Title 
What makes a good itemization good ?, what makes a bad itemization bad ?, how to recognize and pursue ?
Just a few questions that I’ll try to answer (and yes you can reply all you want cause it’s a discussion)
I’ll try to point out some criterias (keep in mind, these are not set in stone) based on past experiences and comparisons
Before going down deeper further I want to point out some “trends” and/or maybe myths that I kinda want to address:
Criteria #1 => itemization is required to 'support' builds
- D1 => 75% about combat, 25% about character stats (nothing about builds or even close something resembling a build)
- D2 => 75% of what D1 did but also added ~20% “secondary combat” and around 5% for “secondary buffs” and that’s all (maybe 1% build-first approach with a few uniques ?, but I think it was just a matter of identity rather than deliberate approach)
- GD => Haven’t played in a while but I think it was around 60% about combat, 10% about character, 10-20% about “secondary upgrade” of character, and 10% “identity”
- NOTE: “Identity” means => you equip what you clean, cleared chaos dungeon ?, chances are your strongest item is now chaos gear, cleared aether dungeon ?, you’ve got some really good pieces of aether gear, e.t.c.)
LE => 60% about combat (both defensive and offensive affixes), 30% about character stats, and only 10% about build (yes, 10%, I don’t think I’ve ever seen an item in LE that buffs a passive on a skill tree for example)
PoE2 => 50% about combat, 40% about character, 10% about secondary character (same thing, not build-oriented even here)
So the question is:
- Is D4 trying too much to read into “tea leaves” and go with “build first” when it comes to gear approach ?
Criteria #2 => itemization needs lots of affixes to increase flexibility of use
Again, let’s go through the games one after another
D1 => has up to several common/impactful offensive affixes, about half of that in defensive ones, and about 2-3 more
D2 => has up to dozen common/impactful offensive affixes, about several defensive ones, about 3-4 secondary/utility affixes, but what it had was “uncommon” and impactful inherent rolls
D3 => haven’t played in a long time but I’d imagine has like 10 common character and combat affixes that everyone wants (Crit trifecta, CDR, Skill empower, main stat, HP), and about several extra defensive ones too
LE => 6-7 offensive combat affixes, about almost as much defensive combat ones, and about the “usual” amount of defensive affixes too (in fact even less defensive affixes than D2)
GD => all that LE does but without defensive combat affixes and about a dozen “damage converts”
PoE2 => frankly IDK how many there are but the fact that there seem to be 50 of them (at least to me) none of them seem impactful
D4 => about 10 common ones and about 100 to pick from by chance % of RNG
Criteria #3 => itemization has to have high clarity of gauging individual affix impact
D1 => super high clarity, chance for backfire i.e. roll a negative value
D2 => super high clarity, one of the strongest “traits” of D2 itemization IMO
D3 => partially high clarity, i.e. if serves DPS or build then it has clarity otherwise it’s vague clutter
GD => “identity” provides most of the item’s clarity, but other than that feels like there’s a lot of “try out and find out yourself” types of stuff
LE => relatively high clarity except when it comes to mechanics that require more than one component to work (like Crit % for example, can increase baseline %, can increase chance multiplier %, maybe could also increase Crit damage but not sure)
D4 => we’ll have to go with two-phased approach here cause loot 1.0 is not the same as loot 2.0… Loot 1.0 was super vague and was very hard to ‘gauge’ impact of an individual affix, Loot 2.0 is much clearer in that regard but feels like still not that impactful cause there are too much “complementary” factors
Criteria #4 => rarity carries weight and matters a lot
D1 => has several uniques in the game (in total), a lot of them felt like a big upgrade but also a big commitment in return
D2 => has uniques and sets, and not only that but has 3 ‘tiers’ of items. But doesn’t seem like they carried the weight of itemization system itself on their own
D3 => ayy yaayy yaaay, IDK, I just never want to think about this tbh… It’s about maximizing several affixes, several pieces of gear, and then double-maximizing it all… In other words => rarity was not only carrying the itemization but was extremely super mandatory by default
GD => frankly I’ve completely forgotten… I know there were the white-blue items that were supposed to be the ‘rare’ ones and maybe complement each other but not sure. Some of were entirely character-changing with a higher downside of commitment… As far as I’m concerned yes, there were some rare items but I don’t think they ‘carried the weight’ in character building but not sure
LE => I don’t feel uniques and rarity carry the weight here either… I mean I know that set items were just weird and not with that big of a use. What I found for LE to be the case was exalted items i.e. the ‘deeper blue’ items with a double/tripple crit on one single affix tended to carry the “weight” for me the most
D4 => Uniques carry A LOT of weight in character build-up contribution. Some of them are even build-founding like TR, but in general not to the point that D3 did
PoE2 => I mean ‘meh’, so far got like ~30 uniques and none of them felt ‘whoa’ to me. Mainly because the mantra they seemed to use is 3-4 really safe and ‘global buff’ affixes with one very unique kiss-curse which felt like the point was not the kiss mechanic but find ways to battle the curse
Now let’s do a “vague” grading in terms of actual implementation of criteria mentioned for each game involved:
- D1 => C1 => 0/10, C2 => 3/10, C3 => 10/10, C4 => 6*?/10
- D2 => C1 => 2*?/10, C2 => 4/10, C3 => 8.5*?/10, 3/10
- D3 => C1 => 9/10, C2 => 6/10, C3 => 3.5*?/10, 10/10
- GD => C1 => 3.5*?/10, C2 => 10/10, C3 => 0*?/10, C4 => 2/10
- LE => C1 => 5*?/10, C2 => 8/10, C3 => 4*?/10, 6.5*?/10
- PoE2 => C1 => 1*?/10, C2 => 10/10, C3 => 4*?/10, 4/10
- D4 => 7.5*?/10, C2 => 2.5/10, C3 => 5/10, 8/10
Now for the “bias” verdict I have, I prefer itemizations of those games in the following order:
D2 > LE > GD > D4 > D3 > PoE2 > D1
Whoops
I mean I know, it is a bit of reach but PoE2 itemization didn’t hit any special strings to me: if you get a hammer most important is the damage and stun buildup, if you get a spear most important is damage and AS, e.t.c… In other words most of the time you can just ignore the “affix soup” below
What this means (at least to me) is the following:
C3 > !C2 > C4 ~~ C1
i.e. in plain english
- Clarity of individual affix impact > Non-clutter > Build-orienting ~~ Rarity
That’s the “result” so to speak as far as I’m concerned
Why do D2 and LE take the “cake” the most ?
