Play now:

MaxBeamer
Member

1. Why have you taken Lupina tokens out of the treasure rooms?

2. When are you going to put them back?

2. When are you going to put them back?

Tagged:

0

## Comments

2: as soon as you get an AMAZING treasure chest

Sir William: http://forum.dungeonboss.com/discussion/56856/sir-william-rework-epic-idea#latest

Abigail: http://forum.dungeonboss.com/discussion/56807/abigail-rework-epic-idea#latest

Do you choose the same chest position every time?

www.twitter.com/Darth_Craig

Are all plugs shameless?

Not really. The chances are always 33%. Changing your position changes nothing. You could tap the middle one over and over and never see it.

Favorite hero: Bovus

Favorite element: Nature

Currently: Trying to max out Epics

www.bossfightentertainment.com

Pay attention. Theoretically.

A truly random distribution means that, statistically, 33% of the time you will see Lupinas tokens is you select the same chest over and over.

Choosing chests randomly means that yes, you have an infinitesimally small chance of selecting the correct one, EVERY TIME! The odds are slightly higher that you’ll pick the wrong one until Big Fish goes out of business.

Unfortunately, even though perfect success and absolute failure are relatively low probability outcomes, picking randomly also means that the odds of you selecting the correct one will average out less than the 33% you would get by sticking with the same chest over and over again.

None of this factors in decidedly broken RNG, which might actually be a factor with this game, but it is a factual post based on statistically random distribution of rewards.

I'll refer you to a simpler example - 2 coins. You flip one and I flip one.

The ONLY possible outcomes are H-H, H-T, T-H, T-T.

We're not interested in getting heads or tails specifically, but rather that we "match" The criteria for this is H-H or T-T, which is no surprise, exactly (2/4), 50%.

So now, when we take this theory to a "3-sided coin", we know the only possible results for our chests are Left, Middle, or Right.

The match criteria is that the two events come up with L-L, M-M, or R-R.

The only possibilities for the two independent events are L-L, L-M, L-R, M-L, M-M, M-R, R-L, R-M, and R-R - for 9 in total.

Low and behold 3 match criteria, out of 9 possibilities, is exactly (1/3), or 33% chance.

This is based on a perfectly random sampling from the user, which unfortunately due to the human brain is not possible because of "selection bias".

One may not even realize it, but what they think may be randomly choosing a chest may in fact heavily skewed and the middle one chosen twice as much as either side. So in a game where the "success chances" on RNG are already so poor, I take the extra variance out, which will guarantee that every single time a Lupina token is "available", I will have the best possible chance to get it.

That being said, the OP is likely experiencing a combined circumstance with low "availability", as well as a bit of selection bias.

www.twitter.com/Darth_Craig

Are all plugs shameless?

If you want to feel like your presumption of which chest should have the token actually affects the outcome, try using the Monty Hall problem. Pick a chest, then don't open it and pick one of the other two. That way your odds go from a 1/3 to a 2/3 chance.

Of course, they actually don't do that, but it if helps you feel like your odds are getting better then by all means try.

Favorite hero: Bovus

Favorite element: Nature

Currently: Trying to max out Epics

(And I think it always goes middle)

My Lupina is now six-starred. Must've done something right.

I'm not trying to say the % is NOT 33% though.

What I am saying though, is that selection bias is real.

To achieve the 33% rate you MUST choose one of the chests at random, which the human brain cannot do.

Ergo, to remove selection bias, you have to use some sort of predetermined rule to guarantee that you are not favoring one chest over another. Choosing the same chest each time is one such rule that achieves this.

www.twitter.com/Darth_Craig

Are all plugs shameless?

You're wrong, @JackHallow666. And you know what, I don't care. You do you. Be confused and misinformed all you like, I really don't care.

Then switch to a different one, rinse and repeat.

I'm just worried that one day I'll pick a chest and there'll be a dead cat in there.

I'm not misinformed. That's just how probability works. Your previous picks don't affect your current or future ones. Picking the middle every time won't guarantee you anything. It's a 1/3 chance every time. I get what you mean though, if you were to try it 10 times, the odds of you getting at least 1 token from picking middle all 10 times is extremely high.

But then again, you could get more tokens by not picking middle every time because the probability that it will be in the middle all 10 times is extremely low.

Favorite hero: Bovus

Favorite element: Nature

Currently: Trying to max out Epics

Well by doing that you're of course going to get biased results. 10 times each with each side, that's 30 times total. Of course with such a small sampling it will appear one chest is more preferable than another.

But do that 100 times each, or have 9 other players do the same test, and you should find each chest is equally likely.

Favorite hero: Bovus

Favorite element: Nature

Currently: Trying to max out Epics

Well yeah, there's obviously a system. No such thing as true randomness. That's what RNG is for. It's essentially artificial chance. However a good RNG system should be able to accurately simulate randomness to the point where a person can't tell it's a system unless huge (and I mean massive) sample sizes are picked and studied.

