Raven at the Door Loot Strangeness

[Certon]

Today, I ran a Raven at the Door on 5 characters, 3 times each, doing the optional to have 2 shots at getting a Legendary Phasecloak, because it is the only cloak in the game that will upgrade one of my apex toons.

Results:
11 Legendary Bracers of the Fallen Hero
5 Legendary Reflective Bloodstones
1 Legendary Skulled Ring
1 Legendary Memory of Hiding in Darkness

It's a pretty small sample size, but even this suggests the cloak is rarer than the bracers and the bloodstone and may be as rare as the ring and the docent.

SO FRUSTRATING. Ugh.

[KoobTheProud]

Today, I ran a Raven at the Door on 5 characters, 3 times each, doing the optional to have 2 shots at getting a Legendary Phasecloak, because it is the only cloak in the game that will upgrade one of my apex toons.

Results:
11 Legendary Bracers of the Fallen Hero
5 Legendary Reflective Bloodstones
1 Legendary Skulled Ring
1 Legendary Memory of Hiding in Darkness

It's a pretty small sample size, but even this suggests the cloak is rarer than the bracers and the bloodstone and may be as rare as the ring and the docent.

SO FRUSTRATING. Ugh.

It's just random. If you were looking for the bloodstones you probably would have gotten 5 phase cloaks instead.

[rabidfox]

It's a pretty small sample size

Can stop right there; that's covers what you saw.

[Bjond]

SO FRUSTRATING. Ugh.
Most MMOs use a pRNG algorythm -- psychic random number generator. It knows what you seek. It wants you to press that astral shard re-roll button. Give in. Feed it shards. And you shall be rewarded.

The rat that presses the button gets the cheese!

[Carpone]

RNG on a tiny sample size isn't strange.

[Eantarus]

Today, I ran a Raven at the Door on 5 characters, 3 times each, doing the optional to have 2 shots at getting a Legendary Phasecloak, because it is the only cloak in the game that will upgrade one of my apex toons.

Results:
11 Legendary Bracers of the Fallen Hero
5 Legendary Reflective Bloodstones
1 Legendary Skulled Ring
1 Legendary Memory of Hiding in Darkness

It's a pretty small sample size, but even this suggests the cloak is rarer than the bracers and the bloodstone and may be as rare as the ring and the docent.

SO FRUSTRATING. Ugh.

I've witnessed the exact same behavior multiple times. Supposedly(according to the devs), all named items are weighted equally. I believe this is true on paper, but that there is a bug in the RNG that makes items drop in "clumps" like you're seeing there or just with considerably greater frequency than others.

To those continuing to deny it, which would you say is more likely:

-That a statistically unlikely event happens with a high degree of regularity

OR

-That a game riddled with bugs has a bug?

[niknight]

I've witnessed the exact same behavior multiple times. Supposedly(according to the devs), all named items are weighted equally. I believe this is true on paper, but that there is a bug in the RNG that makes items drop in "clumps" like you're seeing there or just with considerably greater frequency than others.

To those continuing to deny it, which would you say is more likely:

-That a statistically unlikely event happens with a high degree of regularity

OR

-That a game riddled with bugs has a bug?

I attribute it to a form of selection bias. The only time that people ever take data like this is when they're experiencing the bad end of variance. Nobody ever posts "statistical analysis" for items they get on their first few runs. Because of this, we get a very skewed image of how the loot tables work in practice. Even the person who posted the Ravenloft item drop data a few years ago wasn't able to disprove the null hypothesis, even after about 300 runs of the saga.

[Wizard1406]

I wonder why loot is reliant on RNG only in DDO (save a few turn ins). It makes getting loot difficult for people with bad luck


- If it is to make content last longer, why not introduce loot via tokens or quest rewards when a (expansion-)pack is older, like over one year?

- If it is to push shard sales for rerolls, why are they limited to 3x per chest (and ransack after 8 runs) , even for older content?

[Weemadarthur]

I've witnessed the exact same behavior multiple times. Supposedly(according to the devs), all named items are weighted equally. I believe this is true on paper, but that there is a bug in the RNG that makes items drop in "clumps" like you're seeing there or just with considerably greater frequency than others.

To those continuing to deny it, which would you say is more likely:

-That a statistically unlikely event happens with a high degree of regularity

OR

-That a game riddled with bugs has a bug?

There is another aspect to this that the playerbase (and devs for that matter) never seem to fully understand. If the loot is set to a set drop pattern (lets say it rolls a d100 to simulate % chance) each time an item drops this would give a 20% chance of each item dropping every time an item can drop if there are 5 items. At a glance this looks like an equal spread and each item has a set 20% chance to drop every time an item can drop. However if anyone spends any significant amount of time rolling a d100 through the UI you will notice most rolls are between the 30-70 range. This is due to the rolls themselves not being actually random but only psuedo random. Extend that same roll pattern to the % chance of specific items dropping and you would see a larger drop rate of the items that are set to the middle of the dice roll than those at the edges. The items that require a roll of <20 or >80 would drop considerably less.

So although there may be a 20% chance of getting a specific item per roll in theory due to weighted dice rolls those items that are linked to the 40-60 rolls would drop most often, the 20-40 and 60-80 rolls would be frequent but lower than 40-60 and those that are linked to the 1-20 and 80-100 would become quite rare.

[Eantarus]

I attribute it to a form of selection bias.

The only problem with this theory is that you can very reliably repeat the OPs result's any time you want. I can open up DDO any day of the week, pick a chest, and ransack it across multiple characters and derive a very clear clumping pattern where specific items drop more frequently than others. I have personally repeated this experiment at least 10 times.

If you want an easy one to try: check out Into the Mist. Its very short and relatively easy. Take a capped character in and ransack heroic elite a few times. You'll see the same thing.



I wonder why loot is reliant on RNG only in DDO (save a few turn ins). It makes getting loot difficult for people with bad luck


- If it is to make content last longer, why not introduce loot via tokens or quest rewards when a (expansion-)pack is older, like over one year?

- If it is to push shard sales for rerolls, why are they limited to 3x per chest (and ransack after 8 runs) , even for older content?

I don't personally believe that the drop rate issues are intentional. I don't think SSG sat down and said "Let's make it difficult for some people". I think this is much more likely a design flaw coupled with improperly-implemented RNG.

Still, though, relying so heavily on RNG is a very poor design choice. Its extremely frustrating for the players across all demographics.



There is another aspect to this that the playerbase (and devs for that matter) never seem to fully understand. If the loot is set to a set drop pattern (lets say it rolls a d100 to simulate % chance) each time an item drops this would give a 20% chance of each item dropping every time an item can drop if there are 5 items. At a glance this looks like an equal spread and each item has a set 20% chance to drop every time an item can drop. However if anyone spends any significant amount of time rolling a d100 through the UI you will notice most rolls are between the 30-70 range. This is due to the rolls themselves not being actually random but only psuedo random. Extend that same roll pattern to the % chance of specific items dropping and you would see a larger drop rate of the items that are set to the middle of the dice roll than those at the edges. The items that require a roll of <20 or >80 would drop considerably less.

