Excuse me but were Mc Tusk drop rates silently nerfed?

[ArubaroBeefalo]

51 minutes ago, Vinja said:

Objectively false.

i have check in case there were new ways of generating numbers and seems like for some stuff are being used atmospheric values to create, what some people claim, trully random numbers so, yeah i was wrong.

Anyways, dst uses algorithms so they are deterministic which means that arent trully random

57 minutes ago, Cheggf said:

"True randomness" isn't real. It is impossible to prove the existence of it

empty comment that has nothing to do with what i said

you can prove that something isnt random

[SapoLover]

15 hours ago, Wonz said:

RNG is never good, you could never get any tusk despite 50% being relatively high chance. I've went 2 winters without a tusk once, nobody supported me saying RNG should be gone

I agree with you. RNG is okay for the world generation, but idk where's the fun part in getting out of your way to kill the same mob over and over again for the chance of getting or not getting the thing you want.

Beefalos and Volt Goats are okay because there are many of them in one or multiple places, but the world  generate or only one McTusk in the moon forest, or four of them if you get the Mctusk biome. (edit: I forgot to mention volt goats and beefalos can be relocated wherever you want)

Spoiler

Don't bother in telling me there are other alternatives of speed bonus because I already use them. I'm tired of praying for good RNG.

 

[Yuuko]

3 hours ago, DeadWhereX said:

Beefalos and Volt Goats are okay because there are many of them in one or multiple places, but the world  generate or only one McTusk in the moon forest, or four of them if you get the Mctusk biome.

  Reveal hidden contents

Don't bother in telling me there are other alternatives of speed bonus because I already use them. I'm tired of praying for good RNG.

 

That's fair - one issue with Mac Tusks is that you have either 1 or 4 in a world, and even with 4 and 50% drop rate, with all the things to do in winter its easy to make it through with 0-2 walrus tusks.  I don't think adding more is a good option, or a super powered version - but maybe the drop rate should be more like 60-75% ?  I don't think it needs to (or should) be 100%  Some drops should be more rare and that's okay.

[Cheggf]

7 hours ago, mr. brj said:

Pseudo RNG is 100% reliable and in my opinion the drops of some mobs in DST, such as the walruses and the goats, could use some pseudo RNG. It still uses true randomness, but manipulates the last possible outcome(s) if the previous outcomes didn't match the desired probability.

Let's take tam's 25% drop chance as an example, and let's say the game checks for 4 outcomes. Then for every 4 consecutive walrus kills, you are guaranteed to get exactly 1 tam, no more and no less. This basically means that if you killed 3 walruses and got no tams, the next walrus is going to drop a tam 100% of the time. If the game checked for 8 outcomes and you didn't get any tams in your first 6 kills, then the last 2 walruses are going to drop a tam 100% of the time, and so on.

That's bad luck protection, not pseudo rng.

7 hours ago, ArubaroBeefalo said:

empty comment that has nothing to do with what i said

you can prove that something isnt random

You're right, you can prove that everything isn't random. A coin flip is determined by the force that you flip it with and the air it flies against, it only seems random because it's difficult to control those variables. Randomness doesn't exist.

[Vinja]

1 hour ago, Cheggf said:

You're right, you can prove that everything isn't random. A coin flip is determined by the force that you flip it with and the air it flies against, it only seems random because it's difficult to control those variables. Randomness doesn't exist.

Let’s say we design a program that will select any number from 1-100, with each individual number having an equal chance of being selected, and I then tell the program to give me a 100 number sequence.

I could give the worlds top 100 scientists and mathematicians all the resources on the planet and 50 years to try to come up with some sort of way to accurately predict that specific 100 number sequence and they will never be able to.
 

Even if they do happen to guess correctly that specific number sequence I generated on that first  day after trillions of attempts and countless years of trial and error it would only ever be due to luck, and if I generated a follow up 100 number sequence and asked them to use the formula they used to crack the last one to accurately predict another 100 number sequence it would fail. 
 

This is randomness in action, by human definition.

 This applies to don’t starve together as well. No matter how many tries I give you, if I killed 10,000 mactusks and asked you to find some formula that would accurately predict the outcome of those 10,000 drops you would die long before you ever found a solution. Why? Because the outcome is random and there is no formula that could predict it.

