Reverse bear traps and Jigsaw's boxes
Which one of the following statements is true?
1) There is one Jigsaw's box common to all Reverse bear traps;
2) every trap has its own box, but a box can be linked to two or more traps;
3) every trap has its own box and a box cannot be linked to more than one trap.
Comments
Evry trap has it's own box if you trap the same survivor that may have the same box
yes but the box linked to the survivor
I think me and my friend try this so many time but with only 4 traps we don't know if was RNG or the code most of the time 2 out 4 trap will be on the same box
I doubt that because more than once I removed my trap where someone else just removed from.
Number 2
Every box has 50% chance when you search it. They´re not linked.
This is...technically true, but not guaranteed? There is nothing stopping a Survivor having a second trap placed on them from finding their key in the same box, but that's just because the game randomly picks a box when you get a trap on. It doesn't, to my knowledge, take into account any previous traps or boxes or anything.
This is definitely wrong. A given trap is linked to a box randomly chosen when it is placed on a Survivor's head.
Thats what everyone thought. Someone on reddit tested 500 traps and the results are pretty disappointing:
https://www.reddit.com/r/deadbydaylight/comments/anq365/pig_rbt_stats_part_2/
Seems like whenever you search a box its just 50/50, until the last one. They're not cleverly linked or so, just 50%.
How should that work with 4 boxes and the way they are distributed? If it has a 50% chance to get off on first try, it should be linked to 2/4 boxes. Do you have a source for the dev statement?
The fact that their results had the key show up in the first box about 50% of the time does not actually indicate that boxes have a 50% chance of getting the key at any given box. There's nothing wrong with the data the person gathered, but conclusion they and other commenters are coming to does not support the official word of the devs because with the way the boxes function according to the devs, you have a 25% chance of getting it off on the first box (assuming four boxes). How do I know this? Simple math. If the game chooses a box to put the key in when you get a trap on your head, there is one box that will 100% give you the key and three boxes that will 0% give you the key. That means if you pick a box at random, you have a one in four chance of picking the correct box, which is 25%. ~50% of the people getting it off on the first box can very easily be attributed to just purely getting lucky on which boxes they walked to. They just happened to go to the correct box first about half the time.
There are three conclusions you could come to with regards to this:
1) The devs are somehow completely unaware of how they programmed the Jigsaw Boxes and do not realize/have forgotten that they programmed them to have X percent chance of success per box instead of just randomly putting a key in one of the boxes.
2) The devs ARE aware of how their boxes function and deliberately lied about it to viewers for no apparent reason and with nothing to gain from doing so, given that the topic of discussion where they brought it up was the Luck mechanic (see the video below).
3) The info the devs gave are correct, and the Reddit study does not actually matter.
I'm going by Occam's Razor here, so #3 seems like a safe bet here.
As for the dev statement:
(go to 29:36 if the video does not automatically put you there)
The exact words:
"If you are doing a Jigsaw Box, it is not a random roll when you complete the action. So it's not the same thing as a hook or a (Trapper's) Bear Trap. It is predetermined. When you get that (Reverse) Bear Trap on your head, the game already knows where the key is."
2x tests over 500 samples with very obvious results speak for themselves. This is not a test over few games, its 500 traps with a very distinct distribution of 50/50. Twice. Thats way out of error range and "reliable" so to speak.
No offense, but dont trust what the devs say. Thats something i learned from the very release of this game. Some of them are not responsible for the programming of certain things. I believe in the math in this case, its a coin flip.
The problem is that the results in that reddit thread are possible in a "the key is in one of four boxes" scenario. Slim? Possibly. Impossible? No.
Let me put it to you like this: If I secretly put a ball in one of four cups randomly 500 times without you knowing, and you managed to guess the correct cup about half the time, does this mean that you have a 50% chance of finding a ball in a cup, or did you just get lucky about half the time? Having the key randomly placed in a box every time a trap is put on does not mean it is impossible or even nigh impossible for someone to guess the correct box about half the time.
Why should I assume the devs either lied about how they programmed the boxes or completely forgot how they did it? This isn't a "Oops, we screwed up a few lines of code" thing. What the devs are claiming with regards to how the boxes are programmed is EXTREMELY different from how you (and others in that reddit thread) are alleging they are programmed to the point where it would not make sense for them to just forget how they programmed it. So either they'd be lying (again, why? for what benefit?) or you're wrong about how they're programmed, regardless of what the results are claiming.
Its almost impossible for an even distribution to look like a coinflip distribution after 500 samples. Twice.
Anyone who knows maths can immediately tell this is a coinflip distribution.
Is is possible that it is an even distribution? Yes, but almost zero chance. Its 99,99...% a coinflip.
I dont think they lie intentionally. They sometimes just dont know themselves, unless its the one who programmed it.
Is it almost a 0% chance? Or are you just assuming it is?
Why would they not know how they fundamentally programmed an entire Killer?
Here is a quote from someone who did the math:
"I ran some stat tests and it turns out the distribution isn't 1:1:1:1 nor does it seem to be 4:2:1:1 as you claimed but rather it seems to be 8:4:2:1.
If we look at the people who got it off (ignore deaths and DCs) we get the following from a chi square test:
p-value assuming the distribution is 1:1:1:1 was <.00001, less than a .001% chance.
p-value assuming the distribution is 4:2:1:1 was .0276, a 2.76% chance
p-value assuming the distribution is 8:4:2:1 was .2716, a 27.16% chance.
