Ep. 094 Genre Study: ARPG

Today we're going to do a genre study on one of my favorite genres, A-RPGs. Hey everyone, welcome to the 94th episode of the Game Dev Field Guide. I am your host, Zach Avelli. You can find me on Twitter @_zachavelli_ and tune in for Game Dev Streams at twitch.tv/zachavelli_ We also have an open community discord, there's a link to that in the show notes. And if you're looking for a longer game jam, I'm about to start the APRAE that is April/May. Monthly game jam, we're going to do like the last two weeks of April and all of May. So yeah, that's an extra long game jam if you've been looking for something that definitely gives you a deadline, but gives you a pretty generous deadline. Maybe consider joining that game jam. I'll release the game jam page with the theme announcement on Friday. With the intro out of the way, let's move on over to the Game Dev Challenge. The Game Dev Challenge is the part of the show where I provide a prompt to the listeners, and it's intended to be like a 15 minute exercise to reinforce the themes of the last episode. If you'll remember, the last episode that is episode 93 was all about math concepts for game developers, and the Game Dev Challenge for episode 93 was to solve a game design or development issue with one of the math concepts from the episode. Show your work so others can follow your process. We got lots of really well explained posts with all sorts of useful math concepts, so many of them I didn't even talk about in the episode. But as you all know, there can be only one winner, and the episode 93 Game Dev Challenge winner is Harness. Now, Harness made a really long post, but it's about something that we talked about before. I don't think it was in the last math concept episode, but I know that I've talked about this before, and that is a Monte Carlo simulation. I believe it might be under a balance episode. That definitely seems like something I would cover in a balance episode. I'll sort of paraphrase Harness's post, because, like I said, it's a long one with many figures, but I really would encourage you to go look through it if this sounds like something that you want to learn, because it'll be useful for your games, and trust me, there are lots of good ways to use this. Harness's post says, "The concept is called Monte Carlo Simulations, which is a fancy name for Ain't Nobody Got Time for Mass, also my computer counts better than I do." So, instead of calculating complex probabilities of certain events, you can just run 100,000 simulations and look at the results, which, with a sample large enough, will approximate it really quite well. I used this when I was developing my first-ever jam game, a dice-based game called Trinity Developer, featured mechanics such as reroll all ones, or reroll all dice on failure. As a game designer, I was interested in the dynamics created by such mechanics. I wanted to know how probabilities of certain roll scores are distributed. Obviously, I didn't feel like digging through textbooks on combinatorics, so I made a super complex Monte Carlo simulator, and I want you to know that the Monte Carlo simulator has got the alternating caps thing to make it seem all special or something. Anyways, mimicking different mechanics with a couple of if/else statements allowed me to learn what I needed. And then he shows a couple pictures of rolling a certain score, or the way that they use this Monte Carlo simulation, and more importantly, the figures where you could get some real takeaways from it. For example, there's a graph that says the result from rolling a one D6, and it says, as you can see, all probabilities are equal. It's called the uniform distribution, and we can see from the graph that, yeah, pretty much everything is all equal. But now when we roll two D6s, we get a different graph. It's more of a normal distribution or a triangular shape. And you can see more and more, the more D6s you roll, the more it fits kind of the bell curve, aka the normal distribution. It's a really good graphic, actually, to just explain what a normal distribution is as well. But with these things graphically shown, we can start to make game design decisions, and it looks like Harness did start to make game design decisions to change things. And the graphs make it a lot easier to see the result of your changes. For example, if you implemented the rule you can roll two D6s and you choose the better results, you get a brand new graph. It's more like a staircase or a straight line up, where six is the highest. You can see that the chance of rolling a one is now below 5%, and the chance of scoring a five or a six is now greater than 50%. It's a really great post of showing exactly how to use a Monte Carlo simulation for this situation. And you know me, I love graphs, so yeah, I thought it was a really good post, and I'm glad that Harness won. So congrats to Harness for winning the Episode 93 Game Dev Challenge. For Episode 94, I want you to design a character class with a good structure for character builds and item progression for an ARPG. Today we're going to talk a lot about ARPGs, and two of the core things we're going to talk about are character progression versus item progression, and the idea of a character build being something that grows on a structure that the designer makes. And yeah, for this Game Dev Challenge, I want you to design the structure. We'll talk more about how exactly to do that later, but I'm sure after listening to this episode you'll