Howdy, Stranger!
Categories
- 3.7K All Categories
- 3K The Midnight Age
- 757 Announce posts
- 125 Newbies
- 6 Class Guides
- 804 Harpy's Head Tavern
- 57 Echo Canyon
- 310 Self-Affirmations
- 163 Town Crier
- 135 The Exchange
- 167 Sparring Grounds
- Combat Guides
- 30 Combat Logs
- 188 Idea Box
- 9 Classleads
- 581 Roleplay Logs
- 107 Aetolia Development
- 5 The Void
- 744 Miscellaneous
- 9 Announcements
- 417 OOC Chat
- 217 Tech Talk
- 47 Scripts
Looking for more active discussion? Join our Discord at https://discord.gg/x2s7fY6
Promos: Buy Be Aware?
This has been stewing about in my brain for a bit, and ultimately I figured I would create this thread, potentially solicit others for any data they may have, or their experiences. Also, wall of text warning. Cool.
There is a kind of problem with promotions. They're broken. Responses to bringing up data-driven analyses of the promotions compared against the documentation has been less than satisfactory. So lets start at the beginning.
Last month when cryptic chests were the only promotion, I jumped onboard. Woohoo! Who doesn't love cryptic chests? I cracked open 50 of those bad boys and noticed a few things off the bat:
- There was no gold awarded
- There were no broken pieces of chests awarded (Stated odds of 10%, more later)
- The minipet list was wrong
- The relic list was wrong
- The secondary prize list was a mess. Notably, no ylem amulets.
- All. The. Fireworks.
A few of these were worded such that they were ambiguous enough that they're mostly a minor annoyance. The secondary prize list specifically was generic and said something to the effect of, "The bag could contain things such as:" similar to the wording as is. But some of the items simply didn't show. Looking at you, ylem amulets.
Now, I'm familiar enough with statistics to know that while n=50 is good, n=100 is better! I reached out to someone via Discord to ask about very specific items, such as gold, ylem amulets, and broken chest pieces. They confirmed they received none of those items. The odds of getting no chest pieces in 50 chests at the 10% stated odds is just north of half a percent. 0.5%. Alright, that's not unheard of. None out of 100? 0.0027%. Tiur did confirm via e-mail that the pieces were broken. Alright, progress!
Gold has also been removed from the documentation as a possibility, as well as ylem amulets and the relic list and minipet lists are now probably correct. I haven't bothered to buy anymore chests because fireworks. Seriously, I don't need 27 chests worth of fireworks. However, on the website where the purchase is made, gold is still listed as a possible prize. This is at best a disservice to anyone in the unlikely scenario of being an actual newbie and thinking they might be able to get gold and chests are clearly the wrong avenue. I realize the probability of having an actual newbie is also pretty slim, but really the documentation in all areas needs to be correct on these things.
Lets fast forward to the present. Elemental coffers. Initially, the documentation stated that the default element would be fire, something about wanting to burn the world? Yeah, no, it was random. I'm not the only one to have wasted at least one coffer opening to get earth or some other equally useless element. Why does this matter? The everburning torch relic pieces are exclusive to fire. A little math exercise, using an arbitrary drop rate of 'x' would reveal that in reality the probability of getting the everburning torch relic pieces is at best 1/6th of the advertised: if I open all of my chests and they were fire, as advertised, I'd have some probability x. With it being random with an equal chance for any element, we're at roughly x/6. This is cutting edge stuff, I know. Hand wavy math, tada.
But Wjoltyr, you said you had data! Give me the data!
Look, coffers are an interesting beast. As best as I can discern from how they work, you start with a roll. You get some item. If the value of that item is less than 20 credits, you get another roll. If the combined value of those items is less than 20 credits, you get a third roll. If the value of those three items combined is less than 20 credits, you get a fourth roll. The minimum value for any piece is 5 credits, so 4 rolls is the most you're going to get. Maybe. It could be 25, it could be 30. But you hit some threshhold, you're done.
And the prize pools are different for each element, so it's really hard to get a meaningful data set from which to figure out odds on anything. Also, I opened a bunch on my phone and didn't have anywhere to record what I got. But lets talk qualitative.
