I got about 45 minutes with Mark Roberts and Octavian ahead of Curse of the Allflame, and I went in aiming to dive a little deeper into the mechanics of Path of Exile: Curse of the Allflame and the future of Path of Exile. As always, Mark and Octavian went above and beyond with their responses, pulling back the onion that is game development and diving deep into the upcoming league and all its new features.
Lightly edited for length and the usual verbal debris.
Sockets
OUTPUT LAG: With the big change to socket colors, not fully going away but being changed, how’s that going to affect the existing stuff? Pre-existing corrupted six links, old uniques with fixed colors. Do they get grandfathered in with those same colors and then get the quality bonus, or are they changing too?
GGG: Every unique that had a particular interaction with socket colors in some way has had a look. Most of them have been changed, with the goal being to make them more viable and buff them within the context of the new system. For instance, there’s a unique ring, the Pariah, that has a different bonus depending on the color of the socket on the ring. Pretty much all those bonuses have just skyrocketed up. I believe if you have a white socket on it now, it can give you like plus 200 something life. It’s a little crazy.
It’s interesting you ask the corruption outcome, because I don’t want to spoil all of them, but pretty much corruption outcomes on gear have all been significantly buffed. It’s still the same outcomes where it can brick a rare or get an implicit or do nothing. There might be a fourth I’m missing. But that outcome where it gets an implicit is significantly stronger now. There’s a lot more really powerful options there for basically every gear slot. And then re-rolling the socket colors and links, that’s inherently more valuable because getting off colors is harder, or getting colors in general is harder.
Some other examples there, and hopefully this is still accurate as of the last time I checked. Tabula Rasa is still just a six link, so there’s no colors, but it’s a six link, which is amazing from the start of the game. That’s still fine. Something like Skin of the Lords and Skin of the Loyal generates with all colored sockets, which can obviously be very powerful.
It’s still unmodifiable and you get random ones. You still are looking for the right color combinations. But the important thing there is it doesn’t restrict what you can put in. That’s the big win. Even if you get one that has four colors you want and two you don’t, well, then use it with four colors you want and two you don’t.
OUTPUT LAG: That’s huge. When I first saw it I was like, oh man. But then it’s like, no, it still matters. You handled it perfectly.
GGG: I agree. I’m very happy with that change.
Charts in Curse of the Allflame
OUTPUT LAG: It’s nine per voyage, right? When you’re putting them on the board.
GGG: Yes. Your very first voyage is an introductory one, you only need four. So it’s a two by two instead of a three by three. Any voyages after that, it’s nine charts to a voyage.
OUTPUT LAG: Cadence wise, say you’re running white maps. Should you expect a voyage every session, or is a full board something where you’re looking at running a ton of maps? And are the charts tradeable?
GGG: The charts are tradeable. Interestingly, the way the charts work is the majority of the properties of the chart are determined on dropping it. So the explicit modifiers on the chart, what kind of chart it is as in where it will lead to on the sea floor. However, the implicit modifier, which is what’s used to determine what bonuses it gives in a voyage, that’s not determined until you run the chart. So that aspect of it is kind of not tradeable. Once you run the chart, you can’t trade it again. And that’s when you reveal what the voyage modifier is. So it’s a split where you can trade the key, but you can’t trade the outcome, if that makes sense.
As far as generation of charts, you have a chart generated every single area within Curse of the Allflame and a 20 percent chance for another one. So you’re getting 1.2 charts on average per area you run. And you need nine for a voyage. So it’ll take seven or eight areas on average to get enough for a voyage. There’s a bit of a margin for error there, because voyages do require you to have your charts placed in a way where they connect with one another properly. It isn’t too terribly difficult to do, but it does mean it’s technically possible that nine specific charts might not work for a voyage. You might need to get a tenth or eleventh to make it work.
From the behavior I’ve seen of players on the internal alpha we’ve been running, most people like to build up 20 or 30 charts and then go run them all in a row, because each chart takes about a minute to run if you have a relatively quick build. And from there they’ll do two or three voyages in a row, because people just seem to like to group their content and how they run it. But if you just wanted to get a voyage done as quickly as possible, you’d do that in a relatively quick session. That’ll take you running eight, nine maps and then running the voyage itself, and a voyage is probably two to three times the size of a map, because it’s a pretty large area. It’s three by three charts.
