Path of Exile week continues here on IncGamers and we had the unique (no pun intended) opportunity to talk to Brother Laz who some Diablo gamers may also know as the creator of the hugely popular Median XL Mod. Brother Laz was also most recently hired by Grinding Gear Games to create their ARPG’s unique items.
In this interview Laz talks about his work on Path of Exile with Grinding Gear Games, his work on Median XL and what he thinks of Diablo 3.
IncGamers: You’re creating some of the unique items for Path of Exile. How did you get that job and what sort of responsibilities do you have working on the project?
Laz: It took talent, hard work, and saying “yes” to an email that arrived in my mailbox one day.
Basically there is a huge variety of base items in the game, and making unique versions of all of them is a process that most certainly takes longer than it looks, especially for a small studio like Grinding Gear. Diamond supporters can have their own uniques added to order, which helps alleviate some of the load, but there is still a metric ton of low and mid level itemisation that needs to be done and is particularly relevant for the levelling experience. This is where I come in.
IncGamers: When you joined the project, how did you get up to speed and gain the expertise required to make proper unique items? You had to learn the overall item strength and values, project what high end characters would need, etc.
Laz: Path of Exile’s character build options are very deep, which sets the game apart from most other ARPGs. Therefore it is important to stay on top of the popular builds and which builds are unpopular because they have no item support. As a relative outsider, character guides have proven to be very useful.
However, not every item should be designed towards a specific build – in fact most should not. Creating items that are mandatory for build X and useless for builds !X does not add value. It is better to make items with a variety of uses and letting the players figure out what to do with them.
IncGamers: How do you and the GGG devs interact and communicate and coordinate, in terms of keeping the uniques consistent in power and balanced compared to other items? Are there issues with the numerical values on items having to change in response to other game balance tweaks?
Laz: There is a process that may or may not include feedback from the Grinding Gear developers and their own balancing steps.
Items should not be balanced based on the rest of the content. Just because build X is weak does not mean you have to make an item specifically to buff it, because then you can’t buff the spells and passives that make up build X without making the item overpowered.
Creating items in the right ballpark is not too hard unless there is some unforeseen interaction or small print about an affix. Or perhaps the Grinding Gear people are sweating day and night trying to salvage my horribly imbalanced items and I just don’t know it. That’s also possible.
IncGamers: Are you assisting with larger issues like balancing the economy and managing item rewards and other gaming experiences?
Laz: Not yet, but who knows what the future brings…
IncGamers: Planned obsolescence for items is a big deal while leveling up characters. Developers want players to find items that feel awesome and can be used for a few levels, until something better is found. In Diablo III this system is inexorable, with item stats cranking up so that anything found at level 40 is underpowered by level 45 and useless by 50. This differs from the system in Diablo 2, where lots of low and mid level unique remained useful even in the end game. How do you approach that obsolescence balance in PoE, and how do features like item enchanting and enhancing and socketing factor into it?
Laz: D3 made the mistake of balancing items based on the WoW paradigm of linear progression, which works well in a game that is about taking one character and developing it for the next two years but not in a game where rerolling is a thing. It is absolutely necessary to have a chance to find something awesome at low levels, otherwise you feel like you are wasting your time. This something awesome tends to come in the form of uniques (or powerful binary affixes like +skill levels).
Awesome low level uniques may include Johnny uniques like Springleaf in PoE and Thinking Cap in D1. The idea is that these items are so unusual they cannot directly be compared to regular items and therefore do not get overshadowed at higher levels. They could also take the form of a “Spike Jr.” item that is intentionally designed to be overpowered when you find it, then fall off later. Everyone likes stomping faces for 30 minutes.
Very high level items should generally be more Spike and less Johnny. Those one in a million items should be useful for people who actually optimise their character for the purpose of farming one in a million items, while build enablers should be widely available at low levels so the player can actually implement that build before the earth gets hit by [DATA EXPUNGED].