90 coffers. Most shaken to get either fire or water, otherwise shaken 3 times and I got what I got.
Out of the first 70, I got 2 relic pieces -- the femur and the strap, so 1 from each the fire and water coffers. The next 10 yielded 3 or 4 pieces and were just a freaking anomaly in terms of what I received. For a moment I had hope that maybe some stuff got switched behind the scenes after some attention was drawn to it. Naw, the next 10 was the same nonsense as the first 70.
I have issued myself over this asking what the specific drop rate is for the relics and was told that's confidential, and coffers are currently spitting out higher expected values than intended, but they wouldn't reduce it in the middle of a promo. Cool, I respect that.
But how many pity prizes have you gotten, Wjoltyr? Because everyone loves salvage and safety mechanisms!
8.
Whoa. Well, what about morphic metals?
I legit have lost count, I still have 36 after using some, selling some, and flat out giving some away.
Others have had similar experiences with an abundance of morphic metals and less than an abundance of the stated relic pieces. And while I get that relics sell coffers (or crates and crowbars!), we're looking at these relics being some extremely expensive relics.
'But Wjoltyr, these things are random. You can't expect to win the lottery every time!' I won the salvage lotto 8 times. Out of 90 coffers. Things. Are. Broken. Some salvage was given out without the fanfare of trombones, further evidence that things are broken. While the total value may be in my favor if we dole it out into credit values, the biggest problem is that the pieces just don't exist. I can't trade a salvage to someone for the relic piece I need if it just doesn't exist. Remember when sip rings were giving out 11%/12%/13% instead of the expected 10%/20%/30%? And nobody noticed for years? How am I, the consumer, supposed to have confidence that the drop rates are what they're supposed to be, and nobody accidentally added a 0 in the wrong place, or shifted the decimal place, and instead of 5% we're at 0.5%?
I really, really don't want to be the guy whining he didn't get what he wanted out of the promotions. Because I get randomness. But I believe things are fundamentally broken, and when money is changing hands, things ought to be not broken and the consumer should get what they were promised and not a false bill of goods, or false odds in the event of loot boxes. If I buy a pack of baseball cards and get NASCAR instead, I'll be pissed regardless of how good those NASCAR cards are.
Now, I hate being the guy who just makes problems. I like solutions, not problems. So lets make some suggestions.
1. Documentation needs to be correct from the onset. Transparency goes a long way. Give us actual drop rate %'s. I know Tiur has done this on gift bags just a few months ago. If we're got a big enough n, we can validate those odds and all will be right with the world.
2. Do what you say you're going to do. If you say you're going to give out replacement broken chest pieces because the documentation was not reflective of reality, do that. Don't say you will and then don't. While 'making it right' is probably good customer service, it just feeds into the problem when you say you will but then don't.
3. Respond to e-mails about promos. I get that life is busy, and you get a lot of e-mails, but I would argue that promotions are basically only a bit below something like allegations of sexual harassment. I'm sure someone will jump on me and point out they're not in the same league, yeah, I got it, the point being yes, respond to safety concerns, then financial stuff. Proper prioritization is key.
I hope that my intentions here are clear. I'd like to improve things going forward. If I were to quantify something akin to a Consumer Confidence Index, it would be pretty low right now for me when it comes to promos. Let's get that higher.
If you have your own data you'd like to share, you can message me here or in game.
There is a kind of problem with promotions. They're broken. Responses to bringing up data-driven analyses of the promotions compared against the documentation has been less than satisfactory. So lets start at the beginning.
Last month when cryptic chests were the only promotion, I jumped onboard. Woohoo! Who doesn't love cryptic chests? I cracked open 50 of those bad boys and noticed a few things off the bat:
- There was no gold awarded
- There were no broken pieces of chests awarded (Stated odds of 10%, more later)
- The minipet list was wrong
- The relic list was wrong
- The secondary prize list was a mess. Notably, no ylem amulets.
- All. The. Fireworks.
A few of these were worded such that they were ambiguous enough that they're mostly a minor annoyance. The secondary prize list specifically was generic and said something to the effect of, "The bag could contain things such as:" similar to the wording as is. But some of the items simply didn't show. Looking at you, ylem amulets.
