Unique rarity tiers [ Needs severe updating ]

"

Malone wrote:

Lioneye Glare T3 ? really ? it doesnt seem as common as other T3 uniques

I think GGG might've lowered its rarity. I have personally found 3 on torment. Back in early beta Lioneye's worth at least 10 times more than they do today... Funny how things works out.

Same as thunderfists... I sold 2 pairs of thunderfists during beta for 6ex/7ex each now they are just some super rare leveling item... worth no more than a few chaos. Was BiS for the oh so famous elemental cleave builds.

Marohi used to be GG item for many 2H builds and it was the 1st "7L" unique ingame. But GGG loves melee and 2H builds too much so no one wants these anymore.

Windscream is likely tier 3.

Greed's Embrace is likely tier 2, I've found only one of them and done a couple thousand runs since its release.

Edit: Woops responded to an old post, free bump for a great resource.

Supreme truth t2.

How could you forget SotS? Belongs at the top of T1....

I think I remember one item being in a tier of its own (1.5)... not sure but think it was goddess

I'll be updating this once more so by all means help me if you can and yes I know I forgot some items just remind me.

Mainly the new shit.

Thanks.

"

I_NO wrote:

I'll be updating this once more so by all means help me if you can and yes I know I forgot some items just remind me.

I'd completely forgotten you'd been compiling a listing; good to see the effort put in so far.

Though I might ask, what are we trying to judge rarity by? Mostly anecdotal examples?

I'd always considered the possibility of trying something a LITTLE more scientific, if somewhat less direct: comparing the popularity of an item versus its price. At least for non-leveling uniques, this might be a bit insightful, as we can entrust that supply/demand will adjust the price according to those two factors.

Another alternative would be to base upon the number in total circulation: this would be done by calculating the total number of ones posted for trade. While the DIRECT count in Poe.trade is capped to ">99" given that (as best as was can demonstrate) the odds of each roll are equal, we specify specific rolls, and scale the total by the (sub-99) figure it gives us. A couple examples:

• Aegis Aurora has 31 for sale with 100% armor/ES; it's a range of 21 (80-100) which gives an estimate of 651 in circulation (as opposed to "locked out"/in use)
• Shavronne's Wrappings has a Lightning resist range of 11 (30-40%) and there are 51 for sale with 40%, suggesting a total of 561 available. (I go with resist because the ES was changed with a patch)
• Mjölner has a range of 11 for its "Lightning damage" mod (30-40%) with 38 @40, suggesting a total of 418 in circulation.
• Bringer of Rain has a "increased armor & evasion" range of 101 (200-300%) with 19 @300, suggesting a total of 1,919 in circulation.

Granted, this is a very ROUGH estimate; better figures would require multiple samples for each item. But it does give a bit of a hint at how common each is.

For one, it backs up what I thought, that Aegis is actually a Tier 1 along with Shav's and Mjölner; its value is just much lower because it's a very specific build, (which, unlike Mjölner, is a DEFENSIVE based one that requires special investment) so its popularity is much lower.

Since this seems promising, later on I'll try my hand at putting in multiple samples per item, and more items, to see how this sorts out.

Wait you think Aegis is a T1? I've seen many of them even in the new leagues just now.

"

I_NO wrote:

Wait you think Aegis is a T1? I've seen many of them even in the new leagues just now.

The quantity that I see in existence suggests that it's T1. Likewise, it's still worth like 1-2ex for a non-legacy that not very many people use, which is VERY unusual; most "unpopular, but still endgame-viable" hand-slot items will often sink to 1 chaos or below, (see: Oro's, Doomsower, etc.) due to sheer supply outweighing demand.

So mostly, my conclusion, (partly backed by seeing that the game has close to an equal number of Aegis in trade as Shav's and Mjölners) is that the price is dictated more by demand than supply: a factor of 40-50 when the difference in rarity between T1 and T2, if I recall, is only like 4-5x.

Going on that, and my preliminary searches, I'd guess that, using my method, (assuming these are items that are traded pretty often) we'd see roughly the results scale along the number being traded:

• Tier 1: 250-750 in circulation
• Tier 2: 1,500-2,000 in circulation
• Tier 3: 4,000-5,000 in circulation
• Tier 4: 10,000+ in circulation

Of course, the lower-tier ones will be harder for a few reasons: I suspect people are more likely to not circulate them, (either use them or just vendor 'em) and the MUCH higher number may make my proposed method that much harder to use. (as if there's only 11 possible rolls for an item, ">99" could mean ANY number of 1k+)

Similarly, I recognize it'd be useless for maps, as those are consumable.