honest mistake or unfair advantage or cheating?
- Thread starter TrEkIeV
- Start date
- Status
NoPoke
Active Member
So I roll a die 100 times. If it is a fair die then I should expect to get between 40 and 60 heads.
Or are you going to continue to insist that is not random?!?
bullados
<a href="http://pokegym.net/forums/showthread.php?
If this post is off-topic, just LMK and I'll delete it.
I'm honestly curious about something. What exactly do you consider to be a "good" shuffling technique? Here's the pre-game technique that I've been using for years without incident...
1) 5-pile. 1-2-3-4-5, until I hit 60. This is mostly done to ensure I've got 60 cards.
2) Take piles 3 and 5, and mash shuffle them at least 7 times (usually closer to 12 or 15, depending on how much time I've got). Call this Pile 6.
3) Take piles 1 and 4 and mash shuffle them at least 7 times (usually closer to 12 or 15). Call this Pile 7.
4) Take Piles 2 and 7 and mash shuffle them at least 7 times (usually closer to 10 or 12). Call this Pile 8.
5) Take Piles 6 and 8 and mash shuffle them at least 7 times. Continue mashing until opponent presents their deck.
6) Present.
This all usually takes about a minute or two
Is this a sufficiently shuffled deck? Or do I need to change my technique?
During games, I usually mash about 7-10 times after a search where I know my next deck manipulation is not going to be a Search. Like, if I'm going to Level, then Ultra, I generally won't shuffle between the two in order to save a bit of time. But if I'm going to Level then Juniper, I'll shuffle after the Level.
The whole reason I mash instead of riffle is that a mash takes significantly less time for me than a riffle. I can mash about 7-10 times in the time it would take me to riffle, and the randomization is close enough between the two that I feel additional mashes make up for the slightly worse randomization. BTW, Mash shuffle.
If my assumptions are wrong, just LMK, and I'll change them.
I'm honestly curious about something. What exactly do you consider to be a "good" shuffling technique? Here's the pre-game technique that I've been using for years without incident...
1) 5-pile. 1-2-3-4-5, until I hit 60. This is mostly done to ensure I've got 60 cards.
2) Take piles 3 and 5, and mash shuffle them at least 7 times (usually closer to 12 or 15, depending on how much time I've got). Call this Pile 6.
3) Take piles 1 and 4 and mash shuffle them at least 7 times (usually closer to 12 or 15). Call this Pile 7.
4) Take Piles 2 and 7 and mash shuffle them at least 7 times (usually closer to 10 or 12). Call this Pile 8.
5) Take Piles 6 and 8 and mash shuffle them at least 7 times. Continue mashing until opponent presents their deck.
6) Present.
This all usually takes about a minute or two
Is this a sufficiently shuffled deck? Or do I need to change my technique?
During games, I usually mash about 7-10 times after a search where I know my next deck manipulation is not going to be a Search. Like, if I'm going to Level, then Ultra, I generally won't shuffle between the two in order to save a bit of time. But if I'm going to Level then Juniper, I'll shuffle after the Level.
The whole reason I mash instead of riffle is that a mash takes significantly less time for me than a riffle. I can mash about 7-10 times in the time it would take me to riffle, and the randomization is close enough between the two that I feel additional mashes make up for the slightly worse randomization. BTW, Mash shuffle.
If my assumptions are wrong, just LMK, and I'll change them.
Last edited:
Otaku
Active Member
Proving once again how contrary I can be, while I agree with your sentiments in the full context of the discussion, the statement was sound. He wasn't talking about wagering - he was talking about the simplest, non-invasive method of determining if a die is weighted.
While a legal die can indeed significantly differ from coming up "even" 40 to 60 % of the time (though perhaps as we are discussing dice we should count how often it lands on specific facings), the main goal of this is to verify that a die isn't weighted, as opposed to proving a die is.
When playing the game, I heartily agree; a legal, successful (e.g. not rolling off the table into a crack :lol
flip or roll will be "heads" or "tails". A good player will plan for either result, both in deck building and in play, and woe to him or her if the game comes down to a single coin toss.
NoPoke
Active Member
Throw 100 fair die or flip 100 fail coins and the number of heads or tails will be between 40 and 60 95% of the time.