have some more ideas, and if you do, you should just go on over to the Community Discord and leave a post under the Game Dev Challenge channel. With the Game Dev Challenge out of the way, let's move on over to the body of the episode. Today's episode is a genre study about a genre that is near and dear to my heart. It's a genre of games that I've loved ever since my childhood, and I'm so excited to talk about them today, and that is the action role-playing game, or the ARPG genre. Now, right off the bat, there may be some confusion here, especially for those who are unfamiliar with this genre specifically, because there are games that are called action RPGs, and then there are ARPGs, and they're two distinct things. For example, Hogwarts Legacy right now on Steam calls itself an action RPG, and I would say that accurately describes it, but Hogwarts Legacy is definitely not an ARPG. I know that's really confusing. At some point, the two things split, or whoever decided to name these got some overlap, but yeah, I wanted to get that out of the way right up front. They both have the same words, they're just two different things. If you see a game with it spelt out, like action RPG, it's probably in the vein of something like Hogwarts Legacy, Dark Souls, Witcher, Assassin's Creed, etc. If you see a game with it abbreviated as ARPG, then it's likely going to be more like the topic of today's episode. ARPGs are like the Diablo series, Path of Exile, Torchlight, Last Epoch, and Lost Ark. ARPGs are often defined by a few key characteristics that make them easily recognizable. The core gameplay loop usually looks something like kill an enemy that drops tons of loot and experience, you search through all the dropped items for item upgrades, you unlock new skills with your XP, that way you can kill better enemies for more loot, for more XP, to kill better enemies for more loot, for more XP, so on and so forth. I've seen the core gameplay loop described as a loot treadmill, and I think that's a pretty good way of succinctly describing it. They often have a huge focus on building your character with large skill trees and millions of procedurally generated items. Ironically, the role-playing part of it is less story-focused and more character-building focused. Think about those people who play D&D for the numbers and the builds and the combat, and not so much the background storylines, which I know for a lot of people is like a D&D sin. Oftentimes in these games, the items are randomized heavily, and most of them will not be useful. You're looking for a few diamonds in the rough of just piles of items, and you're just looking for the couple that are going to work for your build. It definitely scratches this gambling itch that seemingly all humans have, where you're always waiting for that next item drop. From a gameplay perspective, these games are usually played in an isometric third-person view, somewhere where the camera isn't perfectly looking straight down, but the viewpoint is from above. While they almost always have a campaign, many players enjoy them for their endless end-game content, or the end-game here is really the gem of modern ARPGs. Trying to push your character to be game-breakingly powerful is my favorite part of ARPGs. ARPGs are a genre that has been around for a long time, and it's gone through several evolutions. What's interesting about it is it's still evolving and still as popular, or even maybe more popular than it was back in the day. RTSs, for example, are another genre that got popular around the same time, but they've kind of fallen away. Like I said, ARPGs continue to evolve and be amongst the most popular of genres, and I think that really speaks a lot to what makes the genre so great. They can be played over and over, and they are just endlessly compelling. There's always that new interesting thing or possibility right around the corner, that next loot drop, that next skill combo you haven't tried yet, that new end-game goal. There's just so much potential left in this genre, and it's why it's one of my favorites. I think when studying this genre, or any genre for that matter, it's important to learn about the iterations and the evolutions through the years, because it speaks to the game design lessons that were learned along the way. And we can just pick up those lessons, we don't have to relearn them the hard way ourselves. So I think I'll start this episode with a little bit about the history of the genre. And throughout the history, I'll mention some game design lessons we can derive from it. So the genre really starts with the original Diablo in 1996, and according to the creator, David Brevik, he wanted to make an RPG that wasn't so much focused on the story elements, and more about the action and the loot. One example of this is he thought character creation in RPGs of the time took way too long. He wanted in Diablo for the player to get straight into the action in just a few clicks, so they really, really diminished the character creation part. At the time, RPGs were often about making characters with these deep backstories, and the story would interact with your character. But in order to do that, you had to take a long time setting up your character customization. With Diablo, there was no backstories or anything to worry about. You just picked your character class and got straight into the action, and the customization would happen throughout the game via the items and skills and attributes. This, I think, already