[Cheggf]

3 hours ago, Vinja said:

Let’s say we design a program that will select any number from 1-100, with each individual number having an equal chance of being selected, and I then tell the program to give me a 100 number sequence.

I could give the worlds top 100 scientists and mathematicians all the resources on the planet and 50 years to try to come up with some sort of way to accurately predict that specific 100 number sequence and they will never be able to.
 

Even if they do happen to guess correctly that specific number sequence I generated on that first  day after trillions of attempts and countless years of trial and error it would only ever be due to luck, and if I generated a follow up 100 number sequence and asked them to use the formula they used to crack the last one to accurately predict another 100 number sequence it would fail. 
 

This is randomness in action, by human definition.

 This applies to don’t starve together as well. No matter how many tries I give you, if I killed 10,000 mactusks and asked you to find some formula that would accurately predict the outcome of those 10,000 drops you would die long before you ever found a solution. Why? Because the outcome is random and there is no formula that could predict it.

We are both saying the same thing in a different way. You are saying that Don't Starve is random like real life, while I am saying that if you don't believe Don't Starve is random because technically you can look at the inputs that make it come to that outcome then nothing is random because everything has things that lead them to the outcome. Calling Don't Starve pseudorandom or saying it isn't really random doesn't make any sense, it's just an argument against randomness existing altogether.

If anybody wants to see a good example of actual pseudorandomness look at old games like Doom (1993) which have very simplistic methods of determining randomness. They have randomness that is either so predictable as to be exploitable by someone familiar with it, or have randomness that comes with express limitations that prevent certain things from happening altogether.

[Yuuko]

3 hours ago, Vinja said:

Let’s say we design a program that will select any number from 1-100, with each individual number having an equal chance of being selected, and I then tell the program to give me a 100 number sequence.

I could give the worlds top 100 scientists and mathematicians all the resources on the planet and 50 years to try to come up with some sort of way to accurately predict that specific 100 number sequence and they will never be able to.
 

Even if they do happen to guess correctly that specific number sequence I generated on that first  day after trillions of attempts and countless years of trial and error it would only ever be due to luck, and if I generated a follow up 100 number sequence and asked them to use the formula they used to crack the last one to accurately predict another 100 number sequence it would fail. 
 

This is randomness in action, by human definition.

 This applies to don’t starve together as well. No matter how many tries I give you, if I killed 10,000 mactusks and asked you to find some formula that would accurately predict the outcome of those 10,000 drops you would die long before you ever found a solution. Why? Because the outcome is random and there is no formula that could predict it.

That's not true.  Computers generate "random numbers" through deterministic means.  For someone who is involved in the science of computer rng prediction is absolutely possible.  rng manipulation is core to TAS automations.  Speed runners have been caught cheating because the "rng" they force on their instance is literally, provably not possible because the entire chart of all possible "random numbers" is known.  Given a sample of results to test against the table you can see exactly which "random numbers" you were given, and would know exactly what numbers would occur next.  You could even force your position on the table to receive an exact, predictable set of "random numbers."

Even physics models aren't truly random.  Its harder to predict because there are many subtle parameters but they aren't any more random just because their seed is more obfuscated.  Learn the inputs and you can predict the outputs.

[Cheggf]

8 minutes ago, Shosuko said:

Speed runners have been caught cheating because the "rng" they force on their instance is literally, provably not possible because the entire chart of all possible "random numbers" is known.

In really old games. Newer games are far too complicated to chart out rng like that, and cheaters need to be proven to have manipulated the rng in other ways like just calculating the mathematical odds of something happening then saying that's too unlikely.

[reddocc]

the quantum mechanics of walruses

[Yuuko]

5 hours ago, Cheggf said:

In really old games. Newer games are far too complicated to chart out rng like that, and cheaters need to be proven to have manipulated the rng in other ways like just calculating the mathematical odds of something happening then saying that's too unlikely.

Modern games are still highly predictable, largely because there isn't much reason to make them less predictable.  For most people rng appears random enough so there isn't much incentive to invest in extra dev time and system resources to be any more random.

[Vinja]

12 hours ago, Shosuko said:

That's not true.  Computers generate "random numbers" through deterministic means.  For someone who is involved in the science of computer rng prediction is absolutely possible.  rng manipulation is core to TAS automations.  Speed runners have been caught cheating because the "rng" they force on their instance is literally, provably not possible because the entire chart of all possible "random numbers" is known.  Given a sample of results to test against the table you can see exactly which "random numbers" you were given, and would know exactly what numbers would occur next.  You could even force your position on the table to receive an exact, predictable set of "random numbers."

