"Pile" Shuffling

[epeeguy]

Originally Posted by onlainari

This is why I'd have a word, I want to know why they mana weave, and what you've just told me has set off alarm bells. Mana weaving is actually improving your draws. You get less clumps. That's cheating. If you told me that, I would be quite okay looking at your deck after presenting it to an opponent.

Provided the player has not done anything inappropriate during the shuffling process described (such as being aware of the order of the cards as the shuffling is taking place and somehow maintaining the order of certain cards), the initial sort isn't relevant nor is the "appearance" of the deck itself relevant after the shuffling has been completed. What matters is the procedures the player used to shuffle the deck before presentation.

Especially as breaking up groups of cards that may been "together" in the graveyard reduces that "togetherness" when the deck is being shuffled before presentation. The concept is discussed here:

http://forums.mtgsalvation.com/showp...1&postcount=52

The concept isn't an easy one to grasp, but if the goal is presenting a sufficiently randomized deck that the player has no reasonable way of knowing the order of the deck (other than clumps of cards may have been broken up), there is no issue.

[Solaran_X]

Originally Posted by XDarkAngelX

In the meantime, you might not mean to be, but you're a cheat. I know, the vast majority of you don't *want* to be cheaters. You are. I'm not going to sugar coat it for you, because you're complacent into your culturally accepted cheating and you need a good shock therapy. You don't want to cheat, but you're currently a cheater. Now do something about it. Improve your shuffling techniques, learn about probabilities (so you're not confused into thinking 'random' means 'uniform', and you stop thinking land clumps means your deck isn't properly shuffled), and stop being a cheater.

You CAN be a cheater from ignorance, or inaction.

So by following the published Floor Rules to the letter, I'm a cheater?

By doing exactly what the DCI itself says is legal before the proper shuffling and presentation for cutting, I'm a cheater?

I should just let my deck clump and work against me on purpose in long tournaments, and completely ignore a legal and available method to prevent that and thus encourage my deck to perform as intended?

What I do is completely rerandomize (to the best of my ability) my deck between matches, or games. This is because after the conclusion of a game, when you pick up your cards you tend to clump creatures, permanents, and lands together into their own groups when cleaning up the board. If you do then, and then just toss your graveyard with them and "reshuffle", your deck is not randomized due to the clumping. Yes, shuffling reduces this clumping a bit - but it still exists.

My shuffling method eliminates the post-game clumping that happens to decks, by restoring it to the pre-game state of randomness it was in before that game.

[epeeguy]

Originally Posted by Solaran_X

So by following the published Floor Rules to the letter, I'm a cheater?

By doing exactly what the DCI itself says is legal before the proper shuffling and presentation for cutting, I'm a cheater?

I should just let my deck clump and work against me on purpose in long tournaments, and completely ignore a legal and available method to prevent that and thus encourage my deck to perform as intended?

The problem is, and it's a rather persistent one unfortunately, is that what you are describing gives the "appearance" of something being "wrong". Because the actual effect on the deck isn't well understood by most, as well as the ease that it can give people who perform the actual manipulation during the shuffling process itself. Hence why people will call this Cheating--Manipulation of Game Materials or Tournament Error--Insufficient Randomization. And why, if you want to do it, you should ask the Head Judge of the event to make sure this won't be a problem during the event.

I know enough to know not to get too hung up on the initial sorting, and focus instead on the shuffling (where the real problems are likely to occur). I'd invite others to do the same.

[XDarkAngelX]

Originally Posted by Solaran_X

So by following the published Floor Rules to the letter, I'm a cheater?

Yes. I don't care what are the written rules. You're not properly randomizing your deck. The proof is in your own assertion that weaving reduces clumping and that the effect wears off over the course of a tournament. If you were shuffling properly that wouldn't be true.

Originally Posted by Solaran_X

I should just let my deck clump and work against me on purpose in long tournaments, and completely ignore a legal and available method to prevent that and thus encourage my deck to perform as intended?

It is not a legal method. You are coasting on the fact that your stacking is impossible to prove by an official. It is still stacking. It is still insufficient randomization.

Originally Posted by Solaran_X

What I do is completely rerandomize (to the best of my ability) my deck between matches, or games. This is because after the conclusion of a game, when you pick up your cards you tend to clump creatures, permanents, and lands together into their own groups when cleaning up the board. If you do then, and then just toss your graveyard with them and "reshuffle", your deck is not randomized due to the clumping. Yes, shuffling reduces this clumping a bit - but it still exists.