is an important lesson in pacing and building something different. Diablo, from its inception, was built to be something faster in pacing than normal RPGs of the time, and so there was no need to keep that vestigial feature of an RPG of the deep character creation. The game is just paced much better when you can jump right in and start killing monsters. And you don't really lose a core strength of the game, because this isn't story-aligned roleplay anyways. So again, this is more of a good example of playing to the strengths of the game. Another thing that came up during the development was turn speed. The Diablo original was heavily influenced by true roguelikes, which are all turn-based at the time. And David Brevik, the creator, thought that it being turned-based was key to the game's identity, because of the tension that was built when your character was on low life and you had to make a decision. This idea of, "Oh, this is a really big turn," and the game pausing for turns was important. It was an important tension-building moment. But after getting some notes from the team, some members suggested a real-time approach without turns. David was very much against this idea. To him, it broke the tension and didn't give the players enough time to soak with the decision. And there was lots of back-and-forth about this issue. Eventually, they put it to a vote, and David's side lost, despite being the creator and director, and this going against one of his core ideas and core tenets of the game, they turned it into a real-time game. He was actually so frustrated about this, that he sent everyone home on a Friday afternoon, and he decided in his frustration to just build a quick real-time prototype and test it on his own. And on the very first combat encounter of this new prototype, he found out he loved it. It was awesome. It worked so perfectly, and turns out it wasn't as difficult as they thought it would be to switch it over. It was the birth of real-time combat in ARPGs, and I think there's another really good lesson in here. Sometimes, as designers, we get so locked in on the core vision, because we know how important having a strong core vision is. And David's vision was strong. He was right about pretty much everything, and birthed a chart-topping genre. One of the things he was wrong about was the idea it should be turn-based, and it's not really even a bad idea when you think about it. There was sound reasoning behind it, with the whole building-the-tension thing, but one small change, one that seemingly didn't work with his vision, improved everything and turned this into what would have been a pretty good game, into an all-time PC game genre-defining classic. I think this goes to show that it is impossible to design a game 100% right from the start. You have to find some of the fun along the way. And when lots of well-informed people see something that doesn't necessarily gel with your core vision, it may not hurt to prototype and see it. I mean, it will hurt, because anytime you add something or change something about your core vision, it just doesn't feel good. Entertaining possible improvements outside the vision is dangerous for the scope and identity of the project, but so is taking no risks or being so locked in on your first draft. This is a skill that I personally really need to work on. I tend to be way too conservative with the original scope and vision, and I think I need to explore more within the bounds of the scope, but certainly outside of my core vision, especially when other people see something that I can't. Okay, after Diablo 1, of course, came Diablo 2. Diablo 2 is one of my all-time favorite games, and I've talked about it so much on the podcast. I think people are likely tired of hearing about it, so I'll sum up this chapter of the history quickly. Diablo 2 took everything about Diablo 1 and improved on it. It had a really chaotic development, but a strong guiding light for design. They perfectly nailed character progression and pacing in many, many aspects, and I think Diablo 2 still has the best pacing, itemization, and progression of any ARPG. It is the perfect mix between complex and streamlined. Episode 12 of GEFG has a lot more about it, if you're curious about that, but yeah, there's a reason why it was one of the first episodes I did, and the system design of Diablo 2 is really something that inspired me to want to become a game developer. There's so much more to say about Diablo 2, about its character progression, how it basically invented and popularized modern skill trees, and how that progression structure is used in many genres today. I think it's an all-time classic video game, and it's a must-play for anyone who wants to be a designer. So that's enough about Diablo 2, let's move on. Eventually, we get what I think is the first fork in the genre in the early 2010s. Diablo 3 and Path of Exile come out, and they approach the designs of ARPGs almost oppositely. Diablo 3 streamlined that progression curve, sped it up, and brought it to an increasingly diverse, larger, but more casual audience. Path of Exile, on the other hand, dialed up the complexity of the progression to an insane and amazingly creative degree. Now, of course, they did this to stand out. Diablo 2 was such a fixture, even after all the years it had been out, the game still had a lot of popularity, and you had to do something different to get people to stop playing Diablo 2 and play the new thing. And so I think Diablo 3 and Path of