IncGamers: PoE is running a promotion where fans can submit their own legendary item designs. Obviously you’ve got to coordinate those a bit, so some guy doesn’t put in a sword of +5000000000 cold damage, or whatever. How is that process handled? Have you seen any really interesting or creative fan design ideas that spurred your item designs in ways you hadn’t thought of previously?
Laz: The fan item design process is not something I am involved in, however the entry barrier for submitting items is high enough to encourage intelligent submissions and the items are looking good so far.
I did read some interesting player comments and added them to my pool of considerations, but I do not directly take item suggestions. For those who want their own item concept included in the game, I recommend purchasing the Diamond Pack which lets you do just that and also comes with a custom made forum avatar and a lot of other swag.
IncGamers: Your relevant experience was creating items for your for Diablo 2. Can you summarize your time working on that mod for people who know nothing about it? It wasn’t just the same game with some tweaked stats; you made mods that were almost total conversions with all new skills, classes, quest events, etc.
Laz: Yada yada self-important bragging blah best mod ever. Tldr: Diablo 2 partial conversion, 2003-2012, new skill trees for all, new items, new monsters including Ni’Va and barrel throwing monkeys, 30-something uberquests instead of Blizzard’s pathetic 2 uberquests, intended to be a more strategic experience at first but quickly evolved into total pandemonium, now the mod has black holes and Viz’Jaq-Taari nuclear calldowns and lets you turn into a giant quill rat and yet it’s still difficult and challenging and also still lore friendly. Also #3 Mod of the Year at ModDB at some point and was played by Blizzard and Riot Games people.
Download now, unless you are playing Path of Exile instead, in which case please keep playing Path of Exile. Latest official version is MXL2012 v005 which you can download at Modsbylaz, though a couple of fans have resumed development where I left off in early ’12 and their modified version is called Ultimative.
IncGamers: Obviously the items are a big aspect of that as well, and they were basically your resume for working on Path of Exile. Can you ballpark estimate how many items you created, including unique and sets and runewords, in the different versions of the mod over the years?
Laz: About 1,500 during the past 9 years, but not many of them are worth remembering. Gaze upon my works, ye mighty, and facepalm.
The difference with a monkey whacking on the keyboard is that I figured out with time why I failed at making good items in the past and what not to do next time. The Von Neumann-like spread of chance to cast items in MXL was a learning experience. I would do things very differently if I were to start over on MXL today, and Grinding Gear have given me this opportunity!
IncGamers: What sorts of lessons did you learn from that? In terms of balancing player wants with game needs and keeping enough reward for fun, item power so players could survive, etc? Do players know what’s best for them? Does everyone want super shiny items that would turn the game into a boring walkthrough if they actually found them?
Everyone wants to win. Everyone therefore wants the stuff that will help him win, and play the game in order to obtain this stuff. But when you actually give it to them they will eat it all and quit. You have to feed the players just enough to keep them addicted but not enough to make them sated.
Yes, kids, we think you are lab rats. Now please don’t complain and if you click hard enough on the boss we may give you a shiny food pellet. Gummy bear flavoured.
Luckily modern games are not just about the grind: some games are mainly about character building and theorycrafting (Path of Exile! Play now!) and others are about immersion and story (Skyrim, GW2) or pvp competition (LoL). Still others just ignore all of the above and therefore fall on their face (D3).
IncGamers: Any other general or larger observations and theories about how items work in an ARPG that fans might be surprised to contemplate?
Laz: When an ARPG features difficult content, this typically means you need to grind a ton before the game lets you win, or alternatively you need to make a very specific build or fail. Easy ARPGs have many viable builds and don’t require much grind, but making the best build for fast farming can still be a challenge.
Note that combat skill is not particularly relevant and challenges that require this kind of skill to overcome are seen as a problem (“remove invulnerable minions!”). These games are about a lot of things, not including mouse skills.