Allflame Crafting
OUTPUT LAG: With Vesper, correct me if I’m wrong, it’s kind of like a Hinekora’s Lock, just with a certain number of previews instead of the one.
GGG: Yeah, it’s quite similar. It’s actually kind of funny. The system specifically doesn’t work with Hinekora’s Lock items, because it would be pointless. Every outcome would just be the same. It’d be like, this is what’s prophesied to occur. So we just didn’t let it work with that, because we don’t want players to waste their sulphur. But outside of that, it is somewhat similar to Hinekora’s Lock, at a much lower one-time fee. Per craft.
One of the things that isn’t gone into detail in the livestream too much is an important limiter on the system. You may have seen the text saying intangibility at the top of an item, and then a value. The value of intangibility on an item is essentially a percentage chance for all Allflame crafting steps on that item to fail to create additional outcomes. So you basically just burn your sulphur and get only one outcome instead of multiple. And only when you successfully do Allflame crafts on an item, it builds up more intangibility, which basically means the more you Allflame craft on one specific item, the less and less efficient Allflame crafting becomes on that specific item. To try and encourage players to use Allflame crafting for powerful crafting steps rather than every step of the process of making a piece of gear.
OUTPUT LAG: Depending on the craft, is it different how much that penalty is applied, or is it a single amount each time?
GGG: Every single currency in the game has an intangibility amount associated with it when you do an Allflame craft. Some of them have an amount that is zero. Some just don’t apply any intangibility at all, and that’s generally the ones you’ll be doing in the campaign. Transmuting or alching an item doesn’t apply any intangibility, because we do want players to engage with the system a lot while leveling. It’s mostly things you’d be doing on really endgame crafts that are going to hit this intangibility system.
OUTPUT LAG: Are all the endgame crafts eligible for it? I’m thinking Awakener’s Orb. Is pretty much everything possible, or are there caps on certain things?
GGG: Basically anything that can work does. Things that can’t work, to clarify, you can’t use a Mirror for instance, because that creates an additional copy of the item and there’s nowhere for the system to put that. So any output that would create an additional copy of the thing, that doesn’t work, because it functionally can’t. And then anything that’s trapped in another crafting system that isn’t a currency can’t work. You can’t do harvest or bench crafts through Allflame crafting, because that’s not an itemized currency that you can bring to the boat. That’s a different crafting system that’s kind of isolated. But if it’s an itemized currency that you can put into the crafting window and is valid for whatever equipment, well, not even just equipment, you could put a map in there if you wanted, then it should be able to work.
OUTPUT LAG: I love crafting leagues. I’m looking forward to it.
Luminary Ascendancy and the Return of Mercs
OUTPUT LAG: I’m really excited about the Luminary. As someone that’s been playing a long time, and I do love PoE 2, but PoE 1 is always going to be my go-to jam. The permanent mercs, do the passives unlock specific options to customize them? What does that look like?
GGG: All of their gear can be swapped in and out, assuming you’re a Luminary. What the passives allow you to do is equip them with uniques specifically. Baseline, they’re not able to wield unique items, but you can swap in rare or magic, I suppose normal if you wanted to, gear onto them without needing any investment beyond being a Luminary.
OUTPUT LAG: Is there anything that, no matter how you go in the tree, you’re never going to be able to equip onto them?
GGG: Not that I’m aware of. There’s certain gear slots they don’t have. You can’t put a flask onto a mercenary because they just can’t equip those. But any of the gear slots that they have, I can’t think of any other restrictions aside from needing to spend ascendancy points to be able to equip uniques in that slot.
OUTPUT LAG: So now that they don’t count as a party member, no life scaling, no rarity and quantity hit. Is there a limit with party play? If you have a full party and you’re each a Luminary, are you each going to have your permanently buffed super mercs with you?
GGG: Yeah, you could. As Mark likes to say, that would be a ruckus.
OUTPUT LAG: I can already imagine the streams and videos I’m going to see of people doing crazy stuff with that.
GGG: The reason we had to take away the party scaling thing is, effectively for each league we have a budget for the amount of items we tack on, so to speak, to the character power. So we had to get rid of the party scaling just because it’s too much extra quant. We can’t just power creep our item drops like that. And then for the power, putting it behind the ascendancy class means it comes out of the class’s power, which is the budget that gave us the capacity to add it there. I thought a lot of players would be a bit disappointed losing the party scaling thing, and I’m sure there will be some, but it kind of just has to happen. Otherwise we have to take it away from somewhere else, and I don’t think they’re going to like that. You also just can’t have an ascendancy that has more quantity and rarity than other ascendancies.