Even physics models aren't truly random.  Its harder to predict because there are many subtle parameters but they aren't any more random just because their seed is more obfuscated.  Learn the inputs and you can predict the outputs.

In the scenario I mentioned you would never be able to accurately predict the sequence I generate. 
 

Sure, if the numbers aren’t truly randomly generated and they are based on some sort of pattern or algorithm then you could figure out that method and use it to predict what would come next but then you aren’t actually talking about a random method, you’re talking about a system meant to imitate or seem random aka pseudo random. Randomness is objective, not a measurement of one’s ignorance.

Anyway, I’m done talking about this lol. Have a good night everyone.

[mr. brj]

22 hours ago, Cheggf said:

That's bad luck protection, not pseudo rng.

Dang it you're right, I somehow mixed up the two as I was writing my reply. Then let me rephrase: pseudo RNG sucks almost as much as true randomness, and some sort of a bad luck protection for certain situations would be really nice to have.

[Yuuko]

9 hours ago, Vinja said:

In the scenario I mentioned you would never be able to accurately predict the sequence I generate. 
 

Sure, if the numbers aren’t truly randomly generated and they are based on some sort of pattern or algorithm then you could figure out that method and use it to predict what would come next but then you aren’t actually talking about a random method, you’re talking about a system meant to imitate or seem random aka pseudo random. Randomness is objective, not a measurement of one’s ignorance.

Anyway, I’m done talking about this lol. Have a good night everyone.

That's...  That's exactly the point lol  How do you think computers generate random numbers?

Maybe the misconception is that "random numbers" exist and you're setting up a No True Scotsman situation to deflect the idea that anyone could predict or manipulate "true rng."  There are no true "random numbers," only more complicated, obfuscated, and difficult to manipulate algorithms.  Of course the more you turn that up the more resources it takes just to generate random numbers, and for most games that isn't a wise investment.  I've seen companies put in safeguards to prevent manipulation but not prediction.  With data mining that is pretty much a lost cause.  You'd have to host it server side and protect the algorithm to prevent prediction, probably only worth investing in if you were an online casino or something.

[Vinja]

4 hours ago, Shosuko said:

That's...  That's exactly the point lol  How do you think computers generate random numbers?

Maybe the misconception is that "random numbers" exist and you're setting up a No True Scotsman situation to deflect the idea that anyone could predict or manipulate "true rng."  There are no true "random numbers," only more complicated, obfuscated, and difficult to manipulate algorithms.  Of course the more you turn that up the more resources it takes just to generate random numbers, and for most games that isn't a wise investment.  I've seen companies put in safeguards to prevent manipulation but not prediction.  With data mining that is pretty much a lost cause.  You'd have to host it server side and protect the algorithm to prevent prediction, probably only worth investing in if you were an online casino or something.

I have no idea the point you’re trying to make here. I never said anything about “true random numbers” whatever that means. I’m using the term random as human society as a whole has defined it. 
 

Also games can generate random numbers on a small scale. In fact, a game you’ve probably heard of titled “Don’t Starve Together” does this to decide loot drops from mactusks, believe it or not (which happens to be the original topic of this thread). This is why if I went in game and booted up a world and asked you to tell me what my next 10,000 mactusk drops would be you would never for as long as you live be able to predict them. Each time this happened, you would fail. This is because computers can randomly generate outcomes on a small scale. 

No matter the scale though, it would still be “random” as per the agreed definition of the word accepted by society at large. 
 

Now, imagine I tell you I’ve made a program that will randomly generate numbers 1-10

I then tell the program to output the numbers 7,2,5,9 in that order.

if you use the program, at first you may think it’s random. Then you generate a second sequence and it gives you the same numbers in the same order. This happens again, and again, and again.

This isn’t evidence that a computer can’t generate random numbers, it’s evidence that my program can’t generate random numbers. Just because one game uses algorithmic patters that can be learned, deciphered, and exploited doesn’t mean that every game does, and it doesn’t mean that it’s impossible for a game to do so. This is why using other games as your example is meaningless. And this is scenario I was talking about when I mentioned games or programs that weren’t “truly random”.