Of course it still exists. A random deck has clumps. If it has no clumps, or less clumps than expected by chance, it is not random. It is stacked.

Originally Posted by Solaran_X

My shuffling method eliminates the post-game clumping that happens to decks, by restoring it to the pre-game state of randomness it was in before that game.

Randomness: I don't think that word means what you think it means. See my example of random vs uniform earlier in the thread. Random = clumps. If you don't have clumping your deck isn't random. You are profiting from deck stacking.

IF you were randomizing properly, whether you weaved or not, whether you put all your cards post game in land/spell/creature clumps or not, would have NO effect. That step would thus be superfluous, useless. The fact that you use that step, and that it benefits you *per your own claims*, PROVES you are not properly randomizing your deck and PROVES you are cheating.

Weaving is either useless OR cheating. No buts, no ifs. If you're using it I consider your deck suspect, because someone who gleaned no benefit from it would have foregone using it as a waste of time. I WILL riffle shuffle it multiple times (I already do anyway, but whatever) so that it is properly randomized, and your weaving advantage is nullified.

Even if the only advantage you gain is purely superstitious, because the rest of your routine does properly randomize your deck, I consider that shady (from an ethical standpoint). You're actually *wishing* that you were gaining unfair advantage through deck stacking over me. The fact you're not *actually* doing it does not nullify intent.

[epeeguy]

Originally Posted by XDarkAngelX

Yes. I don't care what are the written rules.

I've previously cited the DCI policy, as written in the Magic Infraction Procedure Guide, and would remind everyone to read what it says. As explicitly stated, any kind of weaving or manipulation prior to shuffling is allowable. What matters is the procedure the player uses when actually shuffling his or her deck before it is presented.

If you are deliberately choosing to ignore the cited policy, then this is a disservice to everyone reading the thread and contributes to spreading disinformation. Especially as any kind of enforcement should be done based on that stated policy.

Originally Posted by XDarkAngelX

It is not a legal method.

It is a legal method. Your personal opinion is moot, when the stated policy is otherwise.

Originally Posted by XDarkAngelX

A random deck has clumps. If it has no clumps, or less clumps than expected by chance, it is not random. It is stacked.

The presence, or absense, of "clumps" does not prove anything. An insufficiently randomized deck can have "clumps" just as easily as a sufficiently randomized one. "Chance" or "randomness" proves nothing in this regard.

Originally Posted by XDarkAngelX

Even if the only advantage you gain is purely superstitious, because the rest of your routine does properly randomize your deck, I consider that shady (from an ethical standpoint). You're actually *wishing* that you were gaining unfair advantage through deck stacking over me. The fact you're not *actually* doing it does not nullify intent.

Both the act and the intent are relevant. The fact that the shuffling does sufficiently randomize the deck is sufficient for all intents and purposes. Do not get hung up on the initial manipulation of the deck.

[bcclar77]

Dark Angel - while all of your assertions are mathematically correct, they're not that realistic. You know and I know that everyone sitting in a tournament room is playing with a deck which is not mathematically random. I know it's a pet peeve of yours (and mine too, to an extent), but the DCI acknowledges that mathematic randomization is virtually impossible, so their intent is to just make people riffle enough that it is, in their words "sufficiently random" (which I agree with you, is not random).

I agree with you, really, but it's a losing fight. Random, in this case, is unobtainable. It's akin to the fact that there's no such thing as "Black" (because everything black relects some light) or a "Vacuum" (because even in outer space, there's at least one particle of something)...

[epeeguy]

Originally Posted by bcclar77

Dark Angel - while all of your assertions are mathematically correct, they're not that realistic.

His assertions aren't even mathematically correct. Read Condor's post (the link I provided earlier in the thread) as to the why.

[bcclar77]

They are using a different definition of "random".

Dark Angel's is what he stated. Condor's definition is that less/smaller clumps = more random, which is the exact opposite of Dark Angel's argument.

Edit - but perhaps I do not understand the definition of "sequences" in condor's post... he doesn't really explain a lot of what he is using.

[llamatron2k]

Originally Posted by epeeguy

His assertions aren't even mathematically correct. Read Condor's post (the link I provided earlier in the thread) as to the why.

Dark Angel's assertion is an incorrect one. I'll take actual numbers from this week's FNM, but I've noticed that mana weaving, or simple deck stacking before a series of honest riffle shuffles/mashes increases randomness.