Exile, taking Diablo 2 and making something different in opposite directions, is the start of the fork that would define the landscape of ARPGs all the way up until the present day. Currently, I believe we're living in a golden age of ARPGs. There is something on this spectrum for everyone. You have all the way from the extremely streamlined Diablo 4 to all the way on the other side where you have things like Path of Exile 1 or Path of Exile 2, which have insane difficulty, complexity, but also some of the most interesting progression in any game regardless of genre. If you're wondering where I am, where my favorite place to be is, I think I'm maybe leaning more towards the center. Path of Exile can get a little too complex for my taste personally, and I think my favorite ARPG to play right now is probably Project Diablo 2, which is a mod that stays true to the original with some more modern updates. Although I will say I really enjoyed Last Epoch, and Last Epoch kind of fits in between the complexity of Path of Exile and the simple streamlineness of Diablo 4. I'm really excited the next update for Last Epoch actually comes out tomorrow, so yeah, I'm really excited to give that a try. Regardless, with each of these games in this new spectrum, new ideas and systems are figured out and improved. With all of this diversity and different approaches to the ARPG game design, we have an extremely healthy ecosystem where there's something for everyone and games are able to take the best ideas from each other. This is why I believe we're living in the golden age of ARPGs. Just in the time of me writing this episode, we had a new update for Path of Exile 2 that I didn't enjoy so much. Now, tomorrow we have a Last Epoch update that I hope I enjoy. I'll probably jump on the newest season of Project Diablo 2 whenever that comes out here soon. There's so much to play and so many interesting ideas. Yeah, I'm just really thankful to be an ARPG fan right now. At this point in the episode, I think I'd like to move away from the historical stuff and let's start talking about the design evolutions and tips. Like, if we were to make an ARPG right now, maybe your team or you as an individual are working on an ARPG design right now, what are some things we can learn and tips we have for figuring out a sort of guiding light to making these games? So here are some things that you might want to consider. Let's just start with the idea of what's at the core of the design. Essentially, we're looking at some kind of progression curve that is controlled by a steady and consistent character progression, like active skills and passive skills, attributes and character leveling. And then it's boosted with some more erratic spikes and the spikes are caused by the slot machine loot system. The super powerful item or the really great item that drops and gives you that power spike is the compelling feeling that your player is chasing. The character progression is that longer and slower but more consistent burn. And the loot slot machine is more like the firecracker that keeps things exciting. All of your system design decisions I think should be done with this progression curve in mind. The character structure and progression is the consistent baseline and the loot is the exciting spikes. Let's take a second and talk a little bit more about character progression. In the sense of character progression, I mean that in ways to progress the character without items. This usually means something like leveling up and getting points for a skill tree or points to add to your attributes like strength, dexterity and intelligence. This works to form the backbone of your progression. I like to think of it as like a supporting structure that the item progression or the character build then grows on. And we can design different ways of supporting that progression. For example, we might have a skill tree with lots of things that make your character more tanky. Armor abilities, passive abilities that automatically activate when you have less than 50% life, strength bonuses, things like that. This would all build a supporting structure for a very tanky character. One that I might imagine where it comes to loot, they would be looking for big chunky armor pieces, maybe a tower shield or a big two-handed sword. Conversely, we could go a different direction with the supporting structure. We could have a skill tree and attribute bonuses that focus on moving really fast and dodging enemy attacks. You're doing lots of smaller attacks but more of them. You could then build a skill tree with that in mind, maybe a few versions of that idea. We could do one that does lots of small hits and you're hoping to get a critical hit with so many chances. Or one where each individual hit poisons the enemy and while one hit and a little bit of poison is not that big of a deal, the poison builds up really fast because you're hitting the character or the enemies rather really fast. With a couple versions of the core idea, we've built the framework for the item progression to grow on and for the player to have some interesting decisions and directions they want to take it. Of course, what I'm describing here is basically character classes and they've been a part of RPGs for years. But I think the idea of building the structure that the item progression grows on is unique to ARPGs. For those that don't know, a big part of ARPGs is the theory crafting and coming up with the builds of the characters. We refer to all the