[ArubaroBeefalo]

some people are mixing "not having capacity to predict it" with "not deterministic"

[Yuuko]

3 hours ago, Vinja said:

This isn’t evidence that a computer can’t generate random numbers,

Lets bring this back down to earth, and quell your "not true rng" fallacy.  We're not talking about mythical, imaginary programs we've "written to do x," we're talking about DST and tusks.

DST is made in lua.  Math.random in lua is 100% deterministic.  You feed it a seed, and from there every single "random number" is derived.  Give two different systems the same seed they will generate a completely identical set of "random numbers."  If you know the seed value you can automatically know every single result it will generate.  This is why world seeds let people generate the same world.  The "world seed" is just the seed value fed into the rng when worldgen starts.  Feed it the same seed and we get an identical "randomly generated" world every time.

Seeds can be set in different ways, dst does it with 

Spoiler

tonumber(tostring(os.time()):reverse():sub(1,6)))

and from there everything is 100% predictable.

The next step would be to manipulate the rng by starting it at a specific time (or spoof os.time on start) and plotting on what and how often it will be called to drive our 100% tusk drop rate play through.  You could also observe all random results and decipher where you are in the pattern, predicting what the next results would be.

This is a common process employed by TAS designers, manual speed runners, mmo grinders etc.  When I meld talismans in Monster Hunter Rise I can take my results and check against their loot table allowing me to predict exactly what my next results will be, informing whether I continue melding on this seed or take measures to change the seed and try again.

There was a Minecraft speed runner who was caught cheating because the drop results he received during his record run were impossible according to the rng algorithm minecraft employs.  They took the results from his video and reversed the process finding that no seed could have generated his results.

Spoiler

 

ps - No True Scotsman fallacy - https://en.wikipedia.org/wiki/No_true_Scotsman#:~:text=No True Scotsman%2C or appeal,by excluding the counterexample improperly.

What you have been saying is that a "real rng" system can be predicted or manipulated.  This "real rng" system doesn't exist.  It is very difficult for a computer to generate a random number that can't be predicted.  Game devs don't even try, its not worth it.  Maybe if you ran an online casino you might invest in the resources to go harder, there are ways to alter the seed and use a more complex algorithm if you want.  Its just not worth the cost, Klei doesn't care if you can predict the rng here lol

[Vinja]

7 hours ago, Shosuko said:

Lets bring this back down to earth, and quell your "not true rng" fallacy.  We're not talking about mythical, imaginary programs we've "written to do x," we're talking about DST and tusks.

DST is made in lua.  Math.random in lua is 100% deterministic.  You feed it a seed, and from there every single "random number" is derived.  Give two different systems the same seed they will generate a completely identical set of "random numbers."  If you know the seed value you can automatically know every single result it will generate.  This is why world seeds let people generate the same world.  The "world seed" is just the seed value fed into the rng when worldgen starts.  Feed it the same seed and we get an identical "randomly generated" world every time.

Seeds can be set in different ways, dst does it with 

  Hide contents

tonumber(tostring(os.time()):reverse():sub(1,6)))

and from there everything is 100% predictable.

The next step would be to manipulate the rng by starting it at a specific time (or spoof os.time on start) and plotting on what and how often it will be called to drive our 100% tusk drop rate play through.  You could also observe all random results and decipher where you are in the pattern, predicting what the next results would be.

This is a common process employed by TAS designers, manual speed runners, mmo grinders etc.  When I meld talismans in Monster Hunter Rise I can take my results and check against their loot table allowing me to predict exactly what my next results will be, informing whether I continue melding on this seed or take measures to change the seed and try again.

There was a Minecraft speed runner who was caught cheating because the drop results he received during his record run were impossible according to the rng algorithm minecraft employs.  They took the results from his video and reversed the process finding that no seed could have generated his results.

  Reveal hidden contents

 

ps - No True Scotsman fallacy - https://en.wikipedia.org/wiki/No_true_Scotsman#:~:text=No True Scotsman%2C or appeal,by excluding the counterexample improperly.