Because of the importance of rewarding players for making good choices, blindly striving for a high level of balance between items and builds is a bad idea because it makes build planning and item finding much less important. You need weak items and builds for people to feel good about having strong items and builds…
Diablo III and ARPG Design
IncGamers: Diablo 3’s item system was probably the most criticized feature in the game upon launch. The uniques (legendaries) especially were singled out for condemnation, and Blizzard obviously agrees, since they’ve revamped the item system in the last several patches, by improving the legendary items, giving them some mods that rares can’t get, and improving their drop chances. Those seem like very obvious fixes, ones that many players said needed to be made when the first legendary items were revealed, months before launch. Any idea why or how the Diablo III devs bungled the item system so badly upon launch, making what seem like such obvious oversights?
Laz: The D3 developers were heavily influenced by World of Warcraft and implemented a ton of MMO good practices that unfortunately fail in a Diablo game, notably the idea that items have to be balanced instead of fun and inspiring. Misguided attempts at “balance” were actually fairly common among noob D2 modmakers and inevitably led to boredom-induced disaster while I went the opposite direction. It did turn out that people seek out fun games, not balanced games.
Of course the whole auction house fiasco wasn’t helping. It basically means anyone has access to every item. The hassle of trading in D2 and crucially the fact that you needed good gear to trade for good gear made it somewhat less likely that you’d just buy everything off other players. And then the gold farmers came and it got worse.
Interestingly the D3 item design minimises the risk that people would just buy the best item off the auction house and ignore everything else, ensuring repeated purchases and therefore repeated cuts for the house. I won’t claim this is intentional, but you never know with Bobby “Washing Powder” Kotick.
IncGamers: Now that you’re working on a game yourself, do you see or appreciate difficulties that you never noticed from the outside?
Laz: Luckily, the boring stuff like paying salaries and debugging memory leaks in cursor code are handled at the other side of the globe, I’m just doing items here. I’m actually not used to having to ask for changes instead of just going into whatever.txt and doing it myself. Growing up sucks sometimes.
IncGamers: A commonly-debated issue is how powerful players feel. Jay Wilson often said during development that they wanted to make players feel “awesome” and that’s something other game developers have said. It’s a lot of fun for a player to find a new item that’s totally kick ass and even OP for the level. How did the Diablo III devs miss that boat so badly on their legendary item design, and do you calculate for it when making unique in PoE? Say by designing an item that’s good for level 25, and therefore awesome for level 20, and then giving it a chance to drop at level 17?
Laz: Intentionally overpowered low level items are a good thing: it feels good to find them and they don’t break balance. In fact, every non-endgame unique that isn’t a fun or niche item of some sort should ideally be overpowered at its level. What is the point of a balanced unique that doesn’t do anything cool either? Might as well grab any random regular item off the ground then and paint it gold.
Just gotta make sure not every niche is filled so twinks don’t walk around with a full set of uniques ignoring everything else.
I don’t know what the D3 developers were attempting to do, but they made legendaries useless and turned inferno difficulty into Super Meat Boy.
IncGamers: Back in early June, just a few weeks after Diablo III’s launch, you wrote a guest article for us, documenting the lameness of the Diablo III item system while classifying all players into three types; The Spike, the Johnny, and the Timmy. Do you still hold to those classifications and did they inform your item designs in PoE?
Laz: They indeed do; I have a little Excel sheet in which I tag every item with Spike, Johnny or Timmy.
Between that article and today I’ve grown to appreciate the Timmy side. They’re the Darius player, the three minute mage, the Third Street Saint, the battlecruiser guy and the Leeroy Jenkins. They tend to be more casual, but are also looking for straight up fun in their video game.
As far as items go, there are perhaps two other target groups: the altaholic who can’t stick with a character and only sees level 1-20 over and over; and the lorekeeper who simply collects cool and rare items, a target group that is not well served by the Diablo games but has found a home in a certain popular war themed hat simulator and GW2’s metric ton of armor dyes.
Blizzard famously decided not to allow any sort of player mod making for Diablo III. If Diablo III could be modded, would you have tried your hand at it? If so, what kind of changes would you have made, to the item system or the game in general?
Laz: There would be little point in modding a dead game, but the most important thing would be to Dota-ize it: give the items, spells and enemies more pronounced strong and weak points. This should enable players to actually use tactics to beat the game (as opposed to “more damage”) and make the game more exciting.