Abyss Rework
OUTPUT LAG: With the Abyss rework, is it the only source of abyss jewels and Stygian Vises now, the same way the fossil rework worked? Does that take them out of Delve cities and the Lich fight?
GGG: Delve cities is a good example. Those might still, just because they’re super thematically appropriate. You might still very rarely find an abyssal jewel chest in an abyssal Delve city. But basically everywhere else I think they’ve been stripped. So Tujen won’t offer them now. There’s not a random abyssal reward icon that can show up in Legions, or from Sentinels that you use, or probably dozens of other places that I’m forgetting.
I’m personally really excited for this, because I really liked farming Delve last league with how profitable fossils were. The baseline cost of just any fossil at all was like 10 to 12 chaos. So you could make insane amounts of profit. When you draw those distinct lines between different chunks of the game economically, you create places where players can specialize and feel rewarded for doing so. That’s an extremely good rewarding feeling. I want to do that in more places.
We have a goal in general, which is two things for each league. You see this with Legion, like incubators, we’re just giving other league loot, right? Lame. So the initial thing for Legion, we’re like, we’ll just remove them. And then it’s like, well, then it’s kind of got nothing. So let’s add something. Our goal for each league is each league has its own exclusive reward system. And our more aspirational goal, which is a bit harder, is every league has their own pinnacle boss. That is our two goals for every league that you can invest into on the Atlas, or not necessarily on the Atlas, ones you can farm at endgame, i.e. Delve, etc. They each have a pinnacle and they each have their own exclusive reward system. That is the goal, and we will get there eventually.
There’s a mysterious figure in the giant crystal in the Legion Domain of Timeless Conflict. Who could it be?
OUTPUT LAG: What happens to Darkness Enthroned? Is it still tied to the Lich fight, or is that getting moved too with the new Abyss being the source of the regular belts?
GGG: That will still come from the Lich fight. The pre-existing Lich fights are still in the game. All of their uniques are still on them. We’ve just added, I believe, seven new uniques that come from the new Abyss boss.
Talking My Favorite Mechanic: Blight
This started as a tangent and turned into the most interesting thing I got all call. For some reason I’m a tower defense guy, so Blight has always been my league, and I’ve watched it sit at the weaker end for years now.
OUTPUT LAG: I’m looking forward to hopefully eventually some sort of Blight rework. The one thing, and maybe this is a quality of life change, but if the towers eventually work while you’re off screen, that would be amazing. That’s my only frustration.
GGG: Also on that note, we do have the art complete for a Blight pinnacle boss. Now we just need to figure out when we’re going to have the time to animate it and do all the skill effects and everything. But that one exists artistically at least.
OUTPUT LAG: So it’ll eventually get in there somewhere.
GGG: We’re getting there. We’ve got all these things getting made somewhere. It’s just picking our battles.
Vestigial, and the Headhunter question
OUTPUT LAG: With the new enshrouding crystals, is there a line or a limitation? Can I enshroud a Headhunter and put the rare mod stealing on a body armour?
GGG: There’s two things in that question that are limitations. First off, the system does not work on jewelry slots. So Headhunter is just not an option. It works on body, glove, helmet, boots, and shield. Those are the valid options. And it only stays within item class. If you enshroud a body armour, you’ll get back a body armour. If you enshroud boots, you’ll get back boots.
OUTPUT LAG: That makes complete sense. I was thinking a little too grandiose there, but I had to double check.
GGG: The reason we limit the slots, by the way, is just so that we don’t occupy all of the design space in one single system. The capacity to do it on jewelry could be something that comes in later as a more uberfied version, or maybe it comes in in a different context in a different league. We do that kind of stuff at some point later. Not to say belts will necessarily get it, because we’ve got some kind of ridiculous belt, as everyone knows.
It’s a thing we learned from PoE 2 as well, that constraining the classes that things can go on allows us to then add those things later through more exclusive means, or different content, or different leagues. And if every single unique, let’s say every single best unique, ended up being a vestigial unique, as in every single slot ended up being that way, it creates this thing where if every single one is in Legion then it’s kind of like, well, why farm anything else? So having armours come from there and jewelry come from somewhere else is something that could happen. And that way it’s like, okay, I need my armours first, I’ll go do that, or for trade value maybe. And then if I want my jewelry, I’ll go to something else. We try and generally split things up like that more nowadays, where historically we wouldn’t have done that.