choices you make in a skill tree and attributes and the items you carry as a character build. And it's important for the game's longevity and interest to have good build diversity. How you structure the character progression is extremely important for build diversity. It's a balancing act between building a robust and flexible structure with many paths to go down and an overbuilt confusing structure with no identity. Sure, you could just give the player all of the options and that does help a lot with the RNG protection for item drops because odds are you can find a way to make just about any items work with a really flexible build. But that leaves the door open to many bad builds or things that you didn't design the rest of the game around or just plain confusing the character. You're not really leading them to a fulfilling a fantasy or coming to a coherent build. On the other hand, if you build the structure to only support one or a small few builds, then it becomes really reliant on the RNG of this. Like, it works great with certain items but you need those items first. You only have the one option and the one option requires this specific item and whether or not you get that item is up to luck. It does, however, build more to that class fantasy and identity. It feels good to pick the night class and then wear the big armor and have the big polearm and hit for lots of damage and be able to take lots of damage. I think that's another important part of class design. It's not just if the numbers work but the roleplay of it as well. After all, roleplay is two of the four letters in ARPG. It's that mixture of the classic progression and the fantasy roleplay that I think really works here. Imagining the character in your head and making the numbers work to achieve that is extremely satisfying. I think the more specific you get with the class design and class fantasy, the better it is. Like, we could just have a dark wizard character, dark magic spells, intelligence as the attribute to scale, they wear a robe, etc. Or we can get a lot more interesting with it, like a warlock class that has one of its first passes in the skill tree, make you use your health instead of mana to cast spells. That singular idea becomes pretty interesting with just that one implication. Do you want intelligence now or would you rather have more strength to build up your HP since you're using your HP as mana anyways? Do you wear robes or maybe something a little more defensive since you're going to be losing life from casting spells? You probably want some armor to protect the little life that you have. As designers, what kinds of skills should we give this player to trade their life for? Is this an opportunity to build a flexible and robust structure and then let the player grow their character on it? Maybe we have some parts of the structure focused on summoning demons. You trade your life to bind powerful demons to your will. Maybe you then can get the life back by stealing it off the demons or having them steal it for you. Maybe another part of the structure allows you to steal the life from your enemies. That way you can get your life back while casting spells from the enemies that you're using the spells on. You're constantly balancing this idea of trading your life for powerful spells that you hope hit the enemy and recoup the cost. What if you use a big spell and you miss or the enemy blocks it? You can start to see that this idea has all these interesting implications. Maybe there's a really powerful passive skill at the end of the tree that permanently sets your life to one but gives you access to unlimited mana. Of course you wouldn't want to take the spend your life to cast spells and this passive because you would have one life/mana. That would be bad. You probably want to skip that first one and then use this one at the end. You only have one life, you basically cannot be hit, but you have so much mana that you'll have lots of ways to solve that issue, hopefully. You can see just from a few passive ideas that kind of change the build, we can come up with five or six or seven different kinds of ways to play this class that all still fulfill that powerful class fantasy. It's a fun structure for the players to grow lots of different but related character builds on. Just by naturally playing the game, the character is going to progress. This provides that baseline progression and it's going to feel really great when they finally get that item that unlocks new growth potential on the structure. And with that idea, I think we should move to the item progression ARPG tips. I think the item progression is maybe the most important part and sort of the face of the genre. The character progression serves as the space or the maze or the structure or maybe like a puzzle. And the items are the pieces used to solve the puzzle, to navigate the maze or build onto the structure. For example, we just saw all the possibilities of character progression with our warlock idea, but it's the items that really solidify and actually build the character towards a specific build. Like most things in game design, we're balancing between two extremes here. If the items are inconsequential to the build, meaning all of your progression pretty much comes from the skill tree and the attribute bonuses, well then the best part of the ARPG and a huge piece for the core game loop, which is the loot, is sort of broken. What's the point of building a strong character in order to kill monsters