Skip to main content

How do you make video game loot feel satisfying, but not too much? We asked the creators of Path Of Exile

"There's honestly a billion little considerations."

Artwork of Path Of Exile 2's Marauder class, who is standing on a pile of skulls and holding up their sword
Image credit: Grinding Gear Games

Legendary-tier monster cards on the table time: I do not like video game loot. I think that the popularity of "looting", an English word itself looted from Hindi during the time of the East India Company, is one of the worst aspects of the modern games industry and especially of the blockbuster live service game, which strives to keep its audience coming back by means of fresh loot injections at regular intervals.

I distrust how the randomisation element of much video game looting flirts with actual gambling mechanics. I hate that structuring games around the acquisition of loot creates a framework and an appetite for microtransactions and arguably, NFTs. But I am kind of fascinated by the art of designing loot, and especially when it comes to action RPGs such as Diablo 4 and Path Of Exile 2, because it seems to trade on some irresolvable contradictions.

A monk prepares to fight a large boss monster in Path Of Exile 2
Image credit: Grinding Gear Games

In games such as these, loot is the idealised commodity that promises the Earth but leaves you wanting more. It has to be at once desirable and disposable. Each piece of loot is essentially superfluous in that the act of looting itself is the real goal, but it seems to me that games can't ever fully 'admit' to this, because after all, people tend to find it reductive and dehumanising or simply tedious when designers treat them like, well, serial looters. Each Sword +2 and Lightning Shield you acquire has to offer a carefully calibrated measure of satisfaction, with handsome art and stats or powers that have a noticeable impact on your performance. It has to feel worth having if you're to feel incentivised to hunt for loot in general. But for the same reason, it can't ever feel like a keeper. The game's various tasks and challenges, from routine skirmishes to chapter-ending boss fights, are essentially a means of modulating that movement from the joy of ownership to dissatisfaction, rather than having any kind of substance in themselves.

That's my armchair-pretentious analysis of loot-driven RPGs based on my abiding dislike of them. You'd do better to talk to Jonathan Rogers, co-founder of Grinding Gear Games, whose 2013-launched flagship ARPG Path Of Exile is one of the most successful and, in Rogers's view, influential ever made. Having chatted to him about the forthcoming Path Of Exile 2 beta, I asked if Rogers had any thoughts about the act of creating "fun" or "meaningful" loot, a decade on from Path of Exile's launch. It turns out he had a few.


RPS: It was fascinating to hear you talk earlier about how Path Of Exile 2 handles loot and your desire to avoid "spamming the ground with so many items that it's ridiculous". One of the biggest RPGs released this year was Diablo 4, and it's been very interesting to follow the player discussion in that game, because it strikes me that ARPGs exist around a conceptual tug-of-war, where you're constantly hunting for the next piece of equipment, but if you make that too much the emphasis, people will think 'every individual thing in this game is like popcorn, it doesn't really mean anything, so why am I still playing?' Given that Path of Exile has been around since 2013, I'm really fascinated to hear how you've approach loot design in general.

Jonathan Rogers (JR): I guess there's a few things going on with regard to the loot in POE. One that I think is really important for satisfaction is the ability to trade. A lot of games have given up on the idea of player-to-player trade in a serious way. But things only really feel like they have value if you believe they could have value to someone else, in some sense. That's something that I think is really, really important, and a reason why I think that loot in POE feels more valuable than in a lot of games.

So that's like one big component. Another big component is it's very important that you never feel like the items that you have are completely compromised by future content that will be released. So this is the kind of problem you get more in MMOs like World Of Warcraft, for example, where an uncommon item in a new WOW expansion might be better than whatever crazy thing you ground for in the previous expansion. So we have to make sure that when we're releasing new content, there's new stuff you can do, but it isn't necessarily the case of that stuff completely destroys all the equipment that you have. A good item is still a good item, no matter how much time passes.

