Fix The Shuffling Randomization

Time and time again, multiple users complain of bad hands caused by the randomization that duelingnetwork gives them. If the same mechanic is used for the hand shuffling (Which at times, doesn't shuffle at all), then it's quite obvious that the randomization is a slight bit off.

I'm not entirely sure if there's an easy fix to this, or if it would have to be something that the system needs to be reworked to fix, but I'm sure many users would greatly appreciate if the mechanic got revamped.

Thanks!

The staff has already stated that unless you can provide legit proof that the shuffler is bad, nothing will be done. Not gonna lie though, I do believe this shuffler is bad. Konami's yugioh online and other dueling systems seem to have much better shufflers.

Would a recorded test of 1000 to display shuffle rates be worth anything as proof? I don't see too many things wrong with the shuffle, could be improved though.

I can easily get some shuffling results in Unrated duels and provide screenshots.

I could do video as well, but that would take quite a while.

The shuffling algorithm to be blunt is not broken.

In fact, it is the most comonly used online shuffling algorithm in the world (including online casino like games), you will experience multiple cards in hands from time to time. This is not impossible nor rare.

If you want to read about the shuffling algorithm operating in Dueling Network you can find it here:
http://en.wikipedia.org/wiki/Fisher%E2%80%93Yates_shuffle

Ill further comment about your appearance of "bad hands" though. I do not know why this surprises you that you can get bad hands. Bad hands are a probable event just like having a good hand. Like members have stated, unless you can provide me statistical evidence of the shuffler being skewed from the probabilities of certain hands then the code will not be revised.

What constitutes as statistical proof. Recording your duels? No. Doing a statistical analysis requires a lot of work. You first need to form a hypothesis. What is it are you complaining about? What event do you think the shuffler is skewed at producing? A "bad hand" if not a hypothetical event. This is based on your own deck building skills first and foremost. If you have a large combination of "bad hands" then yes you will see a skew, but this not because of the shuffling.

After you form your hypothesis calculate a P (probability) value for the expected probability of your event occurring. After you calculate this, you then need to put your expected P value to the test. At the end of your test, calculate an observed P value and compare the two. With enough data you will find these two values to be very close (Ive done this test a few times on a number of different events).

Ill also make another note about online shuffling. When you shuffle your Deck in real life, you are not shuffling your Deck in the way that DN does it. The way you shuffle is probably very biased, by shuffling clumps of cards grouped together at a time. This is not a randomized shuffle but very skewed based on the clumps that you are grabbing with your hands. DN shuffles single cards over and over randomly to get your Deck each time the shuffle function used. DN is more accurate than you think.

Ill link you to this topic to get an idea of how your statistical test could look.
http://forum.duelingnetwork.com/t3328-broken-draw-system#44998

If you want to tell me what you actually think is wrong with the shuffler. Ill provide you with a statistical test you can do and you can try it for yourself.

How would you argue that the probability of draw rate of 6/40 cards is bad? The problem would majorly be either the deck or a series of really bad luck. It happens in real life and out.

+everything det says.

recently in this past 9 days, i constantly getting bad hand, what's funny is taht the start of each duel it's simillar. I have 3 of those trap cards, but i never able to draw it, however, I always start out with 2 fortune lady fire (I only have 2 of them, but I seen to get them all the time). Furthermore, I remember numerous time drawing 3 dark in a roll. Ok so maybe it's my deck is the problem, but oh wait! When i play on tag force or the DS game, man I get more variables.

I don't see why people blame their bad hands on the shuffle.

It's like going to YCS and demanding they let you get a redo because of your own bad shuffling.

becuase when you play other yugioh game on different platform, the cards you draw are fine, however, when here and you draw 3 of the same cards in a roll or starting a duel with almost the same hand or the selected few cards always appear in the starting hand (despite the fact you put only 1 of them in the deck) sort of mean something.

I haven't had that problem at all though. Shuffler works fine for me. Maybe it just hates you, in which case you could apply for harassment charges against it.

yeah maybe I should! Can you introduce me to some good people that could help me out

I mean do you think the shuffler knows which cards are bad? lol

The shuffler is broken cuz Enchanting fitting room doesn't give me vanillas half the time.

I actually have had no problem with the shuffler. Bad hands do happen and a good player is able to work with them.

Although, I do get suspicious sometimes when I risk setting 3+ cards on my first turn and my opponent has Heavy waiting for me. But that could easily be because the Heavy Storm scenarios just stick out to me more than the Non-Heavy Storm scenarios. In actuality, I probably get away with a heavy backrow more often than I don't. XD

But yeah, I think that's a big factor here. You guys are saying that you have bad hands "a lot". Well, I doubt that. Your bad hands only stick out to you because they're...well, bad. And people tend to remember bad things a lot more easily than they remember good things.