What you have been saying is that a "real rng" system can be predicted or manipulated.  This "real rng" system doesn't exist.  It is very difficult for a computer to generate a random number that can't be predicted.  Game devs don't even try, its not worth it.  Maybe if you ran an online casino you might invest in the resources to go harder, there are ways to alter the seed and use a more complex algorithm if you want.  Its just not worth the cost, Klei doesn't care if you can predict the rng here lol

Yes world generation is seeded, loot drops from enemies are not. Two entirely different things. This is why if you were to do one set of actions, then kill a mactusk, then revert to previous save and do the exact same set of actions and then kill a mactusk it would not yield the same results. 
 

Xcom is a good example of a game with seeded actions. If you move to a spot and shoot an enemy, then revert your save and move to that exact same spot and shoot at the same enemy, it will always have the same result. DST does not work like this. If it did someone would have probably cracked it by now.

If the speed running incident you refer to is dreams cheating scandal then it literally means nothing in regard to this argument. He wasn’t using some clever way in game to manipulate the rng to get his desired outcome, he literally changed the coding of the game to increase his drop rates, which I believe we only know with100% certainty because  he admitted to it. They didn’t figure it out because of some “seed” like you suggested that meant his outcomes were impossible, they discovered it because they ran billions/trillions of simulations using the drop rate from the games coding and found that the odds that he got as lucky as he did was like several trillions to one or something, which meant that while his luck was *incomprehensibly unlikely*, it was not *impossible* which is an important distinction.

Assuming this is the scandal you’re referencing I don’t know if you’re misrepresenting what happened intentionally to bolster your argument or you were just misinformed or misunderstood what happened but I’d encourage you to take another look at it.

 Anyway, bottom line is, computers can generate random numbers. That’s a fact. I believe DST can do this for loot drops but I’m not familiar with the code. I would love to be proven wrong though because as a console player primarily I can’t install mods so a way of manipulating rng to get guaranteed tusk drops would be super nice lol.

[Yuuko]

4 hours ago, Vinja said:

Yes world generation is seeded, loot drops from enemies are not. Two entirely different things. This is why if you were to do one set of actions, then kill a mactusk, then revert to previous save and do the exact same set of actions and then kill a mactusk it would not yield the same results. 
 

Xcom is a good example of a game with seeded actions. If you move to a spot and shoot an enemy, then revert your save and move to that exact same spot and shoot at the same enemy, it will always have the same result. DST does not work like this. If it did someone would have probably cracked it by now.

If the speed running incident you refer to is dreams cheating scandal then it literally means nothing in regard to this argument. He wasn’t using some clever way in game to manipulate the rng to get his desired outcome, he literally changed the coding of the game to increase his drop rates, which I believe we only know with100% certainty because  he admitted to it. They didn’t figure it out because of some “seed” like you suggested that meant his outcomes were impossible, they discovered it because they ran billions/trillions of simulations using the drop rate from the games coding and found that the odds that he got as lucky as he did was like several trillions to one or something, which meant that while his luck was *incomprehensibly unlikely*, it was not *impossible* which is an important distinction.

Assuming this is the scandal you’re referencing I don’t know if you’re misrepresenting what happened intentionally to bolster your argument or you were just misinformed or misunderstood what happened but I’d encourage you to take another look at it.

 Anyway, bottom line is, computers can generate random numbers. That’s a fact. I believe DST can do this for loot drops but I’m not familiar with the code. I would love to be proven wrong though because as a console player primarily I can’t install mods so a way of manipulating rng to get guaranteed tusk drops would be super nice lol.

All rng is seeded.  The world seed is separate from the rest of the game's rng seed, but both rng's are seeded the same way using the same function.  Any time the game pulls for a random number it is drawing from the string of possibilities generated from that seed.  There is no other way to do it, and again this is fact.  Lua math.random function is 100% deterministic.  If you are getting different results from rolling back it is because the seed or the actions that have occurred between the roll back and the point where you draw your results have changed.  Control for these and you will find you get the exact same outcomes every time.  That's just how it works.  Its also possible the rng position isn't serialized, but generated per session aka not rolled back with your save.  Again, this doesn't change how rng is generated or the ability to predict results.

As for the minecraft thing - Sounds like you didn't watch the linked video, because it states pretty clearly who was caught cheating, and how.  The how part specifically relates to the fact that there are only 281 trillion possible chest loot seeds in the world, and by charting every single seed against the play through the player submitted for record they found that their loot chest combination did not exist anywhere in any seed aka literally impossible.  If it were "truly random" they couldn't generate these proofs and any combination of loot chests would be possible.

