RNG gear systems, now comes the issue of determining a new means of generating (or at least increasing availability of) RNG gear in such a way that nullifies/lessens its weaknesses while also maintaining its strengths:
Structured vs. Freeform RNG: One of the primary issues with RNG gear seen in the likes of Diablo and Torchlight is that the items generated by said systems can be incredibly diverse due to the minimal or complete lack of structuring in regards to their formation. While said diversity can be a major boon due to the ability to cover a wide variety of playstyles and circumstances, it can also be a major drawback due to the ability to generate cringe-inducingly awful items that go straight to the vendor. By tweaking the item generation system a somewhat more rigid structure, such as by enforcing minimum values at given levels or by bundling particular effects together - an example of which being the item parts system of Borderlands, where items are assembled from a combination of different sections, each being manually crafted with specific perameters that scale with level and the like - the rate at which outright horrible items can be reduced. On the other hand, this also lessens the potential variety of items and requires considerably more input on behalf of developers in order to effectively structure these items.
Player Transactions: While the odds for finding a particular item with particular attributes may be in the millions, the fact that millions of players may be playing a game means that said item will eventually be generated. By providing a means by which players can communicate about and advertise the trading of particular items, players have the opprotunity to circumvent the grind for a particular item by simply purchasing it from someone else. Human nature being what it is, however, the price at which a particular item may sell for (if at all, as observed by the innumerable bragging rights auction listings in Diablo 3) may simply transform the grind for a particular item into the grind for funds or an equivilent item to be used to purchase said item.
Crafting and Enchanting: An item crafting system with considerable depth has the potential to be an extremily lucritive boon to a game in which the RNG is the primary means of generating equipment. In requiring money to create an item, it removes excess currency from the game and thus strengthens it; in requiring raw materials that can be salvaged from items, it creates a new use for the vendor trash that players have become so accustomed to dumping en masse upon the nearest merchant. By allowing players to be able to invest in the creation of an item with the exact particulars they're looking for, it thus alleviates and avoids much of the frustration of farming and grinding for said item, minimizes the downtime spent on said farming and enables the game to remain as whole as possible instead of being broken down to an explicit set of perameters. Even if the creation of items is a process that requires several steps and considerable resources, that players can actually see and measure the progress being made towards that item serves as a great boon versus the blank uncertainty of the RNG.
A poorly-designed crafting system can be a major disaster, however: given too much freedom and not enough risk, players will be able to devise particular combinations of item perameters with the potential to break features and sections of the game, thus eliminating any challenge and risk whatsoever. On the other, a system that is much too restrictive, ambiguous or resource-intensive will be avoided by players and will thus do nothing: the blacksmith of Diablo 3, for example, fails spectacularly in that it is simply a manually-triggered RNG that lacks any structuring to prevent its products from being completily useless, while one popular model of item enchanting, as manifested in Torchlight, actually undoes player progress by destroying the item if the RNG rolls within a specific number range.
TL;DR: RNG-based gear generation systems, while useful, are also greatly flawed: their overuse by developers to serve as a crutch and a game filler is a travesty and the paradigm behind their use needs a considerable overhaul, or at least needs to be paired with other mechanisms to alleviate some of the bullshit they tend to introduce.
As anyone who's played a Diablolike or any general MMO within the last 12 years can attest to, the random number generators that often form the cornerstones of these games can be a fickle to downright bastardly being. Holding the prospect of great rewards and the once-in-a-lifetime find above our heads, its primary pleasure seems to be dumping upon us heaps upon heaps of laughably useless trash, in some cases taunting us by constructing items in such a way that they could be useful were it not for some glaring critical flaw. Forever egging us on, it can enthrall us for days, weeks, even months before finally, in some half-hearted act of mercy - or pity - it finally beswots upon that mathematical improbability that we have sought in exchange for subjecting ourselves to its every whim and fancy.
Anyone who's played the aforementioned games and been subject to the wrath of the RNG has also asked themselves many a time the following question: why do we put up with this? Why must this be the only way? Isn't there a better system? And with the recent release of Diablo 3 and subsequent discussion on the failings and merits of its current loot system, perhaps this is as good a time as any to finally bring this discussion into the open.
Credit Where Credit's Due: Where the RNG Succeeds
While this post's tone in regards to the RNG gear system as seen in the likes of Diablo certainely comes off as hostile, one cannot discuss how to circumvent or alleviate the ills brought about by said system without first understanding why it's used and the merits that made it such an appealing choice in the first place.