and bosses if the loot is bad? The whole reason you're killing the monsters and the bosses, the next part of the game loop is to get the loot. The game loop is build the character to kill the enemies to get the loot, so you can build a better character, kill enemies to get the loot. But if the loot's inconsequential, that arrow is broken. So yes, we need the items to be a core part of the progression. The other extreme is to have the items be the sole progression method, why even bother with the character progression part? This is dangerous because it relies on RNG to guide the character progression. What happens if the player doesn't get any good items from the random drops? The progression curve is gonna flat line. It's actually gonna compound too, right? Because the less good items you get, the worse build you are, which means you can't kill more monsters, which means you can't get more items. So you can see the difference here or the need for that baseline progression that a character progression gives us. Well, what if we just made all the items good, though? We wouldn't need the character progression if every item that dropped was good. Well, if all the items are good, then you lose that special feeling of actually getting a good item. It's like having steak for dinner every night, you lose the appreciation for it. So like I said, the best way to do this, I think, is to balance all things. We know we need a character progression to provide that baseline, and the items are there to juice the progression and complete that game loop and give the player interesting pieces to chase and solve the puzzles with. Now, in episode 12 of the Game Dev Field Guide, I talked specifically about loot design, and I think it's definitely worth revisiting if this topic has interested you. As I'm writing this, I'm already realizing that we're gonna need an ARPG genre study part two, because I have more and more stuff I want to talk about. But let's just finish the item progression ideas with maybe some big ideas, and this might be some more overlap with episode 12. I think if you think about the problems with any of the modern ARPGs, or if you have a bad update, it usually comes down to the imbalance of the curve between the two pieces, the puzzle and the items for solving the puzzle, the character progression and the item progression. Let's talk about some practical tips for getting the item progression right. I think the most effective tips I have all come in the same category, and it's some form of RNG protection, and that just means limiting the downside of bad RNG. And that makes a lot of sense, right? That's the biggest danger of relying on item progression is bad RNG. So what are some good game designs or systems for RNG protection? Well, let's look at some great examples from ARPGs. Diablo II, Lord of Destruction, perhaps the greatest ARPG of them all, I think does a really good job at making sure you have plenty of opportunity for good items. Famously, at the end of Act 1, the boss and Dario is actually bugged when you kill her for the first time. It was intended that if you completed the quest to kill her, you're guaranteed by the game's code to get better items, meaning the item drops from her were basically juiced. Players discovered, though, that you could bug the quest and get this juiced bonus loot over and over and over. Rather than fix the bug, Blizzard left it in, even making it an official part of the game for the remaster. And I think they did this because it supports the progression design they had in mind for the game. Diablo II is constantly but subtly rewarding you with extra increased loot. It doesn't explicitly say it anywhere, but at the time, this hidden loot bonus was new and innovative. Later, games like Path of Exile would make increasing the loot bonuses a core part of its endgame, allowing the player to control this extra juiced loot and hopefully limiting some of the downside of the RNG. Diablo II also introduced the idea of allowing the player to hunt for specific puzzle pieces to help their character build in the form of room words. Basically, you collect runes and you can socket them into items to give them guaranteed properties. Properties like extra elemental resistances or more mana, more damage, so on and so forth. This is an RNG protection method because you can always find a way to at least make some progress towards a specific puzzle piece, rather than solely relying on RNG. In my opinion, Path of Exile is actually the best at this. They really look at the idea of crafting the puzzle pieces to help solve your build. In Path of Exile, there are tons of crafting options that allow you to add and remove properties to equipment. You can get these crafting materials by selling and basically recycling items that you don't want. This is a great form of RNG protection for item progression because you still have the upside of getting lucky and a monster will drop that loot that is the exact thing you need. But if they don't, no worries, you're still making progress if it's an item you don't need because you can recycle it into crafting materials. And eventually all of those bits that you recycled can be used to get that puzzle piece that you're looking for. I think Path of Exile does a really great job at balancing the difficulty of this. Contrast that with Path of Exile 2's crafting as it is right now, and it's very RNG heavy. Like Path of Exile 1's crafting allows you often to get specific modifiers or properties on