Its kinda funny because obviously you understand this.  You say so yourself here

Quote

if the numbers aren’t truly randomly generated and they are based on some sort of pattern or algorithm then you could figure out that method and use it to predict what would come next

Like........ this is literally, exactly how the "random numbers" are generated.  This is exactly why they can be predicted and manipulated.  There is no actual random number being generated.  Computers can't do that.  There are some very resource intensive rng methods that involve sourcing the seed from outside parameters which could be considered "more random" but that doesn't change that its a deterministic algorithm.  Plug in the same seed, get the exact same "random numbers" every time.

Bottom line - computers can't generate random numbers.  They can run very complex algorithms and source seeds in such a way that they appear random, and are random enough for most uses, but are actually deterministically generated strings where the same seed will generate the same results 100% of the time.

[Hornete]

5 hours ago, Vinja said:

This is why if you were to do one set of actions, then kill a mactusk, then revert to previous save and do the exact same set of actions and then kill a mactusk it would not yield the same results. 

You are (likely) correct on that, but it's not for the reason you think.

There are a lot of random calls in the game, and for each time the random function is called, the next random number is generated. There's rng in almost everything such as:

• Choosing which idle animation to play(If the character hits a certain chance or whatever, play their unique idle)
• Timings for a lot of things are randomized by a second or so to make them feel more natural and not the same every time
• The direction a mob chooses to wander in
• Random positions generated, used for many mechanics in the game, one example being choosing a place to wash ashore when drowning.

There are thousands of these calls happening fairly quickly, which is why you wouldn't get the same results of a mob drop most likely. I know some games can seperate the RNG in different mechanics so that one random call in one place doesn't interfere with another. (I think Isaac does this?)

Funny story, a friend had brought up RNG manipulation in Pokemon and we had looked into seeing if RNG manipulation could realistically be done in DST to guarantee a krampus sack despite how many random calls are called in the game, as a silly little project. Unfortunately I had found that even with a lot of things unloaded, Dragonfly's lava ponds continued to update their heat cap for spreading fire off-screen, which has a random call.

This thread got super side tracked but I love technical topics like these so I had to intrude in, haha.

[ButterStuffed]

2 minutes ago, Hornete said:

There are a lot of random calls in the game, and for each time the random function is called, the next random number is generated.

Time is a common base for seed generators to produce a "random" number.

Does DST use time for it's seed generation?

[Hornete]

2 minutes ago, ButterStuffed said:

Time is a common base for seed generators to produce a "random" number.

Does DST use time for it's seed generation?

Yup! Just like a lot of other games it uses time to generate it's seeds. DST uses the current unix timestamp on your computer.

This also mean's its plausible for two people to accidentally generate the same exact world. I think I had seen an incident of that happening on a discord chat long ago! Neat stuff.

[Wonz]

On 2/21/2023 at 6:19 PM, DeadWhereX said:

I agree with you. RNG is okay for the world generation,

Only because 99% time you get all the necessities will be there and because there's workaround for some resources being there where we don't want to like we can replant bushes, build pig houses, even grow fig trees. But for tusk rng there's no workaround. 

About beefs I don't like they can be gone forever if not enough spawned in pub server to satisfy that one greedy wigfrid player. 

[BezKa]

On 2/23/2023 at 8:44 PM, Hornete said:

There are a lot of random calls in the game,

I accidently pushed my friend away just as she was receiving her daily gift. She caught a glance at it, but it was cancelled because she was too far form the alchemy engine. When she came up to receive her gift again, she got a different one.

Since we're taking about side tracking the thread.

[Hornete]

2 hours ago, BezKa said:

I accidently pushed my friend away just as she was receiving her daily gift. She caught a glance at it, but it was cancelled because she was too far form the alchemy engine. When she came up to receive her gift again, she got a different one.

Since we're taking about side tracking the thread.

I can't see how that is possible because the daily gifting is done on Klei's backend servers as far as I know, the gift skin being chosen and determined on their server and then the game receiving it from the server to open and mark it as owned.

[BezKa]

9 hours ago, Hornete said:

I can't see how that is possible because the daily gifting is done on Klei's backend servers as far as I know, the gift skin being chosen and determined on their server and then the game receiving it from the server to open and mark it as owned.

Yeah, it seems strange to me as well but that's what she claims. Maybe we can try to reproduce it sometime on video.