Given those results the last model seems the most likely, but the second one is possible. What this means is that you most likely have an 8/15 chance the key is in the first box, 4/15 on the second, 2/15 it's in the third, and a 1/15 chance it's the last one. I spent more time on this than i should have but it was really interesting. If it turns out I am right then I hope they fix this ASAP."
For the second question, i would say that this is very much typical for the devs
Apologies, but I am not well-versed in statistics to understand any of that. Doesn't mean it's wrong, just that I don't get it at all.
What is typical for the devs? Being completely wrong about how they fundamentally programmed an entire killer, announcing the fundamentally inaccurate information on a Q&A session, and then having no one correct them on that fundamentally inaccurate information for months on end? That's typical for these devs?
"p-value assuming the distribution is 1:1:1:1 was <.00001, less than a .001% chance."
Thats the chance, that an even distribution, as the devs tell us, ends up in the test results. Less than 0.001%. And that goes for two tests. So less than 0,00000001% chance, that it is an even distribution.
If the test results are true, then it is a coinflip after searching a box, until the last one.
They said its an intended feature for doctor to see snapping out of it with nurses calling, when they added that feature. After legion, they removed the doctors ability to see snapping out of it with nurses calling, because it was a bug. Two devs, two contradicting statements.
Dont take their words for granted. Believe it when you see it. And here you see a coinflip for the trap rng.
Fair enough, I will trust that what the person in the Reddit thread said is accurate.
Considering something intended or not is a HUGE difference from declaring "We programmed the boxes to function so that only one of the boxes is the correct one." when the reality is "We programmed the boxes to function through RNG so that each box has an X% chance of getting the trap off." The sheer difference in execution there is not something that both a gameplay programmer and gameplay designer can just get wrong. It's a large enough difference that they made it a point to mention that RBTs aren't affected by the Luck Mechanic because the way they are programmed is incompatible with how the luck mechanic functions.
Yes it is. It only takes half the time its supposed to take to remove a trap. There has to be misscommunication somewhere between the devs. Either the info was wrong and traps are supposed to be 50/50 or the programmer did not programm what he should (Or the person on reddit made up the stats, which i find unlikely)
Only a statement from the devs can clear things up wether its supposed to work that way or not. Anyways, for pig players, this is bad news.
The devs issued a statement already, though. It was a statement they issued in response to all the people who misinformed with how the RBTs were programmed. I linked you their statement in their Youtube video. They said that the boxes are pre-determined when you get an RBT placed on your head. When you search a box, you don't have a 25% chance of getting it off or a 50% chance of getting it off. You either have a 100% chance because you picked the right box, or you have a 0% chance because you picked the wrong box. You find it unlikely that the stats the person gave on Reddit are inaccurate/falsified? Well, I find it unlikely that the devs in the video somehow did not realize that they programmed the Jigsaw Boxes to function on a percentage-based system instead of a "Only one of these boxes is the right one" system, never realized that they could have it be affected by luck despite the fact that they said that the thing about the Luck Mechanic was that they intended for it to function with RNG systems they implemented but they just never actually ended up adding any since the game was young, still decided to announce to viewers at large that the RBTs was not something that was affected by Luck because of the mechanics behind it, and then were somehow never corrected by someone(s) who knew that they were actually mistaken on that claim despite the fact that this was a recorded video that they probably had to have gone over before they uploaded it to Youtube.
I honestly dont understand the sentence, that makes up the second half of the text.
For the first half, if you mean their Q&A at 29:36, i'm exactly referring to that as the current problem. They say its linked to one box. But the tests show, that it's <0,00000001% chance to actually be that way. Its possible that the stats are made up, even though i dont see any reason or motive to do so and the poster seems pretty honest and casual. Not a conspiracy type of person who wants pig buffs. He just "sad oinks"
Maybe the devs communicated it wrong. Someone missunderstood it and it spread around. Maybe it is supposed to be 50/50. Maybe its supposed to be linked to one box, but the devs programm it wrong, thus the devs think its working like it should but it doesn't. Maybe the stats are made up and the devs can clarify that its indeed linked to one box and the stats are a lie.
The devs haven't responded to both topics, but if they read this and know the exact answer, they're very welcome to clarify this.
I don't see how someone at the company could completely misunderstand programming, spread it around, no one bothers to check the code at any point, and then after they make their announcement on video, no one actually checks to see if anything wrong with that information. And as far as a programming error, if they did, say, mistakenly program the boxes to put the key into two boxes instead of one, then it shouldn't be possible to have to check all four boxes because after you check two bad boxes, the key would have to be in one of the remaining two.
Honestly, i can. To get their knowledge from Q&A, they ask each other. They've missunderstood each other in the past several times (doctors+nurses calling), as humans do.
There's no guarantee that the test results are true, someone else would have to test it. And then people dont trust that person and so on. Until you test it yourself or see a whole video of it, we can only believe it or not.
Its a good thing to keep in mind though for Pig, maybe later on someone can confirm or deny it.
I cannot. That, to me, sounds like it falls outside the realm of plausibility. This isn't a minor misunderstanding such as "Was Snap Out of It + Nurse's Calling intended or not?" This is "How did we game developers make our game?" It's getting information regarding a huge chunk of programming wrong. It'd be like if a dev incorrectly said in a Q&A "We programmed the hook to have higher chances of successful kobes the fewer remaining players there are." Those are not just "Whoops, my bad" levels of wrong. Those are "What the heck possessed you to say something we clearly know is wrong?"