Content Generation: In creating an ARPG (or any kind of game, really) a great many factors must be taken into consideration when designing the tools players will use during the game. There are as many different strategies and playstyles as there are players themselves, doubly so considering the great number of different skill combinations offered to players in more modern Diablolikes, and as such manually creating an item index capable of accounting for all of these would be a Herculean task. By being able to automatically generate items of a great variety, the RNG serves as a developer-end timesaving tool in addition to its ability to provide players with content, albeit of questionable quality in many circumstances.
Player Stimulation: The heart of the Diablolike is, in essence, a slot machine. Its ability to generate items of an extrordinarily vast quality spectrum serves as one of its drawing features as players hope that the next item that drops will be one that fits their particular playstyle. The fact that so many items generated fail to do so serves as a means of enhancing the potency of the one item that does, even if the finding of said item occurs only after many failed attempts and a considerable amount of time invested.
Playtime Padding: Of considerable import back in the heyday of subscription MMOs but of dubious quality in light of their decline, the RNG gear system is an effective means of prolonging the player's completion of the available content by creating larger and larger equipment roadblocks that must be circumnavigated. Due to the gear-centric nature of the player's abilities and the ever-increasing complexity that arises from both difficulty and the availability of more and more abilities as the game progresses, navigating these roadblocks becomes increasingly difficult as the range of acceptable items suitable to use against the blocked content becomes smaller and smaller. The longer the player can be stymied and kept in the game, the more the developer can profit, either by extending the subscription period or by prolonging the player's completion of all available content and thus being forced to develop new content.
A Plague of Vendor Trash: Where the RNG Fails and Damages
Understanding why the RNG can be a good thing (to developers, if not players) we can now explore why the dark side of the RNG undoes much of the good it brings:
Clutter Generation: As the number of generated acceptable items a player can use shrinks, the number of generated items that the player can't use naturally grows. Vendor trash, as said items are commonly known, are considerably damaging to games in which they appear in copious amounts: as many such items can be bought by in-game vendors for various amounts, they thus serve as an additional means by which money may enter the in-game economy, devaluing the current supply and causing the prices of goods offered by players to increase, in some extreme cases to make in-game currency completily useless for bartering between players. Said vendor trash also creates a dulling effect, as players who become accustomed to the increasingly-worthless nature of most items can become increasingly prone to missing items that actually are valuable.
Player Attrition: While the RNG's ability to slow a player's progress through available content can be seen as an advantage, said slowdowns may actually prove to be a disadvantage depending upon the nature of the player in question. While those who are more susceptable to the gamble-based nature may continue to play, others who are more self-aware and strong-willed are much more likely to question the nature of their labors and outright abandon them if they feel the effort invested isn't nearly worth the payoff. In addition to the loss of the player on that current game, said attrition may prevent them from investigating future games produced by a given developer: if the player still bears unfavorable views of the grinding from the last title, they may not even investigate the next.
Farming, de-gaming and dehumanization: As the potential for the RNG to generate a suitable item dwindles, players thus develop techniques by which they are able to maximize said potential, typically by increasing the rate and quality of the items generated to as high a level as possible. To do so, players compartmentalize the game and focus upon exploiting a particular section that meets this criteria, usually built around a specific location or specific enemy that players explore and kill as quickly as possible, repeated in the tens, hundreds, and thousands of times in pursuit of the one lucky break. The farming of particular locals and locales is doubly-damaging, as it turns the game, an item used for pleasure, into a labor in and of itself capable of instead producing frustration; said farming, in being reduced down to a specific set of actions and conditions repeated over and over until a suitable item is found, essentially dehumanizes players that partake in it by virtue of them stripping out all factors not related to the very specific act of farming being undertaken, and indeed the more programmingly-astute of players will create their own programs to do said farming automatically, provided they are able to evade detection and ejection.
Binary Progression: When used in exclusion, the RNG gear system creates a progression system that can result in stop-and-go gameplay: either you find the item needed to progress (and do so) or you stop and are barred by whatever mechanism the game chooses (increased enemy strength, primarily). This creates a rough gameplay experience in which the player is extremily likely to find themselves running into walls as opposed to meeting a more gradually-increasing curve of resistance.