So although there may be a 20% chance of getting a specific item per roll in theory due to weighted dice rolls those items that are linked to the 40-60 rolls would drop most often, the 20-40 and 60-80 rolls would be frequent but lower than 40-60 and those that are linked to the 1-20 and 80-100 would become quite rare.

That could very easily explain the problem. In fact based on previous posts I would say this is most likely exactly what's at play, except maybe that the actual order of the table is different sometimes, seeing as some players report being flooded with specific items that are borderline none-existent for others.

[Aelonwy]

That could very easily explain the problem. In fact based on previous posts I would say this is most likely exactly what's at play, except maybe that the actual order of the table is different sometimes, seeing as some players report being flooded with specific items that are borderline none-existent for others.

Not that my family tries to farm much but when we do we definitely see one item dropping most frequently of the bunch... but that item might change the next time we try to farm that chest. For example, we ran Blown Deadline, so many times. So, so many times. One time we walked away with 3 Wildhearts each and only 2 other items dropped at all. The next time through Hammerfists were dropping like candy. Eventually, we came away with one L.Celestial Sapphire ring each and gave up.

[Eantarus]

Not that my family tries to farm much but when we do we definitely see one item dropping most frequently of the bunch... but that item might change the next time we try to farm that chest. For example, we ran Blown Deadline, so many times. So, so many times. One time we walked away with 3 Wildhearts each and only 2 other items dropped at all. The next time through Hammerfists were dropping like candy. Eventually, we came away with one L.Celestial Sapphire ring each and gave up.

I had a similar experience with Into the Mist and the Vistani Fighter's Sash?- first time I went farming it was all that would drop, then recently I couldn't get one in 60 chest pulls. I've got easily half a dozen different examples. My quest to get an Enigma Core from Reach for the Sky **** near made me quit the game.

So obviously its a bit more complicated than each item 1-5 is assigned 20% of d100, but whatever the game is doing on the backend it does not work. Clumps are real. It is not selection bias.

[Bjond]

I wonder why loot is reliant on RNG only in DDO
It's not just DDO. It's most MMOs. VERY few use any kind of currency-buy system in general. GW2 is the most notable exception where each dungeon run acts a lot like raid runes here and there are flat out no random drops. It's all save & buy.

The "why" of it is human psychology. IIRC, humans just plain don't give up even versus horrible odds and will keep "pressing the button" FAR longer when faced with a v.bad random chance at cheese versus the save+buy scenario. It's a deliberate design choice by MMO creators to keep people playing their game.


ransack it across multiple characters and derive a very clear clumping pattern
"Clumping" is normal in a random sequence (both true-random and psuedo). You need a LARGE sample size to prove something is actually statistically broken.


likely a design flaw coupled with improperly-implemented RNG.
A truly flawed implementation would be one where we as players could easily predict the results. I have seen games where you could do that, but DDO doesn't seem to be *that* bad.

My bet is they're using nothing special for their pRNG -- just whatever came with the language library they're using for their server. They may or may not be using it poorly; eg. discarding entropy due to over/underflow in their calculations.

IMHO, drop chances have been tuned to be appropriate ONLY for shard-rerolls and this is done purely to drive shard sales. The evidence to support that is soft, but the fact that chests with zero drops also present a shard-roll button does speak rather loudly as to their desire to deliver fair value versus their desire to drive shard demand.

[Wizard1406]

Many MMOs I played have tokens, often as an alternative or "bad luck" protection. And quest reward gear, 100% chance , for leveling.

SWTOR, FFXIV, , GW2 , WoW used to have it and I think Rift too, , only played a little ESO but it had quest rewards. How is the loot drop in LOTOR ? At least you can get gear for leveling 100% chance from quests I remember.


The drop rates are low and it can take long (both in RL weeks and total number of runs) to get the right item in DDO, because there are hardly any farm groups except maybe when a pack is new. With 6 people you have a higher chance of getting a specific item. Drop rates seem to be balanced around having 4+ people parties or using max rerolls every time.

Anyway, looking on the bright side, you can get BiS level cap dungeon gear in DDO completely solo if you want. (only raids need a group)

[yfernbottom]

This game and many others suffer from odd clumpyness in RNG results. If something absurdly unlikely happens, you are never more likely to see it happen again than the next time you roll.

In this game in particular, a lot of small things seem off. I see odd streaks in loot, and odd strings of successes and failures. I honestly believe there is more to it than the clustering illusion:

https://en.wikipedia.org/wiki/Clustering_illusion

[Certon]

I'd love a token system where you trade in a token drop from a chest for an item of your choice from the end reward.

BUT

If it gets super easy to gear up, the danger of being 'done' with a character's equipment and the ensuing boredom rises.

I honestly don't know the fix or even if one is necessary. Maybe trade in multiple items for one at a 5:1 ratio?

[Eantarus]

It's not just DDO. It's most MMOs. VERY few use any kind of currency-buy system in general.

Actually what's normal for most MMOs is published drop rates - EG "Item X has chance Y to drop from chest Z". DDO is an outlier in that they deliberately obfuscate the drop rates. The published 10/20/33% is only for base named items in specific content, we have no clue what the Nebula Fragment rate is but its definitely not 33% on EE.



"Clumping" is normal in a random sequence

Clumping is normal but not with the kind of frequencies that we are seeing. A few clumps in a sample set is normal, being able to reliably and predictibly produce clumps every single time points to a problem.




IMHO, drop chances have been tuned to be appropriate ONLY for shard-rerolls and this is done purely to drive shard sales. The evidence to support that is soft, but the fact that chests with zero drops also present a shard-roll button does speak rather loudly as to their desire to deliver fair value versus their desire to drive shard demand.

This does seem to be the case.

I know a guy who's done a large number of shard re-rolls and gathered data on it. Comparing my experiences to his - just doing chest-pulls nearly always produces clumps, but doing re-rolls provides a much smoother experience. He can usually get all 4 items out of a quest using 1 character and doing re-rolls before hitting ransack. That's usually about 5-7 chest opens with 40-50 re-rolls, but he does it.

[Wizard1406]

Yeah 5:1 trade in seems reasonable. People with average luck would already have the item by then (in a quest with 5 different items, less runs needed with average luck if only 3 or 4 different items drop), but it would be a very nice bad luck protection.

[Eantarus]

I'd love a token system where you trade in a token drop from a chest for an item of your choice from the end reward.

BUT

If it gets super easy to gear up, the danger of being 'done' with a character's equipment and the ensuing boredom rises.