your items. In Path of Exile 2, it's a lot more RNG dependent. You're basically randomly rolling all of the items, and so you lose this sense of constant progression. You're basically just swapping out the RNG for killing monsters with the RNG of rolling the modifiers on the weapons and equipment. So yeah, I think that's a good example right there of how crafting can go wrong, where at least crafting is not really solving the issue it's meant to solve. If your crafting is also RNG dependent, then it's not acting as a bad RNG protection method. It's actually adding to it. The ideal balance, and I think they do strike this in Path of Exile 1, is to have the crafting take just long enough that it's almost always better to loot a good item than to craft it. But crafting is a good plan B. That way you always have the positive side of the RNG. Oh, I got lucky and got that thing I wanted, and you're limiting the downsides of the RNG. Oh man, I haven't dropped a good item for my build in a long time. I keep getting all these good items for other things. I guess I can just scrap them all, and then I'll use those scraps to build that thing I need. You can't have it where crafting is plan A. If crafting is plan A, then we're breaking our core game loop. The kill the monster to get the loot part becomes boring because you'd rather just kill the monster to get scraps to craft. And then how is that any different than just why even have items? Let's just do points, and then you use the points to get stuff. So yeah, I think it's important that crafting is a solid plan B, but it's never plan A. One of my favorite crafting methods or systems in any game is called the recombinator. You can find it in Path of Exile 1, and I think they just added it to Path of Exile 2. Basically, you put two items in, and the recombinator combines them into one item. It removes some of the properties, let's say depending on the rules. Let's just say for the example, it keeps two properties from each item and makes a new item that has two from each. This is a really fun system for those items that are almost good. It's almost frustrating or a let down to find an item that is almost good, but it has that one property that you don't really like or need. It feels like it's more than just crafting scrap, like it's almost good. It sucks that I have to scrap this. The recombinator solves this issue. It's an awesome way to satisfy that feeling. Maybe you take two almost good items, and you put them in the recombinator. And you might, if you're lucky, get the two best things from item one and the two best things from item two combined on a third new item. Or maybe it takes the worst of both, and it gives you a piece of garbage. Well, at least you know now that one really is just crafting scrap. Either way, it's a satisfying design mechanic, and one I would certainly implement if I was making an ARPG. Lastly, when it comes to crafting, I wanted to mention Last Epoch. It launched its 1.0 version in 2024, and its second update is actually tomorrow on April 17th. I think it's the perfectly non-overly complicated, but also robust crafting system. I think it really maximizes those opportunities for players to protect themselves from bad item drops without overshadowing good item drops. If I was working on an ARPG right now, I would definitely be studying Last Epoch's crafting system. One more idea I want to discuss with item progression is the idea of uniques. Unique items in ARPGs are often the rarest items. Instead of having random properties, they have defined properties, things that designers specifically pick. And these properties are usually build defining qualities that you cannot get in other places. I think there's kind of two schools of thought when it comes to designing uniques. And it's really a question of flexibility. Do you want uniques that are made with a specific build in mind, one that carry properties that are directly related to certain skills or classes? For our warlock example, it would be like a unique amulet called the change of heart. And it might have a property that says demons you summon are now angels with level 15 healing aura or whatever. You can see how this changes a fundamental thing about the warlock class. This is really cool from a narrative standpoint, it's really cool from a class fantasy standpoint, it's really cool. It's really cool from that idea of you are spending your life to cast spells as the warlock. But now you have these angels you summon that regenerate your life, which means you can cast more spells. This is a really unique item, but it's useless to like a knight class or something that has no way of summoning demons. It has the upside of allowing you to create really cool and interesting character changing unique items. However, the downside is there's limited flexibility and use for many of the players. The other side of this is designing uniques with more general and flexible but still strong and distinct properties. We might have a ghostly spirit tower shield, for example, with the unique property of all damage is taken from mana before life. For a warlock character, remember that passive skill and the character progression that sets your life to one but gives you a huge pool of mana? Well now here's the item puzzle piece that really makes a good character build out of this. When you combine this spirit tower shield, all damage is taken from your mana, but you have a huge pool of mana because you sacrificed almost all of your life for