The Payoff: Circumventing and Improving RNG Gear
Knowing the strengths and weaknesses of RNG gear systems, now comes the issue of determining a new means of generating (or at least increasing availability of) RNG gear in such a way that nullifies/lessens its weaknesses while also maintaining its strengths:
Structured vs. Freeform RNG: One of the primary issues with RNG gear seen in the likes of Diablo and Torchlight is that the items generated by said systems can be incredibly diverse due to the minimal or complete lack of structuring in regards to their formation. While said diversity can be a major boon due to the ability to cover a wide variety of playstyles and circumstances, it can also be a major drawback due to the ability to generate cringe-inducingly awful items that go straight to the vendor. By tweaking the item generation system a somewhat more rigid structure, such as by enforcing minimum values at given levels or by bundling particular effects together - an example of which being the item parts system of Borderlands, where items are assembled from a combination of different sections, each being manually crafted with specific perameters that scale with level and the like - the rate at which outright horrible items can be reduced. On the other hand, this also lessens the potential variety of items and requires considerably more input on behalf of developers in order to effectively structure these items.
Player Transactions: While the odds for finding a particular item with particular attributes may be in the millions, the fact that millions of players may be playing a game means that said item will eventually be generated. By providing a means by which players can communicate about and advertise the trading of particular items, players have the opprotunity to circumvent the grind for a particular item by simply purchasing it from someone else. Human nature being what it is, however, the price at which a particular item may sell for (if at all, as observed by the innumerable bragging rights auction listings in Diablo 3) may simply transform the grind for a particular item into the grind for funds or an equivilent item to be used to purchase said item.
Crafting and Enchanting: An item crafting system with considerable depth has the potential to be an extremily lucritive boon to a game in which the RNG is the primary means of generating equipment. In requiring money to create an item, it removes excess currency from the game and thus strengthens it; in requiring raw materials that can be salvaged from items, it creates a new use for the vendor trash that players have become so accustomed to dumping en masse upon the nearest merchant. By allowing players to be able to invest in the creation of an item with the exact particulars they're looking for, it thus alleviates and avoids much of the frustration of farming and grinding for said item, minimizes the downtime spent on said farming and enables the game to remain as whole as possible instead of being broken down to an explicit set of perameters. Even if the creation of items is a process that requires several steps and considerable resources, that players can actually see and measure the progress being made towards that item serves as a great boon versus the blank uncertainty of the RNG.
A poorly-designed crafting system can be a major disaster, however: given too much freedom and not enough risk, players will be able to devise particular combinations of item perameters with the potential to break features and sections of the game, thus eliminating any challenge and risk whatsoever. On the other, a system that is much too restrictive, ambiguous or resource-intensive will be avoided by players and will thus do nothing: the blacksmith of Diablo 3, for example, fails spectacularly in that it is simply a manually-triggered RNG that lacks any structuring to prevent its products from being completily useless, while one popular model of item enchanting, as manifested in Torchlight, actually undoes player progress by destroying the item if the RNG rolls within a specific number range.
TL;DR: RNG-based gear generation systems, while useful, are also greatly flawed: their overuse by developers to serve as a crutch and a game filler is a travesty and the paradigm behind their use needs a considerable overhaul, or at least needs to be paired with other mechanisms to alleviate some of the bullshit they tend to introduce.
Sorenson on
0
Posts
This does cut down some of the variety in the game. There are far fewer ways for weapons to be named than in Diablo.
Since a decent weapon is effective and a really good weapon is only somewhat more effective you feel like less of a winner when you get something good. It is less of a good thing. This is probably made up for by there being no grind.
It effectively removed vendor trash by making item effectiveness closer. There are always decent items to pick ip and use or sell and the money requirements on new buys are never astronomical. In that game you can leave trash where you found it, on the ground (take that environment!).
Edit: fixed some spelling. Mobile posting is hard.
but they're listening to every word I say
I don't think the motivations behind 'RNG based items' is as malicious as you think, diablolikes (great word!) of course inherit their RNG soul from roguelikes.
In the roguelike, all forms of random generation are not because the devs were too lazy to fill out dungeons, their intent with using RNG is to individually tailor an experience to the player. Roguelikes have that amazing 'water cooler' effect where even after losing a character, you have some unique character life story to relate to people. I have yet to hear something similar for a Diablo 3 HC character wipe.
At a certain point of scope, you could just would not put enough time into a game to replicate this without using some generation.
Now I know from what I've seen the items in this generation of Diablo are a step back, but I'd really like to hear some analysis into why. It's tough to buy into a critique where you just slather the system as bad with no data on it. (Granted early after release, its tough to assemble the data)
Especially with regards to dismissing the structuring they use as non-existent. I mean their item generation IS structured, things aren't so black and white. Writing on how the structure is weak in explicit terms would have been a nicer read.