It's probably not going to be a perfect 1/3 chance, but it's gonna be pretty dang close. And by the time we figure out how it works they could just have changed the algorithm for all we know.

Favorite hero: Bovus

Favorite element: Nature

Currently: Trying to max out Epics

www.twitter.com/Darth_Craig

Are all plugs shameless?

Really?

Cuz my Lupina looks a lot spikier now.

I figured that's where they'd hide me.

You quoted the Monty Hall effect, but that doesn't apply unless you remove a choice after knowing it wasn't the right one. I looked it up because I never heard of Monty Hall.

Obviously 30 is not a big sample. 300 or 300 would be better. 30 is just achievable quicker. The point was to keep a record, to see if there really is a difference.

There was someone on here once who thought they could tell the difference from the sparkles on the chests.

I would probably still be trying to 6* Lupina from chests if she hadn't turned up in daily rewards a few times with VIP multipliers. Can't recall if she was ever offered in the aether shop.

Also, go gorgons!

Same situation -- it kept crashing and I kept going back to the beginning, after selecting the middle one twice and getting Great! both times, I decided to select the left one instead because "Why not? I know the middle one has the Great! option." The left one did indeed have her in it.

The thing is that the random seed going into the room will determine her location. If you replay the dungeon but don't use exactly the same moves and targets as you had the first time, it is very likely that your random seed will be different and her location could have changed.

Exactly what I was saying. Choosing the middle chest every time doesn't up the odds, let alone guarantee you a chest eventually. FatCat could run that test, but he'd have to also get two other players to try going all left and all right chests too. I guarantee you the results will be extremely similar after 100 or so chests. Hence, all chests have the same chance, and choosing a different one every time vs the same gives you the same chances of getting an Amazing! chest.

Favorite hero: Bovus

Favorite element: Nature

Currently: Trying to max out Epics

Doesn’t matter whether you consistently pick the left, right, or middle, but choosing the same chest consistently does guarantee you that you are not introducing bias into your selection and thereby skewing the odds of getting an Amazing! chest.

Look, it's okay to just admit you're wrong and learn from something. Changing what chest you choose each time doesn't affect the outcome. There's no "bias" because it's a 1/3 chance every time. Tapping the same chest every time doesn't guarantee anything. You could just end up never getting an amazing chest from the middle. Doesn't mean the odds are wrong, just that you're incredibly unlucky.

Picking a different chest each time doesn't skew the odds at all. There's nothing here to skew. You're looking at multiple interactions like they're a chain or something. Each one is unique. It's a 1/3 chance each time.

No offense but how hard is this to understand?

Favorite hero: Bovus

Favorite element: Nature

Currently: Trying to max out Epics

So because of the quantum state of the Lupina token, in a sense it is not in any one chest, but in all three chests at a 1/3 chance of appearing when observed, and also guaranteed to exist if all three chests could be opened. Therefore, every time you enter a treasure room, the odds are the same. The token may not

actuallybe in the same chest every time, but each individual chest has the same chance to contain the Lupina token.So instead of looking at it like it's in one of the three chests and none of the other two in every iteration, look at it like it is in all three chests at once, with a 1/3 chance of appearing when you open your selected chest.

In that way, every single treasure room encounter is the same. And it does not matter which chest you choose each time, not just because you don't know which one it's actually in, but because it exists in a quantum state in all three of the chests at the same time.

Now, before you yell at me for using this whole 'quantum mechanics' BS, this is actually how stuff works. When talking about quantum states and observations in this way, I'm not saying there's some unexplainable crazy science stuff going on, I'm just saying that the location of the token is unknown to anyone until it is found. And until it's found, it might as well be in all the chests at once, with a 1/3 chance of appearing when opened.

This has nothing to do with the game's own code limitations, as quantum observation is more of a way of explaining how human observation works. Of course, the game's RNG code can only simulate these actual observations, it would take an actual quantum computer to have true randomness. So in a way, you're right. Each selection isn't totally random. After all, it

ispre-determined at the start of every level. However. You don't know which one it is, and therefore the token still exists in a quantum state. The actual coded chances may not be a perfect 1/3 for each chest, but it's likely close enough that it doesn't matter. Or it could actually be perfect! Who knows?The true answer to this problem sort of exists in a quantum state too, and will only exist in your explanation or mine solidly once a dev observes (confirms) one of our theories.

Until then it could be worth your time to quickly look up basic probabilities of series', quantum observations, and the psyche of people and why one would feel like the odds of such a thing are higher when they decide to pick the same door over and over. It's really interesting. I seriously don't mean this in a passive-aggressive manner either. I loved learning about this stuff!

Favorite hero: Bovus

Favorite element: Nature

Currently: Trying to max out Epics

No. Quantum mechanics is not involved. And no, you have no idea what you’re talking about.

I’ll leave it to better people than myself to explain it to you. Your fascination with pseudo science leads me to believe that you really don’t want to understand.