The reason that this does such, is that during play, certain cards will have a tendency to group together(instants/sorceries in graveyard, mana in play...), and if you simply pile these together and then riffle, you will notice that there is a higher chance for them to present in tightly ordered groups. For instance, when playing my bant deck, I notice that it isn't uncommon for me to draw two rafiqs in close proximity to one another. This is linkage, and it represents a tendency for cards to be next to, or near each other, which is definitely NOT random. While pile shuffling/stacking itself may not increase randomness all that much, it will certainly reduce linkage, which has the effect of increasing randomness.

[bcclar77]

But my point is, that the shuffler on MWS or MTGO is more truly random than we human beings can achieve, and even then it's not true random.

And this concept of "linkage" is impossible to measure. What's to say that the 2 Rafiqs being adjacent is "random" or "linkage"?

[Solaran_X]

Dark Angel:
At a previous FNM, I played someone with their Duel Decks (in this case, Elves vs. Goblins) between rounds. Because it was just for fun, I flipped through the deck very fast to make sure the mana was mixed in, then shuffled and proceeded present for cutting.

First Hand: 0 Lands
Second Hand: 0 Lands
Third Hand: 0 Lands.

I then mana weaved, pile shuffled, and riffled almost a dozen times and redrew to seven with his permission because this was just a fun game.

First Hand: 2 Lands, 5 Spells

So you're telling me the first example is "randomized" despite three consecutive hands, with shuffling in between, of no lands and thus implying that the land is clumped together somewhere, and the second example is "cheating"?

[llamatron2k]

Originally Posted by bcclar77

But my point is, that the shuffler on MWS or MTGO is more truly random than we human beings can achieve, and even then it's not true random.

And this concept of "linkage" is impossible to measure. What's to say that the 2 Rafiqs being adjacent is "random" or "linkage"?

Because at any given time, we can calculate what the percent chance of drawing a rafiq out of a pile is if it were to be truly random.

To see what your tendency to draw rafiqs is, you can then take notes during an FNM while only riffling and never doing another kind of shuffle to see if play affects linkage over time(you'll even have three samples per play period as replications)

If rafiqs are drawn closer/further than what one would expect in a significant manner, a correlation between linkage and exclusively riffling is made. If you chart it and the results trend towards linkage during later games, then we can identify the nonrandomness of play as a cause of linkage.

[MisterKaras]

Originally Posted by Sterling Angel

This is very true. I remember when I used to play on MWS frequently, a lot of people complained about the shuffler. At first I did also, but then I started to think about it and realized the reason people were complaining was because of large clumps in their decks. Now something that is truly random will have some rather large clumps, and the MWS shuffler randomized your deck (as much as any computer program can anyways.) However many people tend to mana-weave their physical deck then shuffle several times, which doesn't fully randomize the deck. Hence people were complaining about the fact that they couldn't mana-weave then semi-randomize their decks. Remember folks, there is a huge difference between random, and evenly distributed or averaged.

I am one of those that complain about the shuffler, but not for clumps, its more the fact that one in 10 games you will draw triple of a card in your deck within the top 15 cards.

I find that to be a major problem with testing as its usally the same 2 cards that get picked for this, I dont know how many games I lost last year to people drawing triple cryptic and thinking nothing of it. Or how many games I've won because I've drawn triple blightning by turn 5

[bLatch]

Originally Posted by llamatron2k

Dark Angel's assertion is an incorrect one. I'll take actual numbers from this week's FNM, but I've noticed that mana weaving, or simple deck stacking before a series of honest riffle shuffles/mashes increases randomness.

NO. A thousand times NO! It does not increase randomness it increases distribution. How many times does it need to be said that an even distribution is not the same as randomized?

Originally Posted by Solaran_X

Dark Angel:
At a previous FNM, I played someone with their Duel Decks (in this case, Elves vs. Goblins) between rounds. Because it was just for fun, I flipped through the deck very fast to make sure the mana was mixed in, then shuffled and proceeded present for cutting.

First Hand: 0 Lands
Second Hand: 0 Lands
Third Hand: 0 Lands.

I then mana weaved, pile shuffled, and riffled almost a dozen times and redrew to seven with his permission because this was just a fun game.

First Hand: 2 Lands, 5 Spells

So you're telling me the first example is "randomized" despite three consecutive hands, with shuffling in between, of no lands and thus implying that the land is clumped together somewhere, and the second example is "cheating"?

The first one is, in fact, randomized *assuming you riffle shuffled properly*. The second one is not necessarily cheating, if the riffles were effective in randomizing. The point is though that if the riffles were effective randomization than your pile shuffling and mana weaving had literally no effect on the outcome. If the mana weaving / pile shuffling affected the outcome, it is not randomization.

