Please add an "Empty Affix" to the loot filter and let us assign a value to it

An empty affix often is nearly as good as a T5 of the most valueable possible affix because I (usually) can easily craft that valueable affix onto the item (at the cost of forging potential which is why it is only nearly as good).
And it is often much better than a T5 of a mediocre affix because - unless still progressing through low monos - I would have to try to remove it first via chaos glyph or removal rune.
Right now an empty affix inherently has a “tier equivalency value” of 0, i.e. it does not contribute to an items filter value at all, which results in the item being incorrectly filtered out.

My suggestion would be to add two “Empty Prefix” and “Empty Suffix” on top of the affix selection, each with a drop down (1 to 7) where we could assign a “tier equivalency value” to them because it depends on item and personal preference how much exactly we value an empty affix (but certainly more than 0 in any case).

U P D A T E
It seems, based on the feedback in this thread and the fact that the sum of good affixes is more important than the sum of their tiers (as we can, sufficient FP provided, always improve them), that adding “empty affix” to the list of selectable affixes might be sufficient to easilily improve the filters.
Also note the workaround laid out further below in this thread

5 Likes

That’s come up, and for some reason they couldn’t make it work, but they know it’s something desired and haven’t forgotten about it.

Good to know. Since it is all about priorities it probably does not hurt to bring it up every now and then.

Isn’t this what the # of prefixes and suffixes filters are? Or am I misunderstanding you?

eg if I want 3 tier 6s and one empty suffix, I can specify the the 2 prefixes and 1 suffix in the affix filter, and then say “min of 1 empty suffix” in the suffix filter.

Is that what you’re asking for?

I do not think so:

  1. You cannot specify the number of prefixes and suffixes, only the number of affixes.
  2. If you do not specify 4 as the number of affixes then you tell the filter that you do not care about the remaining affixes. That is, if you specify a minimum number of 3 affixes then the 4th one could be anything, either the “empty affix” or any other from the hundreds of affixes. That gives the “empty affix” the same weight as any other affix.
  3. There is no way to say “min of empty suffix” to the filter.

What I want is “empty prefix” and “empty suffix” to be added to the list of selectable affixes and also to be able to specify how much value (in terms of tiers) I would like to assign to them. The latter would be required because an “empty affix” inherently has no tiers.

1 Like

It is possible:

  1. show specific things
  2. hide items with 4 affixes
  3. show items with 3 affixes
  4. show magic items

Limit it to tier 4 mods, its only really useful for campaign leveling. For exalted items trying a rune of removal is always worth it.

I am not sure I am getting this.
The way I am understanding it, desirealbe items with an empty affix already get filtered out at your stage 0 because, as currently implemented, they do not contribute neither to the affix count nor (worse) to the tier sum.

Example for a specific filter:
1H sword with at least 2 of 13 affixes and tier sum of 11.
When I am looking at a sword with 2 desireable affixes and a tiersum of 10 AND and empty affix then it will still be filtered out even though I could easily craft 5 desireable tiers onto it.
No matter what you do in later stages you would not get it back. Or would you?

If however I could include an empty affix into the search and give it a “tier equivalency value” of e.g. 4 than that item actually had a tier sum of 14 and would pass the filter.

number of affixes only filters to show the affixes you put into the affix list

for example 3 of “void damage, vitality, (every heath and resist affix)” will search for those specific affixes, but it will not be able to tell between an item with an empty affix or one which has an unwanted affix

correct, now all you have to to is add another rule above that says: hide items with 2 of bad affixes

Youre not. I made it too abiguous. My filter structure needs to be replicated for every type of Armour you want it to handle. Your example would not get hidden by my proposed filter as it only generally hides all items with 4 affixes.
(0. show specific things
optional, not part of the open affix filter; this means specific items that you want to see regardless of the number of affixes; any specific types of affix combinations; tier sums for efficient runes of shattering; etc)

I dont understand your example or why youre bringing the tier sum into this.
To show all potentially perfect swords youll need to:

  1. SHOW the perfect Sword with 4 good affixes.
  2. HIDE all 1HSwords with at least 4 of any affix.
  3. SHOW all 1HSwords with at least 3 good affix.
  4. HIDE all 1HSwords with at least 3 of any affix.
  5. SHOW all 1HSwords with at least 2 good affix.
  6. HIDE all 1HSwords with at least 2 of any affix.
  7. SHOW all 1HSwords with at least 1 good affix.
  8. HIDE all 1HSwords with at least 1 of any affix.
1 Like

Now I get it, thanks.
And I did not get it because for me the natural way to build an (ideally single) filter would be the tier sum of good affixes. Hence my desire to assign a “tier equivalency value” to an empty prefix or suffix.
Your example has made me thinking this over. Thinking about it, my issue was probably still having too much POE in my veins were improving an affix was much harder which led me to ignore the fact that the number of good affixes is probably more importat than their tier sum in LE.

I allready have separate filter lines for all item types which are pretty cumbersome to maintain, also because I am swithing between multiple chars.
So the prospect of tripling (at least) the number of filter lines is not particularly enticing.

I am very thankful for this workaround, and also for the incentive to rethink my filter-“design” but I still believe that a better and simpler solution is required.