Now, I'm familiar enough with statistics to know that while n=50 is good, n=100 is better! I reached out to someone via Discord to ask about very specific items, such as gold, ylem amulets, and broken chest pieces. They confirmed they received none of those items. The odds of getting no chest pieces in 50 chests at the 10% stated odds is just north of half a percent. 0.5%. Alright, that's not unheard of. None out of 100? 0.0027%. Tiur did confirm via e-mail that the pieces were broken. Alright, progress!
Gold has also been removed from the documentation as a possibility, as well as ylem amulets and the relic list and minipet lists are now probably correct. I haven't bothered to buy anymore chests because fireworks. Seriously, I don't need 27 chests worth of fireworks. However, on the website where the purchase is made, gold is still listed as a possible prize. This is at best a disservice to anyone in the unlikely scenario of being an actual newbie and thinking they might be able to get gold and chests are clearly the wrong avenue. I realize the probability of having an actual newbie is also pretty slim, but really the documentation in all areas needs to be correct on these things.
Lets fast forward to the present. Elemental coffers. Initially, the documentation stated that the default element would be fire, something about wanting to burn the world? Yeah, no, it was random. I'm not the only one to have wasted at least one coffer opening to get earth or some other equally useless element. Why does this matter? The everburning torch relic pieces are exclusive to fire. A little math exercise, using an arbitrary drop rate of 'x' would reveal that in reality the probability of getting the everburning torch relic pieces is at best 1/6th of the advertised: if I open all of my chests and they were fire, as advertised, I'd have some probability x. With it being random with an equal chance for any element, we're at roughly x/6. This is cutting edge stuff, I know. Hand wavy math, tada.
But Wjoltyr, you said you had data! Give me the data!
Look, coffers are an interesting beast. As best as I can discern from how they work, you start with a roll. You get some item. If the value of that item is less than 20 credits, you get another roll. If the combined value of those items is less than 20 credits, you get a third roll. If the value of those three items combined is less than 20 credits, you get a fourth roll. The minimum value for any piece is 5 credits, so 4 rolls is the most you're going to get. Maybe. It could be 25, it could be 30. But you hit some threshhold, you're done.
And the prize pools are different for each element, so it's really hard to get a meaningful data set from which to figure out odds on anything. Also, I opened a bunch on my phone and didn't have anywhere to record what I got. But lets talk qualitative.
90 coffers. Most shaken to get either fire or water, otherwise shaken 3 times and I got what I got.
Out of the first 70, I got 2 relic pieces -- the femur and the strap, so 1 from each the fire and water coffers. The next 10 yielded 3 or 4 pieces and were just a freaking anomaly in terms of what I received. For a moment I had hope that maybe some stuff got switched behind the scenes after some attention was drawn to it. Naw, the next 10 was the same nonsense as the first 70.
I have issued myself over this asking what the specific drop rate is for the relics and was told that's confidential, and coffers are currently spitting out higher expected values than intended, but they wouldn't reduce it in the middle of a promo. Cool, I respect that.
But how many pity prizes have you gotten, Wjoltyr? Because everyone loves salvage and safety mechanisms!
8.
Whoa. Well, what about morphic metals?
I legit have lost count, I still have 36 after using some, selling some, and flat out giving some away.
Others have had similar experiences with an abundance of morphic metals and less than an abundance of the stated relic pieces. And while I get that relics sell coffers (or crates and crowbars!), we're looking at these relics being some extremely expensive relics.
'But Wjoltyr, these things are random. You can't expect to win the lottery every time!' I won the salvage lotto 8 times. Out of 90 coffers. Things. Are. Broken. Some salvage was given out without the fanfare of trombones, further evidence that things are broken. While the total value may be in my favor if we dole it out into credit values, the biggest problem is that the pieces just don't exist. I can't trade a salvage to someone for the relic piece I need if it just doesn't exist. Remember when sip rings were giving out 11%/12%/13% instead of the expected 10%/20%/30%? And nobody noticed for years? How am I, the consumer, supposed to have confidence that the drop rates are what they're supposed to be, and nobody accidentally added a 0 in the wrong place, or shifted the decimal place, and instead of 5% we're at 0.5%?