This does include powerful and interesting uniques, monsters that are dangerous when fought on their terms but have weak points in their tactics that can be exploited for massive damage, and actual reasons to use different skills on different monsters. Single target skills should destroy single targets, but swarm type monsters should come at you in packs of 20 instead of 5. Etc.
The biggest issue with D3 after all is that it is not exciting in any way.
IncGamers: Another big complaint about Diablo III’s legendaries in the early going was how bland they were. Aside from having weak stats, they had few stats that couldn’t be found on rares, which made legendaries seem not-special. Have you paid special attention to giving the unique in PoE “flavor” as well as stats big enough to be eye-catching?
Laz: Flavour is very important. Uniques that are not special in any way are pointless. If you don’t want an item to have special stats, at least throw in some flavour stats like a small negative affix or a chance to cast gimmick to make it worth identifying.
Big stats are eye catching, but only once per item type. There can be only one Blightsteel Colossus. The vast majority of the items need to be carried by being special, not just huge.
IncGamers: Most players prefer Diablo II’s item system to Diablo III’s, in part since in D2 the loot drops so much more frequently, especially the uniques and sets. That’s often thought to be largely due to Diablo III’s closed, persistent, balanced economy — in D2 items could drop in huge heaps since players rerolled all the time and the entire B.net realms were regularly wiped with new seasons, so it didn’t matter if characters quickly grew very powerful and inflation kicked in. That can’t happen in Diablo III since the economy is meant to last for years without resets, so items have to drop less often lest they flood the economy. You didn’t need to worry about a closed economy in Median, but PoE is aiming to build a long term stable economy on secure servers, so how do you balance the issues of item spam and inflation with player desires to find for awesome loot?
*untaps Text Wall*
Laz: D2 loot drops only appear frequent because everyone is using duped items and because ladder seasons take many months. Compared to other games it still takes ages to find high end items. Granted in D3 you have zero chance to find anything decent so I guess the faint hope of seeing a Pul someday is an improvement. Perhaps D3 was designed for today’s economy where no one has a job.
Whether or not ladder seasons are necessary depends on the game. An online action RPG does not need ladder seasons to remove items from the economy as long as the grind/reward curve stretches to infinity. If there are items that are almost impossible to find, treasure hunters will keep farming for them until the sun explodes as long as you throw them a carrot every once in a while to keep them addicted. And people who care mostly about interesting builds are more than capable of entertaining themselves without having their personalised characters wiped every X months. A ladder reset is a good marketing move though and tends to result in a huge inflow of returning players. This is why League of Legends has ladder seasons.
When an economy gets “flooded”, this is because good items are too easy to get while the best items are too hard, making players feel like they cannot realistically improve their characters or it would not be worth the bother. This indicates a drop rate issue. Gambling machine designers invested a lot of money into research like this and action RPG developers would do well to take their findings into account.
Once you have things set up so that adequate gear is common and there is a power curve upwards that is inversely correlated with drop rates, the only remaining problem is that the game will gradually get easier as more good items find their way into the economy, and you can’t really make the monsters harder or newbies will get owned. Either you go with ladder seasons for the sole reason of fixing this issue, or alternatively you could keep throwing in new and harder high level areas and treat the large number of good items as a form of rubberbanding for new players. This is where games like PoE with their randomised dungeons luck out.
IncGamers: What will a Diablo player think when he first tries PoE and soaks up some of the item system? Familiar and yet different?
Laz: There was a time when people were willing to put up with a game that has more bugs than a 200/200 ling army, broken items, endemic duping, potion spam, PvP consisting of assholes going hostile on you and using glitches to instakill you, imbalanced skills etc.
PoE is intended to be a spiritual sequel to D2, not a “Diablo clone”. It is similar to the rose coloured memories people have about D2, but not at all to the depressingly bad reality of D2.
IncGamers: Have you guys planned out items over the long term, with ideas for new properties, item types, bigger stats, etc, to be added in patches or game expansions?
Laz: More items will be added continuously after release. Other plans include [CLASSIFIED] and the awesome [CLASSIFIED]!