"Pile" Shuffling

[XDarkAngelX]

Originally Posted by tysith

If you only play online,

Was going for the dramatic. I actually do play in *serious* events, just not FNMs, where people just expect to be able to weave without nobody telling them how cheaty cheaty they are.

Originally Posted by tysith

why the hell did you even comment on this discussion and label us all cheaters, and yes you did label all MTG players

Nope. At high level events people actually know how to randomize a deck, funnily enough.

Originally Posted by tysith

because of our "ignorance" of how to properly "randomize" our decks (and without actually providing any information on how to properly randomize, if it is truly possible!).

Riffle shuffle. At least seven times. Already mentionned. If you weave and see a difference post-shuffle, improve your technique cuz it's not working.

Originally Posted by tysith

And man, if you think a computer can truly come up with a random shuffle, you need to study up on electronics and not math.

Annnnd another pseudo-computer expert hangs himself with the little rope he has.

The algorithm Leaping Lizard used (and disclosed) for MTGO has been demonstrated to give a random deck shuffle.

for x = 1 to n
switch card[x] with card[random(x+1 to n)]
end for

The issue is with the RNG, i.e. the random() function, which every person who knows anything knows it's by default (if you use the ones in the official library) a pseudo-random generator. However, yes, that means even LL's most incompetent programmer know about that fact. I don't know what RNG they use but unless you have actual reasons to doubt that they used something truly random (like, say, the least significant bits in a packet reception time... the simplest method IMO for an online game) then there's really no reason to be suspicious. VASSAL and other online play solutions have access to RNG requests to a server which uses either the lowest significant bits from an electronic thermometer, or the input from a measurement of radioactive decay (which is truly random). Both methods have been used. These are also highly random methods.

PGP uses mouse movement to generate a random key (again they use the least significant bytes, since those are truly random even if you do non-random gestures).

Alright lets face it people, there is absolutely no way to truly randomize our decks before a match unless we pile shuffle 56 times, riffle 67 times, normal shuffle 125, let three different vegas dealers have a go, get the head juge to check for any clumpages and provide a reference from an police officer that explicitly states that Johnny, Combo Player, is a truly trustworthy player and above reproach.

Straw. Needs more straw.

Around 7 riffle shuffles should be enough. And pile shuffling 56 times is as useless as 1 time.

Originally Posted by tysith

Riffle shuffling does not randomize a deck any different than a pile shuffle if used on it's own.

That is laughable. Pile shuffle does not randomize a deck at all. Even a single riffle shuffle is more randomization than that, and you need at least seven to have a fully randomized deck.

Originally Posted by tysith

Plus, if riffles are to become the norm for tourny play, how do you handle a situation where a player does not have the dexterity to riffle shuffle?

The 'mash' can be used as more or less a poor man's riffle. You can do a couple 'mash' shuffles for each riffle you would do normally. I don't have the dexterity for riffle but I have never seen a problem of poor randomization from the 'mash'.

Originally Posted by epeeguy

at least one of them (reducing the initial bias in a deck introduced from having certain groups of cards together after a game) is actually a very legitimate reason in order to further reduce the potential or presence of bias in a deck after it has been sufficiently randomized.

That's not at all legitimate and I fear having to go to any event where you're a judge. This does NOT remove a bias. It creates a NEW bias, one in which the cards from the previous games are LESS likely to be together than expected by chance. If you can't see how that could be used to gain an edge in a game, you're lacking in imagination.

Example: In my previous game I got 3 Incinerate and some creature in my graveyard. I pile shuffle and then poorly randomize my deck. This means that if I draw an Incinerate, it would be right to play the game by assuming I am much less likely to draw a second one than expected by chance, and assuming I am slightly less likely to draw that creature.

[burek]

arkAngelX, maybe you should watch this: http://www.youtube.com/watch?v=L5FoKImUjks

I guess you could say that Saito is a high-level player, right?

[KainXavier]

Originally Posted by tysith

KainXavier, I'm not clicking on any external links.

Do you know who Mike Flores is? Y'know... the guy that writes articles for Wizards? It's his personal blog.

Originally Posted by tysith

Riffle shuffling does not randomize a deck any different than a pile shuffle if used on it's own. I would object to an opponent only riffling as you can still stack a deck and mana-weave with this method.

