shuffling

[infectedswampwalker]

  .

[Tyril132]

The problem with your claim is that you don't really seem to be allowing for basic probability.

In layman's terms, something being statistically likely is not the same thing as being guaranteed. If you flip a coin, the odds of getting heads once is 1/2. The odds of getting it five times in a row is 1/32. Statistically, this is unlikely but if you run the test 100 times you'll see that it happens often enough to be noteworthy.

The point is: you can't get any kind of meaningful pattern from observation of just a few games. We can, however, explain what's happening by using  hypergeometric distribution.

Let's say you have 24 lands in a standard 60 card deck. In your opening hand, you drew 2 lands, which leaves 22/53 cards in your library as lands. Here are the chances of you drawing a land over a series of sequential draws:

1 draw: 41.5%
2 draw: 66.2%
3 draw: 80.8%
4 draw: 89.3%
5 draw: 94.1%
6 draw: 96.8%
7 draw: 98.3%
8 draw: 99.1%

Statistically, you are very likely to draw a land within about 5 cards but this is not a guarantee. Out of 100 games, you will have at least one on average where you don't get that land you're looking for even after you draw 8 more cards.

The shuffling algorithm would be broken if this didn't happen at least some of the time.

[Trevor]

This topic has been brought up many times:
http://www.lackeyccg.com/forum/index.php?topic=358.0

No one has come up with any evidence to suggest the shuffler is anything but perfectly random. I have conducted extensive tests, and it is essentially perfectly random.

People have a tendency to ignore the times things work the way you like, and dwell on the times things happen the way you don't like.

If you get too many of a particular card, suddenly it means the shuffler is broken.
If you get too few of a particular card, suddenly it means the shuffler is broken.

Unlikely things are going to happen. If unlikely events never happened, then that would actually mean something is broken.

[Trevor]

What Trevor said.

Chances are, if this happened a few times, it was bad luck or confirmation bias. If it happens consistently, your deck's land ratio probably needs tweaking.

I remember learning in school exactly how to calculate the probabilities of complicated sequences of events, and I wish I remember.

Lackey is completely blind as far as what the card types or names are. In testing randomness, in order to not confuse one card with another, it's helpful to not have any doubles. People with 10 copies of a card may think things nonrandom if they get 2 of them in multiple hand draws, when in reality they may be getting a different 2 each time.

Tyril132, did you delete your post?

[Tyril132]

Whoops, I figured that might come across as kind of troll-y and removed it to be safe. lol

The basic point stands, though.

[Trevor]

Whoops, I figured that might come across as kind of troll-y and removed it to be safe. lol

The basic point stands, though.

I thought I accidentally deleted it or something.

[Tyril132]

well the only thing wrong with your post, is that a deck isnt not completely randomized
a deck is shuffled

lackey needs a shuffler, not a randomizer
thus presents the problem

Shuffling, by definition, is the "randomization of playing cards to provide an element of chance." If a deck is shuffled properly when playing a card game, it is for all intents and purposes completely random.

when you randomize a deck the chances of putting say 4 of the same card in the same spot are high
if you shuffle the chances are much lower unless you started shuffling with an organized deck

I'm sorry, but I'm not seeing any logical basis for your position.

When properly executed, there is no practical difference between artificial vs. real shuffling. The problem (and where you'll tend to see a visual bias) is when a deck is not properly shuffled using real cards, which can happen when the deck hasn't been shuffled enough times when using certain techniques.

For example: you'd have to shuffle a deck about 7 times using a riffle technique to randomize the deck to the point where drawing any specific card in the deck is equally likely. Most players probably don't spend that much time shuffling their decks, but that's likely to result in more clumping, not less.

[Trevor]

infectedswampwalker, proper shuffling is the same as randomizing a deck. If you don't shuffle properly with real cards, then you are cheating. I am not going to de-clump, or some other way fake the shuffle. And even if I did, people would still complain when they got unlucky that it was the fault of the shuffler algorithm.

The shuffler is currently properly randomizing the cards, which is working as intended. If you have evidence to refute that, present it and I will gladly investigate.

There are several things that are better with playing with real cards, as compared to playing online. Shuffling is one of those things that is better with Lackey as opposed to with physical cards. You just click a button, and everything is perfectly shuffled. Often times, playing with real cards, I wish I could just press a button and have my deck be perfectly randomized.


...
On second thought, you convinced me! I will emulate a bridge shuffle, so the cards are shuffled just like a person would with a physical deck. What's more, I have even updated your Lackey immediately. No need to download an update. Your current Lackey now has the shuffle emulation you requested. If you don't think so, prove me wrong.

[Saethori]

If, at any point in your playing career, shuffling is not the same as randomizing, then at that point in your playing career you were technically cheating. >_>

A lot of people like to claim evenness, through methods such as card weaving, but the rules for every TCG I've seen so far ask for a sufficiently random deck. If you know for a fact that there's a card of a specific type within the next three draws, then the deck is not sufficiently random, and therefore you are violating the rules of the game. There's really no two ways about it. People playing physical games are so used to inadvertently cheating while shuffling, that once they find something that is actually random, they cry foul, because it doesn't fit their definition of 'random'.

It's also a very difficult thing to objectively realize, because the human mind is designed to remember notable occurrences more than others, so the times you lost games because of bad draws stick out in your mind much more than all the games where the draws were enough in your favor that you won or at least put up a good fight, thus creating the false illusion that more games than usual had these odd bad draws.

[Trevor]

When a perfectly random shuffle is called "garbage", there isn't much more I can do to try to convince people like that. That's like complaining that water is too watery.

im not the only one who complains about the shuffling

Lots of people thought the earth was flat, too.

I think people will complain about the shuffling algorithm until I implement a shuffling algorithm that stacks the deck exactly how they want it.

[Cyrus]

I think people will complain about the shuffling algorithm until I implement a shuffling algorithm that stacks the deck exactly how they want it.

This, sadly, is the truth. I've never been on a online-ccg program's forum (such as this one) without seeing at least one if not many threads about the related shuffler.

For a good test of how random a randomizer really is, pull up an itunes or similar music library and put it on "random" (not "shuffle", if the program has two settings, because some programs shuffle functions essentially "stack the deck"). Even with almost 20gbs of music I hear the same song within less than an hour of listening. If that didn't happen, it wouldn't really be random, would it?