Anomalies
OUTPUT LAG: They ignore the Atlas tree. What sort of value can you expect out of them compared to a super juiced map? What’s the benefit of running them versus a fully invested map late in a league?
GGG: For sure, if you do a full scarab investment super juiced encounter map, you’ll get more value out of that than your average Atlas anomaly. But your average Atlas anomaly is also meant to take much less time, much less investment, much less effort. The Cadiro one, for instance, you just go in and give Cadiro some gold and you get an item. Even things like the random blueprint wing that can be generated, it’s just the one wing of the blueprint. So it’s meant to be, you’re in there for a couple of minutes and you get a cool blueprint reward and then you get out. Quick little adventures that you get to do between your normal Atlas content. And some of the value in that is that we get to give players opportunities to take little breaks from the really intense, high octane, high juice Atlas content that they might otherwise be doing, which then helps to prevent burning out.
The other benefit is more in the system itself. When you do a map, you can go in, you can leave, and then go back, as in just use portals. But there isn’t the concept of permanence on the Atlas. So Cadiro, for example, you can go there, have a look, leave, and then three days later come back. That’s not a thing that the Atlas has supported. Now it is the thing that we’re starting. As much as Cadiro is the basic example of the use case there, the idea that we can have things that appear on the Atlas that are like, go and do these 20 maps now, and then come back and do the thing, is a concept we have. A map is a short-term goal. It’s I do the map. And then you have your pinnacle bosses for longer term goals. These can provide some medium length goals to do in the future. So even if not all of them necessarily are achieving that now, depending on how we see how people resonate with it, it’s a system we can use for doing some pretty cool stuff in the future.
The other benefit which I really like is, I’m not the kind of person that’s going to, I’ll generally pick other content over doing Heist. Personal preference. And the idea that there’s just a blueprint wing that reveals and I can just go and do that, I’m kind of hoping it convinces me to maybe give it a go. And honestly, if it doesn’t, I’ll be like, all right, well, maybe we need to add that to the league rework list.
I like the idea that we get to take things that are normally not on the Atlas, a sanctum floor, a blueprint wing, and you get to do one of those and get a taste of it and go, oh, okay, yeah, I’m into that. I’m going to invest into that and run into that content now. Or maybe you’re like, next league I’m going to do that, because I actually quite enjoyed that. Giving players, especially less your veteran veterans, the opportunity to get exposed to some of these alternate forms of content, I think is the thing that could be cool for getting them to go, I’m going to give that a go now. I quite enjoyed that. Where they would never have discovered that otherwise. So hopefully that happens. But they also might try it and go, you know what, that reaffirms my hatred for it, and then never touch it again. And that’s also fine.
On niche content, and leagues that are better optional
GGG: Obviously we want everyone to enjoy it. I’ve talked about this before. There are some leagues that are just better when they’re optional. They are excellent when they’re optional and they are not as good when they’re not.
It’s very hard to make an entire league that isn’t the core league. That’s a huge amount of development time for us. The idea of making optional side content league style things that aren’t the core league isn’t really a thing that exists for us. I would love to be able to do it, but I just don’t think it will ever be possible. We’re not going to make a league that we believe is a 50-50 hit rate. Sure, sometimes it’s 70, we’re like, three quarters of players will like this. Sometimes we honestly just go in going, look, some people are just not going to like this league, it’s not for them. And that sucks, we know that’s the case, but we can’t take too many risks. And so what you then lose is the capacity to develop quite niche content, because you don’t want your core league release to just brick because it’s too niche.
So it’s an interesting thing where I would love to be able to create more niche content, but it’s a development resourcing issue. But we’re kind of doing it. The Atlas Anomalies allow an injection point for a little bit more niche stuff, because you could just not do it if you don’t want to.
The New Pacts Gems
OUTPUT LAG: With Pacts, the new exceptional skill gems, the video said they’re used to empower spells. Is that an actual tag or was that flavor? Can Pacts be used with triggers, totems, that sort of thing?
OCTAVIAN: No, they’re kind of meant to be a tool for self casting. It’s an archetype that we want to support more. It works similarly to warcries, if you’re familiar with that system. It gives you a number of empowerments that are spent on using the spell casts.
