Itemization design in D2: what’s the right place for Runewords?

Now that Pandora’s box is open, with talks about balancing and even new Runewords, I think it’s a good time for everyone, including my purist ass, to talk about what aspects can objectively be improved in Diablo‘s sublime yet imperfect design. With the benefit of 20+ years of reflection & hindsight about the game and some game design principles, I’m gonna talk here about Runewords in particular, and what place they should have in the game with a stated goal of making the game as a whole deeper and more interesting. I would love to talk about other item classes another time, as there are also things to say about them and how to make them fit better into the game (mainly set and crafted items).

What do I mean by “depth“? It can be shortly defined as having a high number of viable ways to play the game. Which is a good enough explanation for what we’re doing here.

Runewords in Diablo 2: what makes them a problem

Diablo 2 is perhaps one of the best and earliest examples of how to have amazing itemisation, and what that brings to your game. D2’s itemisation both provides motivation for players to go forward and continue playing even after they beat the game, but also, it’s what allows multiple viable builds to exist for every class, essentially creating what we defined as “depth“.

In Diablo 2, itemisation has multiple types (item “classes”) that each obey their own rules. What’s incredible is that, because these item types compete with each other in a way, it was crucial that no item class is clearly superior to others, and that there is room for all of them to shine given the right circumstances, otherwise, players would never use other classes and just concentrated on the best ones which would make the game less interesting. The problem with Runewords actually starts here, as some of the Runewords items are often best-in-slot, but, crucially, this is combined to other fundamental properties of Runewords which ends up making them objectively OP in the game.

Runewords properties that cause issues

The main things that you have to consider with Runewords, that sets them apart from other item classes as well as creating a problem are the following:

  • Runewords items are craftable
  • …with components you accumulate (runes) that are quasi-currencies with relatively fixed exchange rates, which makes them obtainable in a relatively straightforward way (grind)
  • …Runewords have relatively low variation when crafted (some stat variation, but you roughly know what you get), contrary for example to Crafted items
  • some Runewords are unequivocally best-in-slot items in the game for certain builds (Grief, Enigma etc..)

And it’s the combination of these points that are specifically what set Runewords apart from other classes, and again, what creates the problems with them in the game

To help illustrate the problem, let’s first look at a crude but effective graph I shamelessly stole from a MrLLama stream that basically illustrates the issue with Runewords:

Graph shows relative power/utility vs stat spawn range for all the item classes (i.e. Magic class mostly spawns weak items, but the rare best ones are better than most best ones in other classes)

As you can see, itemisation in Diablo 2 is quite unique in that all classes have different rules as to how they’re created or found and what effects they have.
As a reminder: Magic has 2 effects with a large range of values, Rare has up to 6 effects with a smaller range of values, Set items are authored with synergies as you complete the set and low range of variation, Uniques are authored with often strong effects and low variation, Runewords have authored effects that you can apply to an item you craft and Crafted are obtained from recipes and generated with a very large range of values

The issue with Runewords starts with the fact that some of them are undisputed best-in-slot items for some builds. However, the problem is not that they’re just best-in-slot, as that would just make them valuable and rare (like Unique items or some Magic ones), but it’s that this is combined with the fact that they’re craftable, which means you can obtain them directly, AND that the components to craft them, mainly runes, are a form of quasi-currency that can be easily exchanged in the player market for some other valuable quasi-currencies, similarly to how bank notes have different values but you can exchange all of them for the ones you want and at the end of the day only the “total value” of your bank notes matters, more than the individual bank notes, which makes obtaining these a question of time rather than luck or skill.

The net result of which is that players know that obtaining Runewords is just a question of pure time spent (“grind”), they will eventually have a total Rune value that will allow them to exchange them for the right set of Runes, and just craft the best item in the game.This in turn means that builds ultimately require no adaptation: you just need to aim for a build and you’ll just go for it regardless of the items you find along the way.This in turn affects Depth, and therefore, player choice, autonomy (“freedom”), build diversity, uncertainty.

The other way to look at it is that Runewords being top tier and craftable essentially devalues the whole economy, including Unique items and Sets, especially in the end game, except the components required to make the Runewords which are the Runes themselves and their base items gets devalues.