A Path of Exile 2 screenshot with a player firing flaming crossbow bolts in a collapsing stone city environment
A Path of Exile 2 screenshot with lightning bolts striking the earth around a wizard character in a dark environment.
Image credit: Grinding Gear Games

So that's another really important one. And I guess there's a few other points as well - like, you have to try and make sure the [basic modifiers] on items are not so complicated that people can't understand what they do. And you also have to make sure that the mods on items are high enough in the values that you actually feel them. Like, something a lot of action RPGs have gotten wrong in the past has been, oh, you'll add +2% to some random thing that you don't care about, that you don't feel or like it's just irrelevant. You kind of need the number to be like, +20% damage before you actually feel it. That's really important as well. So making sure the numbers are big enough on the mods that they actually have some kind of satisfaction.

Another one is 'mod dilution'. Like, if you have too many on your character, then that means that the effect of any given item can get too low to the point you can't actually feel the effect of it. There's honestly a billion little considerations that we've learned by analysis of other games as well as just stuff that we're going through with our game. It's a big area, but those are a few random details!

RPS: I'm picturing a monstrous spreadsheet where you have all this stuff mapped out, but I'm sure it's not just a spreadsheet. How do you track and visualise all those considerations around the value of loot during development?

JR: We use a few different tools. We do have models that we make in Excel. We also - and I actually need to break this out again, because it's been a while since we used it - but we have a simulator for loot drops. And we can query it, like we can effectively say, 'OK, if you do 100 maps, what are the results of that?' We can effectively use that to determine 'Does this seem right?'

Then there's just the kind of testing where it's like, you kill a boss just with cheats, and you say, 'Does the drop that I got here feel satisfying?' And you do that, like, 20 times and say, 'Does that feel satisfying every time? Did I sometimes get good stuff?' There's that kind of empirical testing. You kind of have to do a lot of different things, a lot of different ways. I haven't actually got the item simulator out since we started working on POE2 - we did that in the POE1 days, but I need to get that code working again. That's another thing we definitely need to do before we deploy. But yes, there's there's quite a lot of different tools like that that you can use.

RPS: Loot is integral to live service games with seasonal models that are trying to keep players coming back. Have you learned anything from other productions like that in the ten years since POE released?

JR: I mean, I would say that we're the ones who defined this sort of seasonal model, honestly. I feel like a lot of people are following us, not the other way round. We kind of pioneered the three months season model before I think a lot of other companies did. I would say a lot of that stuff sort of comes from us!

Loot is obviously very important. I think we are more at the forefront of that, than looking at what other people are doing so much. But I don't really think our model has actually changed a huge amount in this regard. One of the things that I think is really important is the full reset every season. I think a lot of games are very afraid to do that, because they worry that if you just reset someone's progress, they'll find that to be a negative thing. But I actually think it's one of the most important things for retention in the long term.

A boss called the Prisoner in Path of Exile 2, in the middle of a fiery melee attack on a player in a shadowy stone arena
Image credit: Grinding Gear Games

What we found when we've [released in different regions] is that you do suffer the first time you reset everyone completely to scratch, the first reset will have fewer users than the initial release. But over time, I think it's the only sustainable way to go. Because a lot of the time when you have these games that exist for a large amount of time, very few people are going to start playing a game that doesn't have the full resets now, because they feel like they're behind and they can never catch up. And so by resetting, there's always this expectation that, 'Oh, like if I come in now, I'm starting from scratch just like everyone - I'm sure they may have more knowledge than me, but we're starting in the same place every time.'

I think that's one of those things that ultimately leads to real longevity for a game like this. And there really aren't very many other games that have lasted as long as Path Of Exile has, and the funny thing is, is that Path Of Exile 1's largest launch, after 40 launches, was this year. We still broke our [current players] record this year, which is kind of crazy.

I would actually say that [resets] are one of those things that a lot of companies still don't understand and they often chicken out of, but I think it's very key.

RPS: Do attitudes towards the loot in Path of Exile vary much between players in different regions? Minecraft players in China and North America have very different expectations for the game, for example. Have you noticed anything similar in your community?

I honestly don't know that I've observed a subtle difference myself. I won't necessarily say that it's not different, but we honestly don't tend to break things off by region that much, when we're looking at things. Like I tend to look at the playerbase as a whole and not like segment them too much, which is probably something a lot of data scientists would say is terrible. But either way, I tend to just sort of see players as players in that regard.

Read this next