They’re intuitive, immediately able to gauge impact of use, being a part of a greater system to immediately recognize if good or not… Rarity and Build-orientation aren’t that much important
In fact D2 has the biggest “non clutter” in all games except D1, one of the most-definite features in D2 is that you never have “granularity” at all:
- 25/50/75% Poison reduction
- Half/Immune freeze
- Ignore target defenses
- Prevent monsters heal
e.t.c.
LE has something similar in its “inherent” rolls
- Chance to crit/bleed/shock/burn
As very clear and obvious offensive tools of combat as well as:
- Chance to knockback/stun/apply-frailty/lifesteal
On the defensive end of things
The moment you get the item you can immediately “gauge” its impact just by looking at the inherent value (not just weapons but jewelry and gloves also)
Armor on the other hand is harder to “gauge” in either of these 2 (in fact D2 has/had most useless armor in all of them), PoE2 for example has an entire encyclopedia based only around armor you equip (but does it really matter that much ? 
you can get away with 100 armor if have high regen in that game IMO) 
Now let’s get back to answering the initial question/s:
- What makes a good itemization good ?
-
- Very clear and easy to gauge impact of individual affix rolls
- What makes a bad itemization bad ?
-
- Abscence of the first one (hard to gauge impact, too much “puzzling” and combinatory logic on gear alone)
As for the 3rd question itself:
- How to recognize and pursue ?
I’ll just say what D2 did better than every other game (even till this day)
- It “grouped” items by type in a system that they belong in, and each individual system that those items belonged in can tell the player “this is bad” or “this is very good”
What do I mean by this ?
- White items were either socketable for runewords or had quality increase for Charsi Imbue (that’s it, see an item with superior quality of 10% ? => Imbue, see a “crude” item with -5% quality ? => leave on the ground)
- Blue items had a self-gauging of their own, 1/3 ?, bad, 3/3 => good
- Rare items were in general the “common” ones, and this is where players decided what they wanted to “focus their strenghts” on for their character (cast rate, knockback, lifesteal, ignore armor, e.t.c.)
- Unique items were… well, unique, most likely unless exactly what you use not gonna use but IF YOU WANT to make it work can “milk” that uniqueness with an adjustment/sacrifice or two
So there it is folks:
The REASON why I did this thread is I want to kinda “omit” reading too much through the “tea leaves” and get to the brunt of the matter
Yes, games have evolved, a lot of games were tried out, but I still think that the criteria D2 excelled at the most (each item belongs into an entirely separate system of gauging values, and super high clarity of what affix does on its own) are still the highest-priority criterias of making itemization good
In terms of D4 ?
- Clarify the system of gauging for each of the item types
- If doing/redoing affixes make sure they’re immediately gauged on impact also
Or in other words:
- Items should be EITHER/OR => temperable, craftable i.e. succumbed to MW, socketable (not all of these at once)
- Whatever affixes on gear rolled have to be super clear what they do with an easy-to-gauge impact, i.e. things like
** Ignores barriers and damage reductions
** Damage over time stacks twice
** LH vs elites increased by 50/100/150%
** Take 25/50/75% less damage from elite affixes
** Overpower resets buff durations
** Cannot be vulnerable
** Basic attack deals double damage vs elites
** -10/15/20 sec CD on Ults
** e.t.c.
I’m not suggesting that these exact things are what D4 needs (in fact if things like this are done chances are would need to start over again lol), but => just wanted to share some thoughts based on long-existing (or a bit newer) games overall
But based on that criteria alone the best thing that D4 can do is:
- Drop mandatory MW
- Add another “gauging” metric which would suggest (but not oblige player) what to do with the item (socket, MW, or Temper)
For example:
- Additional affixes on this item have 5% increased value
- Additional affixes on this item have their value adjusted to at least 10%
- All potencies and chance-based affixes on this item have their value increased by 5%
e.t.c.
P.S., I know I’m a bit biased as you could see (and barely played PoE1 to include)