it. But this spirit tower shield is not totally useless to other builds either. I can imagine maybe a spell slinging character that gets buffed when they're on low mana. With a shield like this, when you get damaged it might automatically put you on low mana. It might be a good way to proc that property. You can see that this spirit tower shield is a lot more flexible in its use case. Admittedly it's not as useful or character changing, but it's still a cool and flexible item. Uniques designed in this way though are often overshadowed by regular rare items, especially uniques with more niche properties that maybe aren't that useful. Now no matter how you decide you're going to do your uniques, I think one thing is clear. They're often key important pieces to character builds. So a major mistake here would be to not have any RNG protection. Like imagine your player was working towards a key piece to their build and they had no idea if they were making progress towards it because the unique they need is just a random drop. Their progression in their eyes could be, well my character could be really good tomorrow or it could be really good 100 hours from now. I have no idea when this item is going to drop. Now in games where there's multiplayer maybe you can trade for it, but that doesn't seem that fun. I mean it can be fun, trading can be really fun parts of this, but it's basically just like click the button to get the power. I think that's why game modes like solo self found where the character or the player rather needs to find all of their own items and can't trade for them are becoming more common. In my opinion a modern ARPG needs some way to target farm uniques, some way to make progress towards getting that unique. Not just a one button click to trade for it and not just a totally random you have no idea when it's ever going to happen. This usually looks like a unique being tied to a certain boss's drop table or some kind of challenge, some kind of progressive challenge or crafting recipe. For example defeat this boss 100 times to get this unique or this boss has a one out of 30 chance to drop the special unique. So you think oh I'll just kill him an average of 30 times and I'll get it. Some way to make a long term progression and target farm the unique is a really good way to solve that issue. Especially if this thing they can do builds towards and improves their RNG or circumvents it all together, especially if that thing is acquired through higher skill or game knowledge. That's a really good way to dovetail those two pieces of design together. So I'm realizing now there's just so much more to talk about with ARPGs and the script is getting quite long. People really enjoyed this episode maybe I'll make a part two sooner rather than later. I'd like to talk more about monster and skill design, character respecking, the emergence of WASD movement, the content grind of working on a project like this and the logistics of actually doing this as an indie. There's so much left to talk about my favorite genre but I think I'll leave that for a part two. Let me know if you really enjoyed this first part and enough people did I'll make a part two soon. Now let's revisit the stuff we did talk about today. Today we did a genre study on ARPGs like Diablo, Path of Exile, and Last Epoch. Remember the core game loop here is kill monsters, get loot, use loot to get stronger, kill stronger monsters. To get more loot, to use the loot to build a better character and so on and so forth. We talked a little about the history of the genre and how I believe we're in the ARPG golden age right now. We talked about how at the core of these games there's a strong progression curve and it's guided primarily by two parts. The stable and consistent character progression and the fun and chaotic chase of the item progression. Remember character progression is the consistent structure. It's a steady source of progression and acts as a realm of possibility for character builds. Remember not to just look at it from the mathematical sense of progression but also in the sense of the class fantasy. The more descriptive and distinct the class fantasy the more interesting the possibilities. The warlock trading their life for summons and spells versus just the dark wizard for example. Item progression serves as the interesting spikes, the puzzle pieces for the build, the carrot on the stick for the game loop. Remember our biggest takeaway from item progression is to try and reduce the downsides of RNG item loop. Allowing player to affect the RNG of the rewards, a good crafting system and target farming important items like Unix are good examples of RNG item protection. Lastly there was just so much more left to talk about this so make sure you let me know if you'd like to see a part 2. I wrote in the script PS I really want to do part 2. So yeah with that I think I'm going to end today's episode thank you so much for listening. You'll have to tell me what you thought maybe tell me about your favorite ARPG. You can do so in our community discord under the episode discussion channel. Sorry this episode was a little bit late. I'm just like drowning in a million things right now so yeah I'm getting it out well I can. And to top it all off this was an extremely long one as well so just bad timing I guess. Anyways with that I'm going to sign off. I have been Zach of Ellie making my 100th whirlwind build and I'll see you on the next episode of the Game Dev Field Guide. [Music] [Music]