I honestly don't know the fix or even if one is necessary. Maybe trade in multiple items for one at a 5:1 ratio?

The fix would be to figure out why the RNG is broken. None of us would be complaining if it didn't occasionally take 500 chest pulls to get a base item.

[yfernbottom]

I honestly don't know the fix or even if one is necessary. Maybe trade in multiple items for one at a 5:1 ratio?

Five to one sounds about right to me. Maybe an NPC that will trade you an item you don't want for a token?

[Eantarus]

Five to one sounds about right to me. Maybe an NPC that will trade you an item you don't want for a token?

Alternative fix: make all non-raid loot unbound. Let me trade my 5 items I don't want/need for your 1 I do. You can feed the stuff you don't want to your sentient gem, and I don't have to do 495 more chest-pulls.

[Bjond]

Actually what's normal for most MMOs is published drop rates - EG "Item X has chance Y to drop from chest Z"
Federal Lottery Laws require the chance for each reward to be published. IMHO, if chests did not have paid rerolls and simply coughed out one set of rewards, whether these laws applied or not would be a very grey area. The other MMOs you're mentioning are being overly cautious with their publishing or doing it to meet requirements for other countries.

OTH, the rerolls are a crystal clear violation in this and other ways. SSG is foolish for even having them in game. Getting enforcement interested might be hard, but once you do, this one would be a slam dunk.

[Wizard1406]

Alternative fix: make all non-raid loot unbound.

This would be interesting for the economy. Prices would start out high but eventually drop low for items from quests that are part of the TR cycle, because they get played so often.

Maybe regular items without mythic +3 / reaper would become very cheap and thus easy to obtain, but leveling items should be easy to get in an MMO IMHO. Lvl 29 items would have a price floor because they are worth quite a bit of sentient xp.

[Aelonwy]

Speaking of clumps, or where random streaks... I have an example that I was reminded of this morning. You see I have a hagglebard and because I like to Mirror things I drop all my armor and cloaks in the shared bank for her to sell next time I log in. So all armor/cloaks I picked up yesterday evening get tried on and sold all in one go. I've been doing this for years. Everyone is aware that armor comes in two parts, right? the under armor and the accessory layer? And cloaks seem to have two colors? even if they're a solid color it just means both colors are the same color. Well I've been noticing for sometime that the colors can come in streaks. That is, out of the ~20 armors and robes I picked up yesterday the accessory layer of 11 of them was flat grey including all 6 of the robes. I only picked up 3 cloaks but all three had hot pink as one of the colors. These sort of color streaks happen often in random gen, at least IME. I don't know what that means in the course of things but its been going on for a long time. Long enough anyway that I get excited when I see a purple/violet streak because its my favorite color and I'm still hoping to pick up the combo below for Mirror purposes:

https://i.imgur.com/DEgvImb.jpg

[Amorais]

Perception bias. Its like thinking every time I try to save vs Hold Person I roll a "1".

But that's the only time I do notice the actual spell getting cast

So...it always looks like I roll a 1 on a save for Hold Person. Reality is, its as random as random can get on a computer.

[Saekee]

well if it makes you feel better, I could not get the right item to drop in a quest after repeated tries until I realized I was in the wrong quest.

On a happier note, got it today on second run with a great PUG on R5.

[Eantarus]

This would be interesting for the economy. Prices would start out high but eventually drop low for items from quests that are part of the TR cycle, because they get played so often.

Maybe regular items without mythic +3 / reaper would become very cheap and thus easy to obtain, but leveling items should be easy to get in an MMO IMHO. Lvl 29 items would have a price floor because they are worth quite a bit of sentient xp.

It would provide the game with an actual economy for a change, which would be nice. Even if its not "most recent", now that they practically gave away Ravenloft whats to loose by unbinding all the gear?

[Eantarus]

Federal Lottery Laws require the chance for each reward to be published. IMHO, if chests did not have paid rerolls and simply coughed out one set of rewards, whether these laws applied or not would be a very grey area. The other MMOs you're mentioning are being overly cautious with their publishing or doing it to meet requirements for other countries.

OTH, the rerolls are a crystal clear violation in this and other ways. SSG is foolish for even having them in game. Getting enforcement interested might be hard, but once you do, this one would be a slam dunk.

It would seem then that DDO us running very much afoul of these laws, since the only published drop rate we have is only for getting a subset of the named items that can possibly drop from a chest, with no hard rate for any one specific item and many rarer items having no data at all.

What, for example, is the drop-rate on Jibbers? We know its not on the same loot table as the 2 named items in that chest and thus does not conform to the 10/20/33% rate published.

[Stormraiser]

I've witnessed the exact same behavior multiple times. Supposedly(according to the devs), all named items are weighted equally. I believe this is true on paper, but that there is a bug in the RNG that makes items drop in "clumps" like you're seeing there or just with considerably greater frequency than others.

To those continuing to deny it, which would you say is more likely:

-That a statistically unlikely event happens with a high degree of regularity

OR

-That a game riddled with bugs has a bug?


This is often due to an inherent bias when searching for an item. When you pull the item on the 1st shot, you don't think about it. You don't create a tracker to track how long it takes to pull an item if you pull it on your 3rd try. Thus what you remember, are really the times when something takes a very long time to pull.

Coincidentally, shuffling a 52 card deck and dealing 5 cards which are a royal flush does not mean the deck is not shuffled, it just means that those are the first 5 random cards. Some people have a concept that cards have to come out all scattered and disconnected for the deck to be truly shuffled, but if pairs or flushes never come out, that's also the sign of a non random deck.

[Eantarus]

This is often due to an inherent bias when searching for an item. When you pull the item on the 1st shot, you don't think about it. You don't create a tracker to track how long it takes to pull an item if you pull it on your 3rd try. Thus what you remember, are really the times when something takes a very long time to pull.

Coincidentally, shuffling a 52 card deck and dealing 5 cards which are a royal flush does not mean the deck is not shuffled, it just means that those are the first 5 random cards. Some people have a concept that cards have to come out all scattered and disconnected for the deck to be truly shuffled, but if pairs or flushes never come out, that's also the sign of a non random deck.

The only problem with this logic is that you can sit down and reproduce the exact same results any time. I've done the test multiple times without looking for anything specific, just ransacked the chest across multiple characters and logged my results. Try it for yourself. Do 100 pulls of the same chest and I guarantee you will finds two things: 1. that the actual drop rates are way lower than the published 33%, and 2. that certain items drop way more frequently than the laws of probability would dictate.

Seriously, how much data do you "selection bias" people have to see before you understand this is not in our heads?

[CSQ]

It would seem then that DDO us running very much afoul of these laws, since the only published drop rate we have is only for getting a subset of the named items that can possibly drop from a chest, with no hard rate for any one specific item and many rarer items having no data at all.