I really don't get how the idea behind this is hard.
I really don't get how some players can claim that random does not come with testable, predictable statistics.
After shuffling your deck if you are not seeing pairs of the same named card together on first search most of the time then you are not shuffling sufficiently. If you are seeing triples or quads of the same named cards anything more than once or twice on first search in a single game swiss tournament then you are not shuffling sufficiently.
Note that with a lot of players in a room someone is going to get the outlier behavior and if they follow the above advice will end up shuffling more even when their shuffle is adequate. But that is a much safer outcome than players performing inadequate randomisation.
==
Bulados: I think I need to explain how to count rising sequences. That is going to be the easiest way for you to check on how your mash shuffles perform.
Not that anyone cares but the minimum number of riffles from Prof Diaconis's work is Ceil(3/2*log_base2(deck_size))
I really don't get how the idea behind this is hard.
I really don't get how some players can claim that random does not come with testable, predictable statistics.
After shuffling your deck if you are not seeing pairs of the same named card together on first search most of the time then you are not shuffling sufficiently. If you are seeing triples or quads of the same named cards anything more than once or twice on first search in a single game swiss tournament then you are not shuffling sufficiently.
Note that with a lot of players in a room someone is going to get the outlier behavior and if they follow the above advice will end up shuffling more even when their shuffle is adequate. But that is a much safer outcome than players performing inadequate randomisation.
==
Bulados: I think I need to explain how to count rising sequences. That is going to be the easiest way for you to check on how your mash shuffles perform.
Not that anyone cares but the minimum number of riffles from Prof Diaconis's work is Ceil(3/2*log_base2(deck_size))
signofzeta
New Member
If you flip a coin, the coin is either heads or tails. If you roll 1 6 sided die, it is a 1, 2, 3, 4, 5, or 6. In this case, we are talking about one game here, and not many games. Your whole idea of flipping a coin and expecting 40 to 60 heads only applies when you are flipping many coins. We aren't shuffling many times here. We are shuffling once. People can shuffle many times before the match, but the only result that matters is the order of cards that is used at the beginning of the game.
If you shuffle a deck, you can expect 60! different orders of cards. Any of these orders of cards CAN happen. Nothing SHOULD happen when it is random.
Just because it is more likely that certain orders of cards, such as clumps of 2 or 3 are supposed to happen doesn't mean that the person cheated when they don't have clumps of 2 or 3.
This is what I think you are trying to say. Imagine a dice that has a red face, and the rest is a white face. You are saying this. It is more probable to roll a white face, so you are doing it wrong when you the dice lands with the red face up, assuming that the die is a 100% fair die. That's whats you are saying.
Am I not right by saying that you are saying that your deck order contains no clumps of 2 or 3, or many clumps of 4, which can happen when it is random, therefore you are shuffling wrong, and you should shuffle again until you get the more probable result?
In random, there is 60! permutations, and each permutation will happen close to equal. Any shuffling technique can never achieve this. That is why MTG has sufficiently random, and why fully random is impossible.
In the end, when shuffling, don't think. Just shuffle and play. If you are thinking about your own order of cards, you have an intent to cheat. If you are thinking about your opponent's order of cards, you just want to win by default. The only time you know the opponent is cheating is if they look at their cards, and reorders them, and does a bad shuffle afterwards. So long as they don't look at their cards, they won't know where each card will end up. So long as one single card moves around the deck 7 to 12 times during a shuffle, it is good enough. People want to play, not spend 30 minutes shuffling.
If you shuffle a deck, you can expect 60! different orders of cards. Any of these orders of cards CAN happen. Nothing SHOULD happen when it is random.
Just because it is more likely that certain orders of cards, such as clumps of 2 or 3 are supposed to happen doesn't mean that the person cheated when they don't have clumps of 2 or 3.
This is what I think you are trying to say. Imagine a dice that has a red face, and the rest is a white face. You are saying this. It is more probable to roll a white face, so you are doing it wrong when you the dice lands with the red face up, assuming that the die is a 100% fair die. That's whats you are saying.
Am I not right by saying that you are saying that your deck order contains no clumps of 2 or 3, or many clumps of 4, which can happen when it is random, therefore you are shuffling wrong, and you should shuffle again until you get the more probable result?