In one sentence:
When there are best-in-slot craftable (Runewords) items available for a grindable currency in the game the only viable way to play the game becomes to plan to make your build around these items, and this leads to character builds that are determined not by player skill, or planning, or adaptation, but instead, that are made by following pre-made internet guides, thus reducing all viable and competitive builds down into a much smaller number of builds, which hurts the game overall by making it predictable, boring, repetitive and shallow.

Difference between Runewords and other item classes

This issue makes Runewords completely different from other item classes (Magic, Rare, Set, Unique) for one specific reason: for all these items, obtaining them depends not on grind but instead on stumbling upon it during a playthrough (“RNG”). The unpredictable nature of finding these items leads to players having to adapt their playstyle to the contextof their playthrough, by modifying how they play, what item they use, planning, tactics etc.This also means, even if there was one Unique item that was way above others in terms of Power or Utility, it would still require you a large amount of luck to be able to obtain it, and it would be much harder to simply follow an internet guide to make your build.Which, while it would somewhat put success out of the player’s control, it would offer much more dramatic and “Roguelike” (which D2 is inspired from) games, and would require much more thinking and adaptation from the player

Counter-argument: what’s the difference with Uniques?

One counter-argument might be: what’s the difference between stumbling upon items (i.e. finding Uniques) vs stumbling upon runes (i.e. Runewords)? In both cases you rely on luck to find your gear!

The crucial difference that has to do with the “quasi-currency” nature of runes. Runes have properties that make them much more like a currency compared to all the other “asset classes” in D2, including Unique items. Ultimately, the only thing that matters with Runes is the total “rune value” you have, rather than the individual runes you have, as you will be able to exchange your way to the desired combination of individual runes. This is a bit like having banknotes in your wallet: what matters is not the individual banknotes but the total wallet value.

This in turn means that with Runes, obtaining the desired Runewords is only a question of time and grind efficiency

Whereas, with items, this is much less the case, because it will depend on various factors such as someone finding the item, then having to store it (itself an issue with limited space) and finally finding someone who wants to trade it

The net result is that Runewords tend to allow people to go for a specific builds, whereas having a unique-item-based economy leads to more adaptation to what you find, and even more drama, emergent player-driven storytelling etc, which is all good for the player experience

Just for fun let’s compare Runes vs Items on the criteria that qualifies something as currency (let’s also add gems and gold there):
Properties of a currency:

  • Divisible
  • Fungible: units are interchangeable
  • Portable: in D2 this is probably how much space it takes in inventory
  • Durable [doesn’t apply]
  • Acceptable: by all traders in a market
  • Uniform: units all have same value
  • Limited in supply
Properties Runes Items Gems Gold
Divisible High None High High
Fungible (interchangeable) High Medium Medium High
Portable (space) High Low High High
Acceptable High Medium Medium High
Uniform (same value) High Medium Medium High
Limited supply High High Medium/Low Low

We can clearly see why Runes are used as a currency in the game. We could also add “utility” as a property to compare these 4 asset categories to determine the demand for them and potential to be the main medium for exchange

“Active vs Reactive” progression, and the Paradox of Heuristics without Search

As a bigger picture issue for designers, this is related to progression systems, and the problem of “Active vs Reactive” progression systems in games.

When given the choice, players will often optimize and minmax (“playing” is often just exploring the game space to find the optimal paths, hence “minmaxing”), to the point where having more choices without requiring adaptation means not having a choice as you will be incentivized into following the minmax path.

This in practice means that giving choices without needing adaptation will make players minmax the fun out of the game. They will think they have choice, but won’t have any, and on top of that will have that dreaded “FOMO” (fear of missing out) feeling that if they don’t follow the optimal path they have done something wrong.

A shorthand way to prove this is the existence of internet guides on “how to play”. Whenever you see these, it’s a proof that someone has solved how you should play and the only thing left for you to do is to follow the guide, with no need to adapt your playstyle. The harder it is to make a game guide, the more chances your game is deep.

Countering this requires tying choices to a “context”, hence requiring the player to adapt. When the build choice you make depends on how the game goes, rather than being done abstractly, then the choices are more interesting. In Diablo, this is done through the items you find in your playthroughs, the variable enemy encounters, and the variable player encounters you make.

If you want to go further down the rabbit hole, you can check out the definition of “depth” in game design, and how the ultimate definition of depth involves “semi-orderedness”, in other words, a good measure of heuristics (skill, knowledge) but also enough room for “search” (adaptation)

How to fix this and make Runewords great?

