Which, you know, is the whole point I guess.

Anthem is gone now, but that post series really focused on how bad of a job that game did in this regard. Or perhaps, how utterly complicated the entire thing can be. The design challenge here is multi-faceted, with some core principles.

• The itemization journey must last a long time to keep engagement up.
  • You don’t want players being fully kitted in a single session.
• Itemization should be constructed in a way that players can understand it
  • Meaningless stats are confusing
• Itemization must have incremental benefits that players can feel
  • +1 stat boosts are not meaningful and dis-engage players
• Methods to acquire items must be varied
  • The D2 runs of a single boss room for 40 hours are long-gone
• Itemization progression should have safety nets
  • Pure randomness is not good. Weighted drops (e.g. wizards don’t get bows) and enchanting (re-rolling stats) are good examples.
• Itemization must complement a player build
  • Generic stats are fine, but stats that are more tailored to a build give variety. +fire damage is better than +intelligence if you’re a fire wizard

This whole mess gets complicated quickly when you realize that some principles actively constrain others. If you add a few stats, players can quickly figure it out, but it also shortens the journey. Add too many stats and there’s no real sense of progression as you spend more time in the inventory than in the game.

Safety Nets + Player Empowerment

This is the relatively recent development in ARPGs, and it’s a relatively fascinating one at that. Other games did it first, but Diablo 3 really did a job here when it came to bloodshard gambling. The principle is somewhat straightforward.

• Players do content, collect items and currency for safety net.
• Currency is used to rapidly pull the 1-armed bandit loot machine to acquire more items

Diablo 3’s also implemented stat re-rolls, where you could select 1 particular stat and attempt to get another one. Re-rolling cold resistance for + hit points, for example. These re-rolls cost more and more resources, and the pool of available stats had different weights (+hit points is much more common that +all resists). This safety net meant that rather than look for ‘god rolls’, you could get something that was close and tweak it to be closer to top tier. Math is a big deal here, so let’s look at that.

Let’s say you have an item that can roll 4 different types of stats, of which the list has 25 possible items. For simplicity, let’s say they are all even odds to acquire. The math to get the ‘perfect’ set is 25x24x23x22 or 1 in 303,600. If you can re-roll one of those stats, the odds drop to 1 in 13,800. That’s a ridiculous amount of difference.

It’s important to note that safety nets + player empowerment actively work against a long journey for itemization, as it accelerates the time to get good gear. Balance is quite tricky.

Weighted Rolls & Spans

This was my largest gripe with Anthem, and at the core is due to the lack of understanding of player engagement. There’s a limit to how often a person will gamble without reward, and this is offset by the perception of progress.

Let’s say you’re a wizard and you run a dungeon 20 times. All the gear that drops is either outright for another class (bows and swords), or has stats that don’t help you (strength). So not only were you ‘unlucky’ in not getting good stats, you effectively were playing the wrong class all along. Engagement tanks here.

Alternatively, let’s say you do get a drop of an item that you can use. You have an rare version, with +100 to a core stat. A legendary item has dropped (yay!) and you look at the stats and it gives +80 to the core stat. As a player, you question your sanity that a less common item is somehow worse than your common one.

Weighted rolls ensure that on average, an item that drops is useful to you in some sense. Items that drop are both things you can use but also have stat rolls that benefit your class. It is VERY obvious when games don’t do this. This may seem an obvious thing today, but it most certainly was not even 5 years ago

Spans limit the high and low portions of a given stat roll for an item class so that on average, the more rare and item, the better the stats. This means that when you see a shiny sparkly item, it will nearly always be better than a more common alternative. This is just basic design, when Anthem messed this up it was a simple matter of poor design choices with too little time.

Diablo 4 vs PoE 2

In terms of itemization, there’s some parts here that are similar. PoE 1 and Diablo 3 are more like stepping stones and not really too much of the conversation here, Diablo 4 and PoE 2 are the next iteration.

Diablo 4 took what was there before and adding more knobs for players to tweak in order to get the right item combos. There’s a ton of safety nets, multiple valid paths to acquire items, re-rolls, tempering, gems, and a pile of options to tweak gear to fit a build. I’d actually say there’s too much here as 4 options to tweak an item is 4 ways to mess it up. The good (and impressive) news is that all the systems intersect with each other and are generally well balanced. Clearly a lot of experience and thought here.

PoE2, there are issues. There are some safety nets, but relatively few here. As an archer, you will find an abundance of melee weapons you will never use. The stat rolls are generally in line with expectations. You can’t re-roll gear in any meaningful sense as you can in D4, but there are some tweaks possible. The core issue is the way stats on items are distributed and then how those stats impact actual gameplay. The physical damage on your weapon is scaled across every other skill you have, making it the ultimate stat for character power, absolutely dwarfing everything else. Secondly, some stats are so important that even though they have no real value offensively, they are the only way to actually survive (temporary HP for instance). The good news here is that gems are no longer socketed into gear, which is a major positive output.

The net result here is that D4 has little friction in itemization and a feeling of continual progress, while PoE2 has significant friction causing fits and spurts of progress. Either you’re god-mode, or you need to dedicate time to farm new gear to get to the next challenge. Conversely, this means that D4 has you reaching the ‘end’ much faster than in PoE2. I’d guess you reach the end of the D4 journey before you’d even complete the entire PoE2 campaign. This generally means PoE2’s journey is more complex and fulfilling and much more painful to restart when a season starts.