In random, there is 60! permutations, and each permutation will happen close to equal. Any shuffling technique can never achieve this. That is why MTG has sufficiently random, and why fully random is impossible.
In the end, when shuffling, don't think. Just shuffle and play. If you are thinking about your own order of cards, you have an intent to cheat. If you are thinking about your opponent's order of cards, you just want to win by default. The only time you know the opponent is cheating is if they look at their cards, and reorders them, and does a bad shuffle afterwards. So long as they don't look at their cards, they won't know where each card will end up. So long as one single card moves around the deck 7 to 12 times during a shuffle, it is good enough. People want to play, not spend 30 minutes shuffling.
Last edited:
NoPoke
Active Member
What? In your pokemon career you have only ever shuffled ONCE???
And again with a sufficient shuffle lots of things not only should happen but do happen. Further I have never said that a lack of pairs means cheating. If a player shuffles sufficiently then they should expect to see pairs. Very different statement!
In the end players absolutely should be thinking about how they shuffle and in particular if what they are doing leads to sufficient randomisation or to stacking.
And again with a sufficient shuffle lots of things not only should happen but do happen. Further I have never said that a lack of pairs means cheating. If a player shuffles sufficiently then they should expect to see pairs. Very different statement!
In the end players absolutely should be thinking about how they shuffle and in particular if what they are doing leads to sufficient randomisation or to stacking.
Last edited:
signofzeta
New Member
We are talking about one isolated situation here. The other times you need to shuffle isn't part of this discussion. For every isolated situation that you need to shuffle, and as a end result, have a certain order of cards ready before the gameplay starts or resumes.
In random, you don't expect. Oh noes, I don't see any pairs, whatever should I do? In random, there is no such thing as should. You don't expect the higher probable result to happen. You expect any result to happen. If that result happens to be in the low end of the bell curve, then it is a result. You don't discard it just because it isn't the higher probable result.
In the end, that guy was clearly cheating, because he looked at this cards and did a poor shuffle afterwards. If he did riffles after, then he wouldn't have cheated. I also say that achieving full randomness is impossible, and that people should stop wasting time trying to make their own deck or their opponent's deck random. You can't. In any shuffling technique, certain distributions would favor others. The only true way to make things random and to ensure randomness is through shuffling non-stop in a very long time period. Anyone can be paranoid or OCD and move cards around, but so long as the last step in the shuffling process is the actual good shuffle, not something where you do a couple hand over hands, like what that guy did there, then it is all good.
NoPoke
Active Member
In random of course I can have expectations!
but since you say I can't then you will take a $10 bet on a single trial of 100 dice rolls? I win if the number of heads is between 40 and 60? After all it is a single result and anything is possible in a single result? On second thoughts I'm giving you most of the numbers so I pay you $10 and you pay me $20.
It is random so I may lose, that is part of my expectation too.
but since you say I can't then you will take a $10 bet on a single trial of 100 dice rolls? I win if the number of heads is between 40 and 60? After all it is a single result and anything is possible in a single result? On second thoughts I'm giving you most of the numbers so I pay you $10 and you pay me $20.
It is random so I may lose, that is part of my expectation too.
signofzeta
New Member
No. From either this thread, or some other thread, you said that the deck should have pairs of 2 clumps, or something like that, if it doesn't, you should shuffle again. That is what I heard. I am saying that just because something that is unlikely to happen, happens, it doesn't mean the result should be discarded. I don't think you seem to understand this.
In the end, we are looking for only one result per shuffle. There are times when you don't yield the typical pairs or triples in your deck, but if that kind of result doesn't happen, it is still a result. I probably have lots of pairs, triples, and the occasional 5 energy in a row thing most of the time, it doesn't mean that I should not have the quads, or singles every day or two of playing many games.
I quote this again. The first paragraph is what I call a delay of game. I am so going to take your advice. I shuffle, look at my deck to see if I have your "sweet spot" distribution. It doesn't. I shuffle again. I look at the deck to see if it has the distribution. It does. Since I looked at my cards, I have to shuffle YET AGAIN. Now I check if it has the distribution. It does. I already looked at my cards so I have to shuffle again. Does this make any sense to you?