Ignorance is bliss I guess. I'm not going to copy/paste the entire blog. #1: It's his blog, and #2: It's pretty lengthy, and had visuals to aid the comprehension of what you're doing (read: not doing). But when substantial information is being provided opposing your theory, and you simply ignore it, you end up just looking like a fool. The world is still flat, yes?

[tysith]

Was going for the dramatic. I actually do play in *serious* events, just not FNMs, where people just expect to be able to weave without nobody telling them how cheaty cheaty they are.

So, you mis-quoted yourself or something?

Nope. At high level events people actually know how to randomize a deck, funnily enough.

And you personally checked all the decks?

Riffle shuffle. At least seven times. Already mentionned. If you weave and see a difference post-shuffle, improve your technique cuz it's not working.

We know your mentioned riffles, but you did not provide any evidence that it truly randomizes a deck better than a pile shuffle.

Annnnd another pseudo-computer expert hangs himself with the little rope he has.

I'm talking electronic engineering, not computers, if you have studied electronics, I have by the way, you will see the flaw in randomizing this way.

Straw. Needs more straw.

I know what you need more of and it isn't straw!


Mods, please lock this thread as it has gone way off-topic with self-righteous idiots labeling us all cheaters.

[discoransom]

2 Riffles
2 Clumpy Mash
2 Riffles
2 Clumpy Mash
2 Riffles
1 Clumpy Mash
1 Riffle
Cut
Present

Less than a minute, random as it needs to be. 1 Pile shuffle of each deck at start of first game. Sometimes I pile shuffle to give my head time to analyze previous game.

[epeeguy]

Originally Posted by XDarkAngelX

This does NOT remove a bias. It creates a NEW bias, one in which the cards from the previous games are LESS likely to be together than expected by chance. If you can't see how that could be used to gain an edge in a game, you're lacking in imagination.

The bias it introduces is closer to the expected norm for a deck, especially in comparison to a deck where the majority of certain types of cards have made it to the battlefield and are in one large "clump" when it is put back onto the deck. The same is true for cards that have made it to the graveyard or are left in your hand. In that respect, I believed it safe to say "reduces the bias". If you wish to pick that nit, then that's fine and I suppose I should have been more clear on the direction I was getting at.

In fact, it doesn't take a lot of work to show the difference between these initial starting points and the end results. For example:

(1) Take a deck, put all the "Lands" in a group, all the "Creatures" in a group and everything else in its own group. Place each group together, then perform seven riffle shuffles. Look at the end result.

(2) Take a deck, sort with "Land - Spell - Spell" for its entirety. Then perform seven riffle shuffles. Look at the end result.

Compare the bias in each versus the starting bias in the deck.

[XDarkAngelX]

Originally Posted by epeeguy

The bias it introduces is closer to the expected norm for a deck

It is closer to what people (and apparently, judges too) *naively* expect to be the norm for a deck.

Originally Posted by epeeguy

(1) Take a deck, put all the "Lands" in a group, all the "Creatures" in a group and everything else in its own group. Place each group together, then perform seven riffle shuffles. Look at the end result.

(2) Take a deck, sort with "Land - Spell - Spell" for its entirety. Then perform seven riffle shuffles. Look at the end result.

Compare the bias in each versus the starting bias in the deck.

Bias at the end should be inexistent. If it's not, learn to shuffle.

We know your mentioned riffles, but you did not provide any evidence that it truly randomizes a deck better than a pile shuffle.

Start from any configuration. Note it down.

Pile "shuffle". Give me the number of piles you've done and the order in which you organized the piles. I will give you the exact order of cards in your deck. Totally 100% non-random.

Start from any configuration. Note it down.

Take the worst riffle shuffler in the world. Have him do ONE riffle shuffle. Even in the worst ones where huge clumps of cards will stick together, I will be unable to predict the exact order of cards. So, at least *some* randomness.

There. Wasn't that hard.

[SpiritMirror]

Originally Posted by tysith

Mods, please lock this thread as it has gone way off-topic with self-righteous idiots labeling us all cheaters.

There is some seriously delicious irony in someone who clearly doesn't understand math or randomness calling others idiots.

[Memnarch]

Rather than let this fire continue and throw red cards all over the place, I'm closing the thread per OP request.