What, for example, is the drop-rate on Jibbers? We know its not on the same loot table as the 2 named items in that chest and thus does not conform to the 10/20/33% rate published.

I think it's been mentioned before, but it is worth noting that drop rates only have to be listed for items that are directly purchased under the law that Bjond references. Technically, even rerolls wouldn't fall afoul of most laws because you're not purchasing the item, you're purchasing a chance to try again- the initial chance is "free". There also isn't any federal law (in the US) about item boxes- there are a ton of games that don't publish even the RNG rates for cash shop items. There is a law somewhere (in China if I remember correctly) that requires disclosure of RNG rates associated with premium features, which is why games with international markets (like Warframe, for example, which publishes its loot tables to very precise levels, because they have a universal drop boosting item that would violate that law with its wording). There's also a law in the UK specifically for loot boxes (which would technically apply to Daily Dice) but since DDO isn't on consoles and SSG doesn't operate locally in the UK it isn't going to be subject to that restriction. However, in the US where SSG is located they have no legal obligation to disclose it. If I'm wrong on this, I would love someone to actually cite the law (it would be public record) so I could look at the wording, because it's likely that DDO would still be in the clear because it's simply a reroll instead of actually buying loot.

It is also worth noting that, at a point in the past, a developer has broken down the drop system. I don't remember who it was, and while they didn't mention the specific drop rates for rare items (i.e. those that don't drop on the "primary" drop tables) there is some information there worth noting:

1. DDO has primary/secondary/tertiary systems for loot, each with different impacts and qualifications:
There is "random" loot that generates in chests, which is the product of quite a few different rolls on tables (basically, determining the quality, number, and type of randomly generated loot like weapons, armor, gemstones, etc. that show up in chests guaranteed). This is what is impacted by, for example, treasure hunter elixirs.
There are, for most quests past a certain point in DDO's development history, primary named item lists. These have the 10/20/33% rates that people mention (this rate actually increases in Reaper, but I can't remember the details here so I'm not going to guess at it). This is impacted by discovery elixirs and includes rolls for mythic/reaper bonuses if appropriate.
There are also secondary item lists, which can draw from different factors. The Cursed Blade of Jack Jibbers is, for example, a very rare (about 1-2% I would guess) drop from "A Legend Revisited" (the technical name for the epic version of the Legend of Two-Toed Tobias, even though no one calls it that because we all know the quest from communal suffering- I was cursed by the RNGesus and must have run that quest at least a hundred times farming for it). However, there are also other secondary lists- tomes dropping from raids, Siberys Dragonshards, some items that can be found from any quest in certain level ranges, etc. all have unique lists that don't impact named item lists (though this might not always be true for older quests, because anything before the 10/20/33% rate implementation is a bit unclear, I'm pretty sure this is true in almost every case). These generally are not impacted by items (so, for example, you won't get a Cursed Blade of Jack Jibbers more often using a discovery elixir) or other modifications if I remember correctly.

2. RNG loves confirmation bias. Unless you've logged every attempt, and you've absolutely made certain that you recorded your data right, you might be incorrect. For example, I mentioned that I ran "A Legend Revisited" like a hundred times before getting it. Now, I'm fairly certain based off the number of times I ransacked the quest that's accurate, but I couldn't give you a concrete number because I didn't write it down or log it particularly well, so I couldn't tell you how many times I rerolled the chests or the exact number of runs I did, so take it with a grain of salt. I know that I ransacked it for over a month, so it had to be at least a bare minimum of 40 on that attempt alone (and I didn't get it until a long while after that, ER/TR'ing several times in between with more, less dedicated failed attempts). I can also tell you that I don't remember how many times I've dropped it since (and yes, I still run it, Stockholm Syndrom is a thing apparently), so clearly I may have just had a dry streak offset by later (wasted) wealth.

3. As with everything RNG, it is ultimately random. I have been in farm groups where people find items that I'd ransacked quests for on their first runs. Now, that doesn't mean that there shouldn't be conversations on, say, implementing bad luck protection (something SSG has been better about, at least for raids, lately) because RNG can be very frustrating, but it is worth noting that people don't go to the forums to talk about how bad their RNG is unless they've had particularly bad luck (or extremely high expectations). For every player who suffers an absurdly long time farming a particular item, there's someone else who gets it without even trying (well, as a figure of speech- depending on the RNG rates that might not always be true, but *relatively speaking* it's probably still true across rarities).

[Weemadarthur]

3. As with everything RNG, it is ultimately random.

Points 1 & 2 are perfectly valid but unfortunately this part is pretty much wrong. I explained in a previous post about why I think "clumping" takes place like it does but I think here I will explain a little further. 1st up RNG in computer games is never actually random it's more pseudo random in that there are patterns to be observed if you look for them. If you roll a d100 in chat as many times as you feel is fit to sustain a truly random amount (I have only logged 100 at any 1 time but have done this a lot. I will come back to this a bit later though and explain why) you will notice that certain totals will be more regular than others. If you take the totals of all rolls made over 100 attempts you will see that the totals will follow certain rules. Over the course of 100 rolls you will never see more than 10% of the rolls being over 90 unless the majority of rolls are below 50. By the same token you will see that the only time you roll under 10 more than 10% of the time is if the majority of the rolls are over 50. If you do this enough times you will also see a pattern emerge where the majority of rolls are within the 25-75 range. This is not something that happens occasionally but something that is pretty much the norm. I expect if you roll a d100 enough times (probably a few hundred thousand at my best guess) you will see that the rolls do their best to total 100 as an average.

Now I started looking at this a while back (this is the later I mentioned earlier lol) to work out if great sword or great axe was a better option on my barb. I did a lot of extensive testing to see if the crit range or crit multiplier gave better results. My finding were that the great sword seemed to crit a lot more frequently than the extended crit range would account for. This lead me down the path of testing crit chances using my rogue mech to see how big a difference it made. I equipped him with 100 bolts and started with a regular x-bow (10% crit chance) and after several trials (10 to be exact using 100 bolts each time) recording how often I crit I found my average was 7% for crits. I then took the enhancement to increase my crit chance by +1 (15% total) and did the same test. Average of 14%. I then reset my enhancements and took imp crit (20% total) and got 21% as an average. I then took both (for 25%) and got a whopping 29% average.

Now I have done the same test multiple times using 1000 bolts for each bracket each time and always ended up with similar results. This then lead me onto doing the /d100 test which I have recorded now over 6000 rolls. Now I know this is not conclusive evidence and is still imo a small sample size but with combining the findings from both it is fairly easy to see that there is a pattern to the "random" rolls that the game makes. It is also pretty clear that the dice are definitely weighted towards the center numbers (25-75 rolls contain a lot more than 50% of rolls I have recorded to date regardless of how many rolls I have done at any point in time (68% at this point)).