If you actually do have more bad hands than good ones, then the reason is that you are not dueling enough. There's something called "The Law of Large Numbers". For example, when used in reference to something that has a 50% chance of doing one thing or another. Like a coin flip. You could flip a coin 4 times and each time you get heads. That's not a 50% chance of either outcome then, is it? Well, the problem is that you haven't flipped enough. The Law of Large Numbers states that the more you flip the coin (numbering into the hundreds and even thousands) the more even it will come out (eventually you will have, say, 456 heads and 428 tails, and get even more close the more you flip).

For less confusion, I'll quote the wikipedia article: "According to the law, the average of the results obtained from a large number of trials should be close to the expected value, and will tend to become closer as more trials are performed."

Of course, your deck build changes the expected value, as Det explained.

In short, the shuffle is perfectly fine. I don't even find it strange if I have a terrible run of bad hands. It can happen. If I NEVER got bad hands, THEN I would be questioning the shuffler.


Here's the wikipedia article, by the way: http://en.wikipedia.org/wiki/Law_of_large_numbers

ok another scenario, 3 duels, In this deck i have only 6 trap cards in it and like 7 spell. Amasingly, for my starting hand I was able to draw the 6 traps in the beginning. Furthermore, it seems like whenever I slap more than 1 set card, the opponent seems to able to draw a heavy storm the next turn. Another amasing thing was one time I used fortune lady light to search out another fortune lady (dark), i saw another fortune lady on the top of my deck. Personally I think, hey the deck will get shuffled and I wouldn't draw into it, however, the next turn, the card i draw was fortune lay light. There's another case is when I slap APprentice magician in def, however, the next 2 card i draw was apprentice magician (that is some good shuffling). In addition, don't forget about fire, i have 2 of that card and i always draw into it the first turn.

We cannot stress enough that the shuffler itself, is not a living thing that gives you bad hands.
That's just silly.
Aside from that, every shuffling program will never be truly random, since that is impossible for programs and computers.
The program used on DN is as close to random as you can get with a card-based shuffler, and works on a randomizing algorithm that would confuse most of us.
Long of the short, unless you want to split an atom into 40 pieces, deal with the shuffler.

Everyone who gets a band hand here and complains should try thoroughly shuffling in real life (not the crappy shuffling most of you guys do) and comparing a lot of hands. They should be roughly the same to be honest. The shuffler here is just a bit more random.

I'm okay with the shuffler. I mean I had bands a few times, and the same with good hands. I don't see a problem with it.

I always drew bad hands while anyone else can use the same deck and get a hand that's a lot better than mine.

But that's just my lack of sacking skills

Random is Random. Altough random in computing is rather arbitrary in a sense. It is not necessarily random. Usually a they start from a seed value that is uniquely chosen to make it seem random, like system time, etc. Also, keep in mind some RNG can be easily predicted due to some clever pattern analysis. I would assume DN uses a more cryptic and popular RNG.

I will not complain about the shuffler giving a bad hand.
I feel on the occasion, though, that it's not quite random.

Like when I draw FuFu in the opening hand in 5 out of 7 duels in a row using the same deck.
Drawing a limited card from a 40 card deck in the opening hand is 1/8.
The chances of drawing FuFu in the opening hand 5/7 duels is 7C5 * (1/^5 * (7/^2, which is roughly 0.049% chance.

And then there are those games where you get a bunch of pairs. Makes me want to think the shuffler likes to put cards that are next to each other, next to each other (since I don't tend to jumble my decks).

IDK. I'm taking Stat right now, so maybe one of these days I'll make a statistical analysis of the shuffler to turn in as extra credit.

I think people believe "random shuffling" to mean that, instead of an equally probable chance of drawing each card, they should always have a hand that would be reminiscent of a "random" shuffle. In this case people want a very unlikely result to occur, such as 6 different cards in the opening hand consistently. If more people took and understood probability and statistics this would not be a problem. Because its random seemingly "unprobable" things will happen, this is a sign of a healthy shuffler. If you believe this events occur unrealistically you either:

1) Need a more objective take on it ("Wow... this is the 20th time this week I've had a very good hand out of 25 duels. Plus I built this deck terribly, which makes this even more rare.")
2) Do the research, document proof ("I have drawn all 3 Tengus in my opening hand 700 times for 1000 duels. This is an unprobable event that should not be occurring. Here are screenshots/video/proof.")

storyteller wrote: Makes me want to think the shuffler likes to put cards that are next to each other, next to each other (since I don't tend to jumble my decks).

That is the way it is suppost to be, however it isn't like that since in that case using "Sort" button would cause players drawing into doubles and tripples way too often

Mod im Disapointed, Shocked you did not see dat necro post