There are a couple of things to consider when fixing Runeword items:

  • Their utility
  • Their power
    • Especially in comparison to the power of Unique items (authored items in general, which includes Set items too)


The first principle is that, in terms of power, Runewords should never EVER be best-in-slot, compared to all other items in the game, especially Unique items.

This is because otherwise Uniques will become useless as soon as you obtain your targeted Runeword item, which removes all thinking (except build planning, which is largely done by online communities), adaptation, uncertainty etc.

This applies for example to Grief: the weapon is so damn strong, that often, there is no alternative to having a Grief, and it’s such a no brainer that there is no build diversity possible to play around it. Grief basically kills Depth and Variety in D2

Ideally, the usefulness of Runewords items, even when they’re weaker than a best-in-slot Unique or Set item, should come from the fact that the Runeword item can be obtained much more easily by crafting it from runes. It’s more easily accessible than specific Uniques.

Which would mean, if their power level is well balanced so that unique items are still better, that power-focused Runeword items would have their utility as slightly weaker versions of unique items that have the benefit of being craftable

Balance-wise, I would make Runeword items at best as good as a low or medium rolled Unique item


The second principle is that, the more utility a Runeword item has, the lower the power of that item must be, otherwise the item will simply be too competitive and always be best-in-slot

This applies perfectly to Enigma for example. As the utility for it, teleportation, is so high, in combination with the nice stats, that this armour Runeword is one of the highest sought items in the whole game, and often there is no good alternative to it. Not to mention the positive feedback this generates (“snowballing”). Enigma too lowers Depth in D2

Utility is the second and best use for Runeword items: offering alternative and viable ways to play your characters and creating new builds through it

A great example is actually also Enigma, which, if it was not also quite powerful stat-wise, would be a good example of this: teleportation changes so much the game that even if the armour stats and effect benefits of Enigma were low, it would still be worth it to have it, though it would involve more risk-reward, more thinking and planning. In short, it would require more player involvement, skill, knowledge and make the game more fun and less straightforward. An Enigma with low stats would increase Depth in D2.

Bad design patterns for new runewords in D2R:

Given everything we talked about, and, as I currently don’t know what the new designs for D2R Runewords will be, here are what I would consider bad designs for the new Runewords:

  • OP, or too powerful Runewords: this would lead to what is called “Power Creep“, old Runewords would stopped being used, new ones would become ubiquitous, and the whole item economy would be devalued into oblivion until nothing has any value left
  • Bad power-to-rune-rarity ratio Runewords: while having OP Runewords is bad, having the opposite, weak Runewords, is sad, as these will, like all the never used Runewords, and won’t bring anything into the game as no one will use them
  • No “utility” Runewords: introducing new Runewords is exciting as an idea. However, if there are no Runewords that fundamentally changes how classes can be played, they might as well not exist

Good design patterns for new D2R runewords:

  • Power-focused Runewords that are well balanced: as I said, I think the best place for them is to design them so that in the best case, the best Runeword (with the best roll) is as good as a medium rolled top 20% Unique item
    • This would allow players to have a useful item that they can relatively easily obtain, without breaking the game and devaluing Unique items
    • Ideally, all item types should potentially be able to spawn best-in-slot weapons, though with different odds (even Magic and Rare items)
  • For non-top tier Runewords, then balancing the cost is important: making it affordable, so that players can use them early on as a good alternative to Uniques they can find, but not so powerful that this becomes the only alternative
  • Utility-based: items with good and interesting Utility, such as unique skills, stats etc, and a correspondingly lowered bunch of effects and stats, even perhaps negative stats, for that good old feeling of risk-reward
    • Maybe even making an extremely high utility Runeword that also has severe drawbacks and pushes you to potentially take risks and die (lowered Res/Health? changed behaviour?)
    • E.g: item that makes you invisible from everyone, but has garbage stats, can be used to stalk on players, escape tricky situations, get past mobs, but don’t get caught in combat with it!
  • If there is going to be OP runewords, there is a solution to balance it out somewhat: restrict it to Hardcore mode. Indeed Hardcore mode is the only mode that has a natural “sink“, in other words, where value is destroyed in the game economy, which is useful to balance item values
    • And, to be fair, the existence of shared loot stashes counter-balances that anyway and makes Hardcore character loss much less “hardcore”
    • Essentially, shared loot stashes act as “permanent progression“, similarly to Roguelike and Roguelite games, such as Hades or Dead Cells