[Certon]

This is often due to an inherent bias when searching for an item. When you pull the item on the 1st shot, you don't think about it. You don't create a tracker to track how long it takes to pull an item if you pull it on your 3rd try. Thus what you remember, are really the times when something takes a very long time to pull.

Coincidentally, shuffling a 52 card deck and dealing 5 cards which are a royal flush does not mean the deck is not shuffled, it just means that those are the first 5 random cards. Some people have a concept that cards have to come out all scattered and disconnected for the deck to be truly shuffled, but if pairs or flushes never come out, that's also the sign of a non random deck.

Not true with me. When I've pulled the item I wanted on the first try, it happens so rarely that I remember it! I'm like "Wow! This almost never happens!"

[Certon]

Swim at Your Own Risk is a great quest to check for clumping because it is so quick and has 4 unique rewards. I've done it 100 times.

In that time I've pulled 15 Legendary Crocodile Skin boots
8 Legendary Crocodile Tooth Trinkets
5 Legendary Crocodile Maw handwraps
and 2 (TWO!!!) Legendary Belts of Myriad Pockets

It feels off pretty consistently. You'd think in 100 pulls the numbers would be more even, all things being equal. And this isn't unique behavior.

[ahpook]

... Now I have done the same test multiple times using 1000 bolts for each bracket each time and always ended up with similar results. ...

I don't doubt your testing but I would have thought that any system using RNG this heavily would have unit tests for this kind of clumping.

Hey devs, if you have such tests and you can confidently say that Weemadarthur is more than wee mad, great. But if you don't have such tests in your system, you really oughta write those ....

[Bjond]

rerolls wouldn't fall afoul of most laws because you're not purchasing the item
Money is not a qualifier on either side (spend or reward). The qualifiers are CHANCE and value, not money; eg. donating time to a charity for a raffle ticket to win a date with a local celebrity still qualifies and still requires published chances. It's an "eye of the beholder" type of deal; if that date lotto enticed a rabid fan to obtain lots of tickets and the chance was not published or was wrong, there are serious consequences both civil & criminal.

IMHO, the only thing protecting folks doing it right now is the attitude of the justice department, which is mostly "uh, lol?" for online games. Eventually someone with sufficient political pull with get seriously irked and insist on enforcement and whoever irked them will get hammered. Then the rest of the industry will adapt very quickly.


take the totals of all rolls made over 100 attempts you will see that the totals will follow certain rules
The 100x1d100 test you're using has issues due to 1) sample size MUCH too small, 2) unknown range limiting function, 3) unknown relation to looting, and 4) RNG is likely "shared" (ie. you compete for numbers with others, you don't get one series all to yourself).

IMHO, based just on DDO's quality issues, I'd guess (LCG mod range) is what they're using (ie. worst possible way to use it), but showing that via external testing would take far more work than is worth contemplating and it would be effort to no point. They're not changing the way they use their RNG based on forum posts. The only useful goal would then become prediction and point (4) above makes that unlikely.

By all means try, just don't get your hopes up with ultra-tiny sample sizes -- aim for something like range^3 (1M for 1d100) for decent results.

[Weemadarthur]

The 100x1d100 test you're using has issues due to 1) sample size MUCH too small, 2) unknown range limiting function, 3) unknown relation to looting, and 4) RNG is likely "shared" (ie. you compete for numbers with others, you don't get one series all to yourself).

IMHO, based just on DDO's quality issues, I'd guess (LCG mod range) is what they're using (ie. worst possible way to use it), but showing that via external testing would take far more work than is worth contemplating and it would be effort to no point. They're not changing the way they use their RNG based on forum posts. The only useful goal would then become prediction and point (4) above makes that unlikely.

By all means try, just don't get your hopes up with ultra-tiny sample sizes -- aim for something like range^3 (1M for 1d100) for decent results.

I did state in the post you quoted that I agree that the sample size was too small for anything definitive to be taken from it. The point was to highlight that there is a lot less random to the random rolls than some people believe. The fact is though that although I wouldn't by any measure call it conclusive evidence it does support the fact that "clumping" on the loot table does have a logical reason behind why it may happen other than (or as I believe to go along side) observational bias.

In short I don't disagree with those that are saying it's observational bias as this will most definitely be in play. That doesn't invalidate the fact that the observations that players are basing their bias on are likely also correct. Clumping causes bias which makes the clumping more noticeable reinforcing the bias.

[boredGamer]

Swim at Your Own Risk is a great quest to check for clumping because it is so quick and has 4 unique rewards. I've done it 100 times.

In that time I've pulled 15 Legendary Crocodile Skin boots
8 Legendary Crocodile Tooth Trinkets
5 Legendary Crocodile Maw handwraps
and 2 (TWO!!!) Legendary Belts of Myriad Pockets

It feels off pretty consistently. You'd think in 100 pulls the numbers would be more even, all things being equal. And this isn't unique behavior.

Wait, since when have the devs indicated the chances for each named item is equal ? I thought it was 33% chance OF a named item. We know for sure Jibbers, etc don't have the same chance, why do we expect all named items to have the same chance?

[Eantarus]

Wait, since when have the devs indicated the chances for each named item is equal ? I thought it was 33% chance OF a named item. We know for sure Jibbers, etc don't have the same chance, why do we expect all named items to have the same chance?

Why the 2 month thread necro?

Per devs: 33% of a named items, all named items are weighted equally. This was confirmed by Steel in a dev post some years ago.

Addendum: while the devs have stated all items are weighted equally, its pretty clear this is not true. There is strong evidence of a problem with clumping on DDO's RNG that seems to affect some players more than others.

[Weemadarthur]

Why the 2 month thread necro?

Sorry that's my fault. Bored and myself where discussing this in another thread so I linked him to this one so he could see what I was talking about for reference. He does raise an interesting point though that if there are items like Jibbers that we know definitely don't follow the same drop % rules as other items it could potentially add another factor to the equation. There has been to my knowledge no official statement that drop rates are equal for all items. Devs have been wrong in the past and as much as I like Steel he could be wrong on this.

[boredGamer]

Sorry that's my fault. Bored and myself where discussing this in another thread so I linked him to this one so he could see what I was talking about for reference. He does raise an interesting point though that if there are items like Jibbers that we know definitely don't follow the same drop % rules as other items it could potentially add another factor to the equation. There has been to my knowledge no official statement that drop rates are equal for all items. Devs have been wrong in the past and as much as I like Steel he could be wrong on this.

Also TWO months necro thread is what we are worried about eant? There’s a ten year necro at the top of the list.

Yeah , where is this all items drop equally quote ?

[Eantarus]

Sorry that's my fault. Bored and myself where discussing this in another thread so I linked him to this one so he could see what I was talking about for reference. He does raise an interesting point though that if there are items like Jibbers that we know definitely don't follow the same drop % rules as other items it could potentially add another factor to the equation. There has been to my knowledge no official statement that drop rates are equal for all items. Devs have been wrong in the past and as much as I like Steel he could be wrong on this.

