Ancient Items Roll Like Crap

General Discussion
Ancient Items Roll Like Crap

Hey Meteorblade. Here's another of the thousands of examples I can point to that tell me that the weighting on property values of Ancient items is different than normal Legendary items. Ancient items are most likely to roll in the bottom third of the range given, while Legendary items usually roll in the top third. Here's an example.

The first Ancient item of the Season dropped for me just now. An Ancient Legendary Witching Hour belt. Here is what rolled on it, with the ranges:

565 Vitality [550 - 650]
IAS +5.0% [5.0 - 7.0]
CHD +29.0% [26.0 - 50.0]
Entangling Shot +12% [10 - 15]
+6574 Life after Kill [6536 - 7725]
1.4% Chance to Freeze on Hit [1.0 - 5.1]

Every one of these properties rolled in the lower half of the possible range, and most of them very near the lower third of the range, if not lower. Now, except for a few glorious RNG exceptions, I believe this is the way the devs intended Ancient items to roll, so that even after we found one, we would be motivated to continue to play and find a better one. That makes sense. Also, it's pretty obvious after watching tens of thousands of these items drop, that my inductive conclusion is likely correct.

Now, don't just give me the statistical likelihood of a belt dropping with these property values. Give me a logical counterargument other than,
"You're wrong, because math."
Also ancient item with legendary power. For example, yang's with 150% MS, Condemn shield etc..
05/20/2019 11:59 PMPosted by StoneOld
Now, don't just give me the statistical likelihood of a belt dropping with these property values. Give me a logical counterargument other than,
"You're wrong, because math."


Thats pretty much the definition of a logical argument.

The alternative is to wait for someone to get a drop and say "wow, look! Awesome stats!", which is meaningless.
If you look for low stats, you will find low stats, if you look for high ones you will find them.
05/20/2019 11:59 PMPosted by StoneOld
Now, don't just give me the statistical likelihood of a belt dropping with these property values. Give me a logical counterargument other than,
"You're wrong, because math."


You're asking for something impossible since there exists nothing more logical than math.
Of course they do.

We are looking for very specific affixes, on top of nice rolls.

Out of the available pool, especially for off hand items, we are always going to have the odds against us.

Honestly, it would lose a lot of the "OMG LOOK AT THIS!" feeling if they were much more suited to our builds optimal stats.

I'd be up for a little nudge in our favour to the ancient/primal drop rates though.

As awesome as a great find can feel, it's just as horrible to find a primal you can't even use due to horrible rolls, even when it's your BiS item which I'm sure most of us have experienced.

To this day, I have found exactly 1 perfect primal. But oh is it glorious!
Deja vu all over again. Stone(head?) shows up with these posts every season.
I think getting ancient is matter of luck and chances.
I got so many ancient so far in season. My necro almost has all ancient on him and some items I got as ancient so many times. Theya re not pefect with stats, but they are ancient allowing me to finish 80 so far, and Im not a person who has luck in this game.
Some people just have more luck then others, for example I played 20 rifts with a friend, for that time I got 3-4 ancient items, he got 3 primals... thats luck...
In my opinion this season have more ancient drops then last season, I might be wrong but in just 3 days I got 30 ancient items...
Here you are complaining about rolling low stats. As long as they are not static they can roll anything. If people seem to find almost perfect ancients your argument is invalid.
05/21/2019 04:01 AMPosted by Singularity
I think getting ancient is matter of luck and chances.
I got so many ancient so far in season. My necro almost has all ancient on him and some items I got as ancient so many times. Theya re not pefect with stats, but they are ancient allowing me to finish 80 so far, and Im not a person who has luck in this game.
Some people just have more luck then others, for example I played 20 rifts with a friend, for that time I got 3-4 ancient items, he got 3 primals... thats luck...
In my opinion this season have more ancient drops then last season, I might be wrong but in just 3 days I got 30 ancient items...


Last season at the start my friend found 3 of the 2 piece set swords in 1 gr run. 2 of them rolled ancient. While I was still running 2 regular ancient swords on my HotA barb. Simple luck
05/20/2019 11:59 PMPosted by StoneOld
Give me a logical counterargument other than,
"You're wrong, because math."

When people provide maths and evidence that contradict your conclusions the logical thing to do is to change your conclusions. Instead, you insult the people providing the maths and evidence and stick with your original conclusions. So, as requested, here you go...

"You're wrong because you trust your feelings more than facts."

05/21/2019 01:54 AMPosted by Kirottu
You're asking for something impossible since there exists nothing more logical than math.

Logic. Logic is more logical than maths.

05/21/2019 02:33 AMPosted by Jazz
Deja vu all over again. Stone(head?) shows up with these posts every season.

Just because something gets repeated a lot doesn't make it true.
RnG

Nothing more. Nothing less.
The devs intended items to roll like crap for two reasons.
    It's special to receive a well-rolled item
    It gives longevity to the 'loot hunt'
05/21/2019 01:54 AMPosted by VPenguin
05/20/2019 11:59 PMPosted by StoneOld
Now, don't just give me the statistical likelihood of a belt dropping with these property values. Give me a logical counterargument other than,
"You're wrong, because math."


Thats pretty much the definition of a logical argument.

The alternative is to wait for someone to get a drop and say "wow, look! Awesome stats!", which is meaningless.
If you look for low stats, you will find low stats, if you look for high ones you will find them.
If I found high ones as often, or even half as often as I find crappy ones, then I wouldn't write these posts. Or do you think I find one in one thousand that are bad, and then immediately generalize all Ancient items...
05/21/2019 02:19 AMPosted by Jay
Of course they do.

We are looking for very specific affixes, on top of nice rolls.