It is entirely possible for a truly randomized deck to draw 0 cards three tiems in a row. Its also possible for all of the land in a randomized deck to be touching other land as one big clump. Achieveing an even distribution of land thorughout the deck is not randomizing the deck. its Ordering the deck.

[XDarkAngelX]

Originally Posted by epeeguy

It is a legal method. Your personal opinion is moot, when the stated policy is otherwise.

Weaving prior to a real shuffle is 'legal' in the sense it's not forbidden.

If there is evidence that the shuffling method of the player results in weaving giving a benefit, that means the *shuffling* is not legal. It is not properly randomizing the deck. I'm not saying anything about weaving at that point.

Originally Posted by epeeguy

The presence, or absense, of "clumps" does not prove anything. An insufficiently randomized deck can have "clumps" just as easily as a sufficiently randomized one. "Chance" or "randomness" proves nothing in this regard.

The *sustained* absence of clumps, or reduced probability of presence of such, is evidence of poor randomization. The claims of the OP is that his method of shuffling produces less clumping in a sustained manner, and that it fades away as the tournament (and number of shuffles, i.e. randomness of the deck) progresses forward. A single time is not evidence (too little sampling) which is why it is unprovable to a judge. However if it is a pattern it is strong evidence of cheating.

What judges can or cannot detect is not my problem. It's not mana weaving itself that is illegal, it is the poor shuffling method leading to a stacked deck which is *evidenced* by the fact the mana weaving accomplishes something instead of nothing. And if a player persists in mana weaving, and if they even claim it helps them, I'll treat them as they should be treated: a cheater. They can coast on the fact that you can't afford the time to do statistical analysis over multiple entire tournaments to detect it, but I don't have to treat them as upstanding players when they are cheating and simply doing it in an undetectable manner.

Originally Posted by llamatron2k

This is linkage, and it represents a tendency for cards to be next to, or near each other, which is definitely NOT random. While pile shuffling/stacking itself may not increase randomness all that much, it will certainly reduce linkage, which has the effect of increasing randomness.

Pile shuffling does not merely "reduce linkage". It increases distance (makes the four copies of a card more uniformly distributed in the deck), which is NON random.

If there is a difference in the probability of two Rafiqs being together whether you're pile shuffling or not, you are not properly randomizing your deck and your shuffle is bad. Period. Initial distribution should have no effect on the probability of final distribution.

Originally Posted by Solaran_X

Dark Angel:
At a previous FNM, I played someone with their Duel Decks (in this case, Elves vs. Goblins) between rounds. Because it was just for fun, I flipped through the deck very fast to make sure the mana was mixed in, then shuffled and proceeded present for cutting.

First Hand: 0 Lands
Second Hand: 0 Lands
Third Hand: 0 Lands.

I then mana weaved, pile shuffled, and riffled almost a dozen times and redrew to seven with his permission because this was just a fun game.

First Hand: 2 Lands, 5 Spells

So you're telling me the first example is "randomized" and the second example is "cheating"?

I don't have enough data to go on. You would have to present all 60 cards and we could do an analysis on the amount of clumps you have and how closely this correlates to what you'd expect by chance, and then I'd have to do this over multiple shuffles to have enough data points.

I was just assuming that when you say you have less clumping because of your method (in a sustained manner, over multiple tournaments) that you are saying the truth. In that case you are cheating, yes, and your deck isn't random. Your impression would be defined by multiple trials and have more data points than you're providing me now.

(Finally, 2 lands + 5 spells isn't telling me much. It includes a lot of cases, from 1 land, 5 spells, 1 land, i.e. a 5-clump, to 1 land, 2 spells, 1 land, 3 spells, which are two clumps of 2 and 3 spells, respectively, to 2 lands/5 spells which is two clumps, etc etc etc).

I am one of those that complain about the shuffler, but not for clumps, its more the fact that one in 10 games you will draw triple of a card in your deck within the top 15 cards.

That to me doesn't seem like something out of the ordinary for a *random* deck. I'm assuming 1 in 10 and top 15 are, like, your guesstimations and not serious numbers?

People often don't realize how likely certain events actually are. How many people in a room do you need so that the probability that at least two of them share a birthday is more than 50%? Only 23 people. And how many so that the odds are higher than 99%? Only 57.

How many people so that the odds are greater than 50% that two people have a birthday a week from each other? 7.