I'm going to tell you both what I have pieced together from years and years of dev quotes:

*There is a 10/15/33% chance for an item to drop based on quest difficulty
*This rate only applies to named equip items
*Everything else is on a separate loot table and the drop rates for them are a complete mystery


So, since the jibbers blade is the thing everyone is most concerned about(with good reason), let's go over it. The quest is called A Legend Revisited, it has two named items: the drowned priest's torch and the coffin nail. Let's say you play the quest on Epic Elite, when you open the chest the game does some calculations and decides whether or not you got "a named item period". If it decides you got one, it THEN does a separate roll to determine which item on the loot table you got. That loot table, again, consists of the drowned priest's torch and the coffin nail, supposedly weighted equally.

Anything that's not a named piece of equipment that can possibly appear in that chest(such as the Jibbers blade) is not on that same loot table and does not conform to the above mechanics and we have no idea what the drop rates are supposed to be. Yes, the jibbers blade is "rarer" than the coffin nail, because its not on the same loot table. The same is true for Star Fragments in Saltmarsh and the Nebula Fragment. I don't have a complete list, but the general rule seems to be "If you can't attach it to a character, its not on that 10/15/33% formula".




Yeah , where is this all items drop equally quote ?

I would love to help but I sincerely do not have time to go digging through the last 5 years of dev tracker and linking you to the ten different threads I gathered my information from. If you don't believe me, fine, don't believe me; go make up your own theories. But this is what I assembled out of many years following the devs around whenever they talk about item drop rates.

[boredGamer]

I'm going to tell you both what I have pieced together from years and years of dev quotes:

*There is a 10/15/33% chance for an item to drop based on quest difficulty
*This rate only applies to named equip items
*Everything else is on a separate loot table and the drop rates for them are a complete mystery


So, since the jibbers blade is the thing everyone is most concerned about(with good reason), let's go over it. The quest is called A Legend Revisited, it has two named items: the drowned priest's torch and the coffin nail. Let's say you play the quest on Epic Elite, when you open the chest the game does some calculations and decides whether or not you got "a named item period". If it decides you got one, it THEN does a separate roll to determine which item on the loot table you got. That loot table, again, consists of the drowned priest's torch and the coffin nail, supposedly weighted equally.

Anything that's not a named piece of equipment that can possibly appear in that chest(such as the Jibbers blade) is not on that same loot table and does not conform to the above mechanics and we have no idea what the drop rates are supposed to be. Yes, the jibbers blade is "rarer" than the coffin nail, because its not on the same loot table. The same is true for Star Fragments in Saltmarsh and the Nebula Fragment. I don't have a complete list, but the general rule seems to be "If you can't attach it to a character, its not on that 10/15/33% formula".




I would love to help but I sincerely do not have time to go digging through the last 5 years of dev tracker and linking you to the ten different threads I gathered my information from. If you don't believe me, fine, don't believe me; go make up your own theories. But this is what I assembled out of many years following the devs around whenever they talk about item drop rates.

Well currently based on zero evidence I don’t believe you.

[Oliphant]

If you roll heads 10 times in a row, one can say its a "small sample size". I'd say it's a 1 in 2^10 chance if the coin is fair, less than 1/10th of 1% chance.

[Eantarus]

If you roll heads 10 times in a row, one can say its a "small sample size". I'd say it's a 1 in 2^10 chance if the coin is fair, less than 1/10th of 1% chance.

And if you could take out a coin and roll heads 10 times in a row, consistently, every day, every time you took out the coin, you'd say say there was something wrong with that coin.

[boredGamer]

And if you could take out a coin and roll heads 10 times in a row, consistently, every day, every time you took out the coin, you'd say say there was something wrong with that coin.

And yet, no numbers have been given. Just vague statements, which all seem to contradict each other. Even this "weighted 30-70" data - where is the data or the summary of the data? It was supposed to be covered in this thread and I don't see it at all.

Very curious what the actual deviation of percentages were, even outside of any ordering.

[Weemadarthur]

And yet, no numbers have been given. Just vague statements, which all seem to contradict each other. Even this "weighted 30-70" data - where is the data or the summary of the data? It was supposed to be covered in this thread and I don't see it at all.

Very curious what the actual deviation of percentages were, even outside of any ordering.

On page 2 I put up a brief overview that covers both why I was testing and the results I got from those tests. I know it's not comprehensive (I do say as much in the post) but the figures are there as overall percentages. Now as you keep demanding proof for what can't actually be proven do you actually have anything to back up your own claims? Demanding proof is fine if you can supply your own to act as a point of reference. If you have nothing then sorry but your opinion is just hot air. You are just claiming everyone else is wrong based on your own personal bias. How is that any better than what you keep dismissing?

[Mofus]

Keep at it, you will get it eventually.

[boredGamer]

On page 2 I put up a brief overview that covers both why I was testing and the results I got from those tests. I know it's not comprehensive (I do say as much in the post) but the figures are there as overall percentages. Now as you keep demanding proof for what can't actually be proven do you actually have anything to back up your own claims? Demanding proof is fine if you can supply your own to act as a point of reference. If you have nothing then sorry but your opinion is just hot air. You are just claiming everyone else is wrong based on your own personal bias. How is that any better than what you keep dismissing?

I’m claiming nothing . I am looking at many claims that all seem to conflict with each other. The burden of proof is on people claiming ****, not people being skeptical. You all sound like ghost hunters (*i* don’t have to prove it, YOU prove there ARENT ghosts!)

I’d be much more curious as to the percents by 10% of your 100 runs, but even with the numbers given - unless there are vastly different things at play here :

- DD on login is it’s own ghost (it doesn’t happen for me so this one is just player id driven?)
- loot pulls are their own ghost (18% discrepancy towards 25 - 75 even if true doesn’t describe behavior seen with loot pulls, I would anecdotally agree with non equal named loot drops but I thought they weren’t supposed to be equal)
- always roll a 3 on login which seems the most obvious lie / red herring
- general d100 not giving random results (this seems the most likely and hardest to prove )
- failed TP (this one I anecdotally agree with , f this mechanic). But again, it’s it’s own rng mechanic ?

So you gave some numbers, but they don’t explain or they even contradict the other theories, and those people refuse to give any numbers.

[Weemadarthur]

I’m claiming nothing . I am looking at many claims that all seem to conflict with each other. The burden of proof is on people claiming ****, not people being skeptical. You all sound like ghost hunters (*i* don’t have to prove it, YOU prove there ARENT ghosts!)