On Whether AI has found its way into GGG
I want to be upfront that I asked this for me, not for a headline. AI is being pushed into all indstries these days and everyone has unique feelings on the way it’s being pushed at a lot of places right now. GGG is the studio I respect most in this genre, in all of game dev really. I wanted to know if they were feeling the same thing. I told Mark he could kill the whole answer if he wanted. He didn’t. This is one thing I love about GGG. They’ve always been a great group of devs that are passionate about their game, well games now, the process, and what they’ve built. He was very transparent about everything but the tldr is: GGG does not use AI because it can’t compete with anything GGG already does and has mastered inside their dev team.
OUTPUT LAG: A lot of industries have started to use AI, and I know it’s a hot topic in gaming right now. Are you using it at all? And if not, do you see any future potential use for it, or do you feel like it’s never going to have a place in gaming?
MARK: We don’t currently use it. Short of someone going, what’s some names for zombies or something. And honestly, Google, just Google. It never gives great answers anyway, for some reason. Short of a bit of that here and there, generally speaking, it’s not used at all. At least I have no evidence of that.
Honestly, it really hasn’t come up at our company as a let’s set some rules on that. Generally no one’s really felt the desire to use it. We as a company, our leadership isn’t pushing it on anyone. We’re just kind of like, what we were doing works. So we don’t really have any reason to change it. From our perspective, we’re already just, content coming out left, right and center. We’re obviously doing something right. We don’t need to go and start compromising what we’re doing.
Being objective about whether or not the games industry will use it in the future, I mean, almost certainly. Already a bunch of companies are using it in various different ways. I don’t really pay attention to a lot of that, so I don’t know how much of it is to their success or detriment, to be honest. But it’s safe to say with how it’s going, it’s going to be integrated somehow into the future. Not necessarily for us, just in general. But yeah, we’re just not, ultimately we’re not going to rock the boat. We’re just going to keep going.
I am waiting for the day where it starts to come up and team members, my expectation is one day someone’s like, hey, am I allowed to do this? And then we have to go, okay, no one’s really even asked.
OUTPUT LAG: With you guys, you have such an awesome team already, I just don’t think it’s there yet where it’s actually going to be useful to you.
MARK: From my understanding, though. Again, I don’t pay a whole lot of attention to it, because I pretty much give every single bit of my attention span just to PoE. Even with all of our tools, by the way, I personally rarely look elsewhere to see if some other company’s doing it better or there’s some technology here or there. I get that our engine department will obviously look towards other engines and stuff like that. But I’m just like, from my understanding, and it mostly comes from us trying to recruit people, my understanding is we’re already working multitudes faster than every other company under the sun. Almost. So why should we go and adopt technologies that would slow you down?
It creates a little bit of dinosaurism, where yes, there’s probably some things we could do faster that we’re just not exposed to. But at the end of the day, if what we’re doing is ultimately to the better service of our players, which I believe we’re doing very well generally, we’re going to just continue to do that.
The vibe I get from recruiting people, because we often do things during our recruitment, let’s say we hire some artists. We’re like, how long would this take you? We just get them to say, here’s a task, we’ll give you this, how long will this take? And in our heads we’ve got, okay, our artists will do this in a week and a half, two weeks. Often we get back, oh, okay, so the first pass, eight weeks. And then after first review, another four weeks. And we’re like, what? You guys are just on another level. What are you talking about? How are you even working to achieve it to take that long? And then some people come in and they’re like, oh, four days. And we’re like, oh, okay, we’re in the right universe now. So it’s honestly a very good indicator.
So many people are just taking so long to produce things, and it could be the engine they’re working in, it could be that that’s the norm of the series of companies they worked at, I don’t know. But if we’re doing good, we’re doing fast by industry standard, why even necessarily go and change things up? Clearly it seems like people should be looking towards us, not the other way around.
END OF INTERVIEW
And I completely agree with Mark. As much as I love this industry and am proud to be a part of it, it’s very much in peril right now. The over bloated massive dev teams taking years between content is simply not sustainable anymore with ballooning dev costs especially for dev teams.
It’s a true testament of effeciency and expertise when you look at the GGG team and what they are able to do with both Path of Exile and Path of Exile 2. Every few months we get a massive content expansion for one of them, something unheard of in the industry. I think it would benefit a lot of other developers to try and explore how they are able to create these massive expansions so quickly.
Path of Exile Curse of the Allflame launches on July 24.
Thanks to Mark and Octavian for the time, and for a candid and spectacular interview.