Out of the available pool, especially for off hand items, we are always going to have the odds against us.

Honestly, it would lose a lot of the "OMG LOOK AT THIS!" feeling if they were much more suited to our builds optimal stats.
Well, that's why it's useful that an Ancient Witching Hour was the first Agent to drop the season. The Witching Hour has four primary properties, two of which are going to be IAS and CHD. Except in very rare occasions, one of the other three is going to be main stat or Vitality, and on about a 40/60 split the fourth one will be skill damage bonus/main stat or Vitality.

The secondary properties are a bit more variable, but I remember one Season when I jealously hung onto a Witching Hour that had 4.1% Chance to Freeze, because that was the highest roll on that property on a Witching Hour I had ever seen, and it still is to this day.
05/21/2019 05:42 AMPosted by Dmoney
Here you are complaining about rolling low stats. As long as they are not static they can roll anything. If people seem to find almost perfect ancients your argument is invalid.
But "people" don't. There's probably 50 different people that still use this Forum. So I'm pretty sure that my observations fall on deaf ears. I imagine if I went through the trouble to start with this belt, and log every single Agent item that drops for the duration of the Season, and then extrapolated the results, which I'm absolutely sure would support my observational conclusions, I'm also absolutely sure that 90% of the people that disagree with my findings would do so by simply saying, "Sample size," dismissing out of hand any argument that might convince a "normal" person. C'est la vie. I can't do anything about people who are obstinate. If I said to most of the users of this forum, "Hey, a blue car just drove by," they would say, "No, that was a green car. You just want it to be blue."
05/21/2019 06:35 AMPosted by Meteorblade
Just because something gets repeated a lot doesn't make it true.
Well that right there is something gets repeated a lot, and it is true.

But I'm not just saying something. I am drawing a conclusion based on empirical evidence, that evidence being tens of thousands of items that have dropped for me in this game, while I examine them closely (even Rare items before I scrap them). The conclusions I reach point to a specific weighting curve of property values.

Now it's important that you don't confuse "property values" with "properties." The property is, "Attack speed increased by…" And the property value is, "6%." I completely agree that the properties that appear on an item are totally random and don't follow any kind of fixed pattern. The Witching Hour is a nice example for my argument, because the randomization of its properties is fairly limited. But the values of these properties, when they do appear on an item, follow a fixed pattern of how they are going to roll based on the quality of the item.

Not one person who has ever responded to me negatively on this topic has provided a counterargument based on the same type of empirical evidence I've used, they just say I'm wrong, or they say "math," without presenting a cogent counterargument. Others, who do watch what drops, and form an opinion based on that, tend to agree with me.
I imagine if I went through the trouble to start with this belt, and log every single Agent item that drops for the duration of the Season, and then extrapolated the results, which I'm absolutely sure would support my observational conclusions

This section, right here, demonstrates the problem perfectly.

Instead of looking at the data and drawing a conclusion from it, you've already picked a conclusion and are trying to find data to support it.

The cogent counterargument is that claims made without evidence can be dismissed without evidence. As a claimant, if you cannot meet your burden of proof, your claim will be rightfully dismissed.
I have a file with thousands of legendaries (regular, ancient, and primal) that I've recorded upon identifying. I filtered the whole spreadsheet down to just ancients and I can confirm that the average stat roll is almost dead in the center of the stat range for every affix.

I will admit, I combined all the skill % affixes into one column because I figured that would give a better average. I also didn't check chance to bleed or native damage rolls on weapons because I don't actually know the exact ranges for the minimum and maximum and I know damage affixes changed based on weapon speed and I only recorded 1 handed or 2 handed weapon on my spreadsheet. I'm way too lazy to go through and plug exact weapon types in.
05/20/2019 11:59 PMPosted by StoneOld
Here is what rolled on it, with the ranges


If it wasn't for the freeze chance, that's some waste degree trash and even a that's low roll. Also you can not show a big enough example size with the rarity of Ancient and Primal items to prove your point.

What you're suggesting is biased because, with the 10% rarity you'd never have enough example size to compare roll ranges to regular items. No matter the example size, good and bad items would split it half somewhere roughly and high rarity of Ancients won't give you enough evidence to get your point across.

If you have, say, 100k legendaries; 90% of that would be regular and 10% would consist of primals and ancients; good and bad rolled items would still halve those groups. You'd have a really low amount of good rolled ancients compared to good rolled regular legendaries but they'd exist as a non-exception.
It's not hard to tell because you can only re-roll one affix and smart roll takes like 1-2 possibilities out of 6 already; if after reroll, item still unusable it's a bad item.
This topic would be moot, if Blizzard actually adjusted
ancient versus regular items properly.
Hellfire amulet crit damage is 51-100%,
an ancient should be 75-100%.
Primals should also have secondary re-rolls.
Who wants a Dead Man's Legacy without Discipline.
05/21/2019 11:51 AMPosted by Dreyda
I have a file with thousands of legendaries (regular, ancient, and primal) that I've recorded upon identifying. I filtered the whole spreadsheet down to just ancients and I can confirm that the average stat roll is almost dead in the center of the stat range for every affix.

I'm genuinely interested in how StoneOld / StoneCold / Slamboney is going to attempt to dismiss this actual empirical evidence that disproves his theory that ancients tend to have values lower in the ranges.
You get ten times more legendaries than you do ancients and you burn every single one which doesn’t beat your currently equipped legendary.

By the time you find an ancient, you will already have a comparatively rare set of stat rolls on your legendary.

This is just in your head.
Don't worry, I won't bother you with Math or such non sense!
I have only one thing to say, awesome reasoning because.. Unicorns! :D

Join the Conversation

Return to Forum