I really, really don't want to be the guy whining he didn't get what he wanted out of the promotions. Because I get randomness. But I believe things are fundamentally broken, and when money is changing hands, things ought to be not broken and the consumer should get what they were promised and not a false bill of goods, or false odds in the event of loot boxes. If I buy a pack of baseball cards and get NASCAR instead, I'll be pissed regardless of how good those NASCAR cards are.
Now, I hate being the guy who just makes problems. I like solutions, not problems. So lets make some suggestions.
1. Documentation needs to be correct from the onset. Transparency goes a long way. Give us actual drop rate %'s. I know Tiur has done this on gift bags just a few months ago. If we're got a big enough n, we can validate those odds and all will be right with the world.
2. Do what you say you're going to do. If you say you're going to give out replacement broken chest pieces because the documentation was not reflective of reality, do that. Don't say you will and then don't. While 'making it right' is probably good customer service, it just feeds into the problem when you say you will but then don't.
3. Respond to e-mails about promos. I get that life is busy, and you get a lot of e-mails, but I would argue that promotions are basically only a bit below something like allegations of sexual harassment. I'm sure someone will jump on me and point out they're not in the same league, yeah, I got it, the point being yes, respond to safety concerns, then financial stuff. Proper prioritization is key.
I hope that my intentions here are clear. I'd like to improve things going forward. If I were to quantify something akin to a Consumer Confidence Index, it would be pretty low right now for me when it comes to promos. Let's get that higher.
If you have your own data you'd like to share, you can message me here or in game.
1
Comments
(Web): Abhorash says, "Nerds."
(Web): Abhorash has left your web.
Alela's Affirmations
1. I usually copy over a main announce post to be the helpfile. Obviously I missed my notes here and there, and didn't see that I adjusted things in subsequent posts. Like the chance of chest pieces, that was removed a ways back and I accidentally copied it over. Which is definitely my mistake to fix! Why haven't I given some out? Because you can still buy chests and their contents are changing. I'll hand out pieces when they're not in the shop, lest I get some angry about giving pieces out THEN changing contents, etc. It's happened before, alas.
1a. Full drop rates on things... I am working on it. I've pushed from completely opaque to slightly transparent, to sometimes perfect. Basically you're all guinea pigs for my data going to IRE to show that it is better to show odds than hide. Note this month, while silly, provides nearly 1:1 comparisons of preference in promo type.
1b. Also, when I show full odds, I have to explain them. I know that sounds silly, but it's like... with some transparency comes a need for full transparency. I'm happy to explain when someone asks me why a giftbag shows a >100% chance to roll something, but I end up doing it so often that I wish I never showed it.
2. I will! I just usually try to, like I said above, wait until the item is off the table and static for a while.
3. Sorry I haven't replied! I have no idea, I likely either thought I replied, or wrote a reply and it's sitting in my notepad++ or blech.
Rapture And Math: Rapture prefers integers. Divide 5 by 1000 and it'll get .005 fine, but if you do subsequent math without protecting it (like by passing it to another function), it'll round somewhere and bad things will happen. Most of the time we know how to prevent this (multiply by 1000 and remove it later, so it never gets to decimals) but it's not ideal. That means that sometimes something the game knows is 95% might SHOW as 100% on a display. Like Audit might have rounding errors hiding it. They're almost universally on the display, because the number is calculated in some function, then passed to a display function.
Now, other things, like the pity rewards. I rewrote how Aetolia gives prizes from the bottom up, with checks for data types and protections. Sip rings can have an error from a decade ago and I'd have no idea, but I MADE the prize system. Which is not to say I'm immune to errors, but to say that I wrote everything in such a way that the errors cannot exist. Rapture rounds to integers... so never let it use anything less than a whole number. Those sorts of things.
I'll show a demonstration in another post. It's getting really long and I need the space.
The core of it is trust. Showing percentages makes more questions I have to answer, and largely it's not something IRE does (I need permission when I do it). I worked/work really hard to make promotions fair, but the only proof you have is if you trust me that it is what I say it is.
As to the broken coffers, it's just a misunderstanding about the workings of coffers, so I messaged you the explanation instead of saying it here. I won't air your private spending things like promo results, but you can if you want
I would fix it, but I'd also have to take away all the salvage and other things to start from 0!