I’d be much more curious as to the percents by 10% of your 100 runs, but even with the numbers given - unless there are vastly different things at play here :

- DD on login is it’s own ghost (it doesn’t happen for me so this one is just player id driven?)
- loot pulls are their own ghost (18% discrepancy towards 25 - 75 even if true doesn’t describe behavior seen with loot pulls, I would anecdotally agree with non equal named loot drops but I thought they weren’t supposed to be equal)
- always roll a 3 on login which seems the most obvious lie / red herring
- general d100 not giving random results (this seems the most likely and hardest to prove )
- failed TP (this one I anecdotally agree with , f this mechanic). But again, it’s it’s own rng mechanic ?

So you gave some numbers, but they don’t explain or they even contradict the other theories, and those people refuse to give any numbers.

The numbers I gave were the numbers I have. I am not trying to prove anything but just gave the data I had at hand. The theory I came up with was just that a theory. I will say (again) that it may not be right. The data I supplied is not there to act as proof that what I think is right so if you don't agree no problem your entitled to your opinion.

There is however a big difference between doing your own test on something that can easily replicated and trying to prove/disprove ghosts exist.

Now as to your list here are a few theories that I have come up with to explain some of this behavior:-

DD on login - I think this has to do with doing a roll before you have fully connected to the server. I can only go with what I experience on this one but every time I 1st log in and do my DD straight away there is always a lag effect between clicking the dice roll and getting my result (this is not the case for any further rolls) and those results are always without exception low rolls. This would explain why if I do a couple of d100 rolls 1st though it doesn't happen (enough time has passed that I am now fully connected).

Loot rolls - If as I believe the dice are weighted toward the center numbers this would mean that if the rolls are against a static table that those items nearer the center would drop more often than those toward the edges. Combine that with observational bias and you get a result where players feel they are dropping one item much more often over time. The clumping forms a bias that is then reinforced by the clumping. The clumping wouldn't have to be that major to start that cycle just enough to be noticable.

Always roll a 3 on log in - Nope soz got nothing for this one. I won't say it doesn't happen but just I can't think of a logical reason that it would.

General d100 results not giving random results - At best guess I think this is caused by the random generator trying to maintain a 50 average roll.

Failed scroll usage - this could be one of two things. 1) SSG/Turbine put this in intentionally to sell more scrolls or 2) pure observational bias.

Now all of the above are just theories but they are all plausible. Whether you agree with them or not is irrelevant as they will all be impossible to prove without looking at the actual code to see how the game works.

[boredGamer]

The numbers I gave were the numbers I have. I am not trying to prove anything but just gave the data I had at hand. The theory I came up with was just that a theory. I will say (again) that it may not be right. The data I supplied is not there to act as proof that what I think is right so if you don't agree no problem your entitled to your opinion.

There is however a big difference between doing your own test on something that can easily replicated and trying to prove/disprove ghosts exist.

Now as to your list here are a few theories that I have come up with to explain some of this behavior:-

DD on login - I think this has to do with doing a roll before you have fully connected to the server. I can only go with what I experience on this one but every time I 1st log in and do my DD straight away there is always a lag effect between clicking the dice roll and getting my result (this is not the case for any further rolls) and those results are always without exception low rolls. This would explain why if I do a couple of d100 rolls 1st though it doesn't happen (enough time has passed that I am now fully connected).

Loot rolls - If as I believe the dice are weighted toward the center numbers this would mean that if the rolls are against a static table that those items nearer the center would drop more often than those toward the edges. Combine that with observational bias and you get a result where players feel they are dropping one item much more often over time. The clumping forms a bias that is then reinforced by the clumping. The clumping wouldn't have to be that major to start that cycle just enough to be noticable.

Always roll a 3 on log in - Nope soz got nothing for this one. I won't say it doesn't happen but just I can't think of a logical reason that it would.

General d100 results not giving random results - At best guess I think this is caused by the random generator trying to maintain a 50 average roll.

Failed scroll usage - this could be one of two things. 1) SSG/Turbine put this in intentionally to sell more scrolls or 2) pure observational bias.

Now all of the above are just theories but they are all plausible. Whether you agree with them or not is irrelevant as they will all be impossible to prove without looking at the actual code to see how the game works.


The clumping thing doesn't explain what seems to be 15:2 type differences in various named loots. I will personally be shocked if indeed loot tables are supposed to be equal chance, and are actually coded that way. It makes more sense to me they are not coded that way than the RNG is doing it.

I'll put my interest where my mouth is - I'll record my DD on immediate login (so far 1 value, 48).

Then 100 d100 rolls, they will be immediately rolled together (script) - if it seems clumpy I might do more dispersed values, but at first glance my first set seems distinctly UNclumpy.

So far:
1-10 7
11-20 13
21-30 12
31-40 9
41-50 13
51-60 12
61-70 9
71-80 3 <-- WEIGHTED DICE ;)
81-90 9
91-100 13

Repeats: 2 (same value one after another)

[Weemadarthur]

The clumping thing doesn't explain what seems to be 15:2 type differences in various named loots. I will personally be shocked if indeed loot tables are supposed to be equal chance, and are actually coded that way. It makes more sense to me they are not coded that way than the RNG is doing it.

I'll put my interest where my mouth is - I'll record my DD on immediate login (so far 1 value, 48).

Then 100 d100 rolls, they will be immediately rolled together (script) - if it seems clumpy I might do more dispersed values, but at first glance my first set seems distinctly UNclumpy.

So far:
1-10 7
11-20 13
21-30 12
31-40 9
41-50 13
51-60 12
61-70 9
71-80 3 <-- WEIGHTED DICE ;)
81-90 9
91-100 13

Repeats: 2 (same value one after another)

I suppose that even with your own rolls the fact that the center 20 has 25% of the rolls compared to 21% being the highest of any other 20 digit range doesn't at all fit with what I predicted you would get lol. Break it up into the same 25 digit splits that I used and you would more than likely have had very similar results to those I posted earlier in the thread. Still that's just coincidence right?

[boredGamer]

I suppose that even with your own rolls the fact that the center 20 has 25% of the rolls compared to 21% being the highest of any other 20 digit range doesn't at all fit with what I predicted you would get lol. Break it up into the same 25 digit splits that I used and you would more than likely have had very similar results to those I posted earlier in the thread. Still that's just coincidence right?

I’ll leave that to future rolls . Will update some thread at some point. Just based on those rolls it “looks” fine . I’d be very curious from devs if loot distribution is supposed to be even on that point .

I’d also love to see someone’s actual first dice rolls over say a month that think they are cursed, because I know my last 4 days are all over the map, but I didn’t start recording until today.

[Eantarus]

And yet, no numbers have been given. Just vague statements, which all seem to contradict each other. Even this "weighted 30-70" data - where is the data or the summary of the data? It was supposed to be covered in this thread and I don't see it at all.

Very curious what the actual deviation of percentages were, even outside of any ordering.