My best guess for the rationale behind why it is weak, is that Blizzard bit off more then they could chew and had no devtime to test more advanced item types.
This is because they need the items to be relatively stable for the endgame content of PvP, inferno mode, and the real money market to work decently. (which for only one of those making launch in itself is quite telling).
Also the narrowing of the item system really just reflects the narrowing of the game in general, but I'd rather see a design discussion then a diablo one. But less replicable skills means less item affixes that are deeper. You can't distribute a monk skill through an item to a wizard, but in Diablo 2, this was really common between classes.
I think crafting has interesting connotations, but really doesn't work as a throwaway. It needs to be built into the system quite extensively. It really bothers me that the economy of Diablo is always scaling up though, so any solutions it could bring are a bit iffy, I think.
For example, If you do something neat like let players forge recipes of amazing items they acquire through a drop, and sell the recipes to replicate it that would be a great crafting target.
However solutions like this suffer from the fact that eventually everyone is walking around with the ideal gear, and you've homogenized the items again.
Diablo needs something to create an item 'meta' so there could be a reason to devalue certain gear, but without PvP being competitive I don't know what it could be.
D3 Legendary items are another case of RNG weirdness. People occassionally post in the D3 thread about how they've seen their first Set or Legendary after 100+ hours of gameplay - yet, vendor-sold blues can be BETTER at max level. What the hell?
His tl;dr is fine, professor. It's an in-joke in the gaming community. I think we've all seen a tl;dr on much shorter and less content-filled posts.
Crafting is not something I like that much because it tends to break down into situations where everybody crafts the same optimal things, or many key crafting ingredients are rare drops which results in a grind anyway. Given that the grind is going to exist, novel surprises (e.g. RNG generated loot) are sometimes more pleasant than "well, time to get 9 more of these Godly Dragon Scales". Having random drops that give you big bonuses for things that you didn't expect to use are also sometimes nice because they can cause you to change what things you prioritize in order to maximize the value from the item you just got... something that is more done in respec-friendly games.
Interesting point. Do you think this is a weakness with the Diablo series in general, or just Diablo 3? Why do you think roguelikes have the water cooler effect and Diablo 3 does not?
As user mods like Equivalent Exchange have shown, it's easy to make such a system scalable to have 'epic' tier equipment that a player will see as very valuable but will not be so stupidly rare that a player will consider it an impossibility to acquire.
Really, in my opinion, Diablo 2 was pretty terribly designed from the standpoint of loot. It's about tedium more than it is about exploration, and when you realize that as a player, the game becomes pretty boring (or it does for me, anyway).
The funny thing is that Diablo, the original Diablo really was an attempt to take the roguelike formula and make it more compelling and action based. Over time however the series just became about loot drops. That is pretty much ALL it's about now, nothing else.
A lot of answers about liking and disliking RNG systems will change depending on how the above - and more - are answered. It's not about an entire system being bad and good no matter the circumstance, it's about proper implementation. Same goes for any video game system - random encounters, quick-time events, stat allocation, cover-based systems, etc. Sure, the industry identifies what sells and pushes it to retarded extremes, but that's bad design, not bad systems.
I think the Diablo 2 method - which WAS different from part 1 - ruined a ton of things and created a misinterpreted perception by the industry of what players demand. In my view it created the problem of saturation on two layers. First, too many item modifications, which goes in hand with too many stats. In my view, when you have a game based on mass slaughter, multiple stats aren't necessary. You should really macro manage vs. micro. Stats and counters of that degree matter more when each individual encounter, be it one monster or just a few, matters a lot. It's naturally a more calculating position. I get that it's supposed to be more engaging the way they did it, but it turns out you just had to game particular stats to get the most bang for your buck. The second layer is actual item drop saturation. I don't think, in the first game, items dropped a whole heck of a lot. And when they did, magic items were few. In Diablo 2, they dropped plentifully. It ruins the value of what you have on hand - or your options - because you can just farm drops.
So to sum up my take as briefly as possible:
- RNG should guide character growth subtly, not be content dropped all the fuck over
- RNG is NOT bad or good at face value; what works for Torchlight or Dungeons of Dredmor does not work for the other
- Negatives are a good thing, making decision-making a used real-world skill and increasing item values inherently
- It's okay to not like things because everyone is allowed to have their taste / opinion