By the time you are legally allowed to check the order of cards in your deck, your deck would have already changed so much that you can't use that result to gauge if you are sufficiently randomizing your deck.
You must be using some strange definition of "random".
As pertains to a deck, random means "lack of any prearranged order". Shuffling enough to obliterate any prearranged order (where "prearranged order" pertains to the order of the cards at the beginning of the shuffle) is not impossible (or even difficult) and is not wasting time.
NoPoke
Active Member
@signOfZeta: So without looking at your cards how do YOU know that they are sufficiently randomised? Unless you actually examine the results (aka look) you can have no idea how good or bad your shuffling is. What I was giving you, though you are unwilling to accept this, is what to look for when you carry out the important examination. A simple test that involves checking for the presence of pairs and clumps of three or four similarly named cards.
That may be what you heard, but it was not what was said. What was said, is that the next time you do a deck search, if you don't see some sort of clump somewhere, you should shuffle more thoroughly. It's not (just) a sign that your last shuffle was bad and needs to be redone, it's a sign that your shuffling in general is bad and you should feel bad (and shuffle better from then on).
And again, that is not what was said. You should not look and see if your shuffle was good enough after shuffling; you should consider, when possible, whether there is a pattern of bad shuffling, and if there is, improve your technique.
signofzeta
New Member
...and how would you know that without looking at your deck after shuffling? You can't. By the time you get a chance to look at it, the order of the cards would have already been different.
Another thing is, we also have to take into account the make-up of the deck. Not every deck is the same, so no one person can expect certain distributions. Some people have more energy than others. Some people have more 4 ofs, and some decks are singleton decks. Just because certain distributions don't happen more often doesn't make them cheaters.
Dude, that isn't full randomness. Partial random maybe, but not full randomness. It is difficult to achieve full randomness, but it is easy to achieve partial randomness. If one of 60! permunations never happens, it isn't random, but if such scenario exists, and a shuffling technique never yields that one of 60! deck orders, then it is good enough, and people shouldn't be labeled as cheaters because they can't get that one of 60! deck order. In that case, that shuffling technique will sufficiently randomize your deck, but not fully randomize it.
From what I saw in the video, that guy did one hand over hand shuffle twice. That is, he took the middle portion of the deck and put it on top, and then he did it again. That is horrible in terms of randomization. The hand over hand shuffle can randomize the deck, but it takes longers, and is especially used for people who don't use sleeves, or don't know how to riffle, typically little kids. This is how I do it. I do 8 hand over hands, 12 times. Initially, it may not give me a good distribution of random, but the more games I play, it gets more and more "random" and unpredictable. In the end, if you can't expect what your next 20 cards will be in your deck, then you gain no advantage from having a not so random deck. Hence why, nobody should look at their decks after shuffling.
The purpose of shuffling is to make it so that anyone can't predict what their next cards are. If their shuffling technique can achieve this, it is good enough.
In MTG, there are times where I keep on drawing the same stuff every time, and that is when I know something is wrong.
Last edited:
If you don't look at your deck at some point after you shuffle but before you shuffle again, then you don't know if that particular shuffle was good enough. Nobody said you had to look at the result of every shuffle. The point isn't to see whether any given shuffle was good enough, it's to see if you shuffle well enough generally. You don't need every result to consider that, you can just use the ones you have.
signofzeta
New Member
I'm fairly certain Nopoke meant that I should look at every result of my shuffles. He did refute my point where I said you should never look at your cards right after your shuffle, or even during the shuffle. You can look at your cards whenever you want after the game ends, or when you are searching for something via card effect, but you never look at the order of your cards during or right after the shuffle, before your game starts or resumes. If he meant that I should look at any point where I am allowed to look, then why would he refute that point? It doesn't make any sense.
Last edited:
Glumanda
Rulings Compendium Translator
Oh, i care. This means, for a deck size of 60 cards, we have 8.8605?
Bilbo Baggins
New Member
Sounds like what I do pregame matches. Had no incidents either. Also, I add that I rarely look at my deck before 6 pile (I 6, not 5), and if I did, I shuffle first a bit then pile.
- Status