Numbers have been posted many times, both in this thread and many others. I hate it when people are too lazy to be educated on a problem and I have to do it for them. Well, here we go.


Check the top page of this thread:

https://forums.ddo.com/forums/showthread.php/513124-Spark-of-memory-drop-rate-is-lacking/page2?highlight=drop+rates

Here are the relevant quotes:


So can you confirm that the drop rate of Echo of Blackrazor is the same as Echo of Whelm, that they are not weighted differently on the loot table?


They are weighted the same.

This is only the first incidence of this confirmation I was able to find within 4 minutes of searching. I know there are 10 other mentions of it since 2017.



Here's another thread with someone posting drop rates:

https://forums.ddo.com/forums/showthread.php/532111-Is-this-a-normal-number-of-Dream-s-Edges?highlight=systemshaker1941


I wanted to get a Terror from the Mind Sunderer quest, so I spent the evening farming for it. Along the way I picked up:

8x Dream's Edge
3x Rahl's Might
3x Lucid Dreams
2x Light and Darkness
2x Elocator's Habilment
2x Death's Touch

And only 1 each of Terror and Gloves of Titan's Grip.

These figures do not include all of the items I rolled away or the fact that Terror did not appear in 7 "third time completion" lists. If all the items are weighted equally, then I'm the king of Siam.


Now here's the real rub: this kind of "random non-randomness" happens Every. Single. Time. I go out farming for a specific item. Without fail. And I've found a half-dozen other threads with people complaining of similar behavior.

Why is this problem being ignored?

You can see a very clear clump in there.

Several pages long thread with many numbers:

https://forums.ddo.com/forums/showthread.php/527645-Plea-to-increase-named-item-drop-rates-or-change-the-reward-system-completely?highlight=family+recruit+sigil+drop+rat e

There are numbers literally everywhere if you just pay attention or take 5 entire minutes to look for yourself.

[boredGamer]

Numbers have been posted many times, both in this thread and many others. I hate it when people are too lazy to be educated on a problem and I have to do it for them. Well, here we go.


Check the top page of this thread:

https://forums.ddo.com/forums/showthread.php/513124-Spark-of-memory-drop-rate-is-lacking/page2?highlight=drop+rates

Here are the relevant quotes:





This is only the first incidence of this confirmation I was able to find within 4 minutes of searching. I know there are 10 other mentions of it since 2017.



Here's another thread with someone posting drop rates:

https://forums.ddo.com/forums/showthread.php/532111-Is-this-a-normal-number-of-Dream-s-Edges?highlight=systemshaker1941



You can see a very clear clump in there.

Several pages long thread with many numbers:

https://forums.ddo.com/forums/showthread.php/527645-Plea-to-increase-named-item-drop-rates-or-change-the-reward-system-completely?highlight=family+recruit+sigil+drop+rat e

There are numbers literally everywhere if you just pay attention or take 5 entire minutes to look for yourself.

Your responses aren’t even worth responding to at this point.

8332211 isn’t a crazy distribution.

The quote is about two items having equal distribution, not all , and my anecdotal evidence of those two items agrees. They drop roughly equally.

Two posts later the dev says the super checked into it and it’s WAI. So you just don’t believe that post ?

[Eantarus]

Two posts later the dev says the super checked into it and it’s WAI. So you just don’t believe that post ?

I do not, because it is clearly not working correctly. I believe the devs are basing their data on server-wide metrics which, averaged over tens of thousands of drop rates do probably work out ok. The problem is individual players do not see tens of thousands of drops. We only see 7 chest pulls.

Plus. And. Again. As I have explained. Several times. If its a bug negatively impacting only a small percentage of the player base, that could easily not come up in their data if they are only looking at big picture. Go read about the Wi Flag in Asheron's Call(from which DDO gets most of it's code). Or do I have to spoon-feed you all of that information, too?

[boredGamer]

I do not, because it is clearly not working correctly. I believe the devs are basing their data on server-wide metrics which, averaged over tens of thousands of drop rates do probably work out ok. The problem is individual players do not see tens of thousands of drops. We only see 7 chest pulls.

Plus. And. Again. As I have explained. Several times. If its a bug negatively impacting only a small percentage of the player base, that could easily not come up in their data if they are only looking at big picture. Go read about the Wi Flag in Asheron's Call(from which DDO gets most of it's code). Or do I have to spoon-feed you all of that information, too?

Yeah . I know about it, thanks.

But to be clear now, you’re only talking loot roll anomalies, not your dd login anomalies, right ?

Because one is easily testable and you keep avoiding it.

The other I still don’t have a clear quote on the intent, even though you keep saying it’s everywhere.

[Eantarus]

But to be clear now, you’re only talking loot roll anomalies, not your dd login anomalies, right ?

I've talked about both of them over the course of this thread. I apologize if it came across like I was claiming the two were related. Loot roll anomalies and DD anomalies are two unrelated things.



Because one is easily testable and you keep avoiding it.

Which one? I've tested both many times and found anomalies with both. I see loot role anomalies every single time I farm for items. If I were indeed to do my dd role upon login every day, I would continue to see the anomalies I was seeing when I did that.

And for what it's worth: the /roll 100 does not appear to be using the same backend mechanic as whatever controls the daily dice roles.



The other I still don’t have a clear quote on the intent, even though you keep saying it’s everywhere.

So let me see if I follow your logic. You asked how it works. Based on my own exhaustive research I pieced an answer together and gave it to you. Are you saying that answer isn't good enough because I had to piece it together? Well, sorry bukaroo; the devs have never made a single post completely outlining the loot system. The fact that's its so obfuscated is WHY we have all of these endless threads trying to work it out. Other games simply publish loot tables and drop rates. SSG deliberately keeps us in the dark.

If my explanation is not good enough for you, then believe whatever you want. I don't care enough about you to explain any further. Go do the research yourself or put forward whatever stupid theories you want to believe. I don't care anymore.

[boredGamer]

So let me see if I follow your logic. You asked how it works. Based on my own exhaustive research I pieced an answer together and gave it to you. Are you saying that answer isn't good enough because I had to piece it together? Well, sorry bukaroo; the devs have never made a single post completely outlining the loot system. The fact that's its so obfuscated is WHY we have all of these endless threads trying to work it out. Other games simply publish loot tables and drop rates. SSG deliberately keeps us in the dark.

If my explanation is not good enough for you, then believe whatever you want. I don't care enough about you to explain any further. Go do the research yourself or put forward whatever stupid theories you want to believe. I don't care anymore.

Let ME see if I understand. You told everyone the devs said all named had an equal chance. But it turns out your “exhaustive research”doesn’t back up what you’re saying, and that you made it up.

And you want ME to go do research to refute a thing you made up ?

I’m trying hard to understand why anyone would take anything you say seriously at this point.