"Omg there's 2 of the same card back to back" ignores every other random ordered card in the deck

There was a recent video where a Magic streamer (CovertGoBlue, if it matters to you) was playing a couple games with one of the developers of their client, and asked about their shuffling algorithm, which players have some Gambler’s Fallacies about. And he was saying that, as counterintuitive as it may seem, having odd clumps of cards is actually proof that the shuffle is random. A shuffling algorithm that is intentionally trying to distribute cards “fairly” across the deck is not randomizing the cards at all. People aren’t very good at recognizing true randomness because we’re meant to make sense out of randomness and see patterns inside of it, even when there are none.

Matt Parker once gave a homework assignment to his students to flip a coin 100 times and record the results. Some of them actually flipped the coin (or used a random number generator), while others just wrote 100 Hs and Ts down 5 minutes before the class, in what they assumed to be a random order.

The thing is, Matt could almost always tell who cheated and who didn't: those who actually flipped the coin 100 times would have at least one "clump" of 7+ consecutive heads or tails somewhere in the sequence. Meanwhile, those who "cheated" almost never had more than 3 or 4 of the same result in a row.

For context, I have a masters degree in maths and did my dissertation on card shuffling randomness. There absolutely IS degrees of randomness. You can actually measure randomness of a deck in a couple of ways, some common ones being "Entropy" and "Total variation distance". These are deterministic measures that calculate how much you know about the deck with each shuffle you do. All of that said, Coder is entirely correct that a pile produces ZERO entropy. The entire deck order is known after the "shuffle", presuming you knew the order pre "shuffle".
As for riffle shuffling, to have a suitably "random enough" shuffle, you should shuffle at least 7 to 8 times, with each shuffle after 8 have notably diminishing returns on how much more random the deck gets (note it is still getting more random with each extra shuffle, just by smaller amounts and at some point you're wasting your time to keep going).
"Declumping" is just a measure of cards we recognise having started together being near each other post shuffling. Its pattern recognition, and so we assume just having those cards not be near each other means its more random, but in reality, you're just stacking the deck to the prefered outcome that you don't see those cards near each other (then undoing all your work by insisting on riffle shuffling anyways which actually randomises the deck).

Randomized has never meant Normalized

2:18 I think that's something that tests at school with 'A,B,C,D' choice answers conditionned us to see differently. When you take a tests and you answer the questions then you see you answered like 4 'A' in a row, 'it feels weird, you feel like it's not right, if the order of the questions and which letter is the right answer is truly random, then why did I answer 4 A's in a row?'. It's a fallacy cause there was actually the exact same chance that the order of answers was 'A,A,A,A' or 'C,B,A,D'. It's just that because it's 'the same', kinda like the pink sleeves, it feels not random when in actuality it is.

The only time I “pile shuffle” is after a match to count the cards in my deck to make sure I have the same number of cards as I started with

It's like picking a bunch of dice up and setting them all to 6 in your hand before throwing them: It's wasting time when the dice is going to be random when you roll it. And if you're specifically rolling it in a way that's going to actually depend on how they were before you rolled them you're cheating.

people really can't fathom the idea that random means that there's a inherent change for cards to be close together because it was RANDOM, having engine pieces perfectly spaced out after "shuffling" a deck with the said intent to undo clumps is the contrary of random

Ive had this argument as well as arguing about the master duel coin flip - basically humans inherently arent wired to comprehend statistics and random probability. We see a pattern and immediately go "NOT RANDOM, RIGGED" And almost no amount of explanation will convince them otherwise

If pile shuffling really made a difference with randomization, casinos would implement it.

It HURTS me to see that people as a whole, really doesnt know how randomization works

People arguing that pile shuffling before riffle affects the outcome are essentially arguing that the way you put marbles in a bag before shaking it affects how much the shake randomized the marbles. It doesn't do anything other than waste time.

Oh shit yugioh has mana weaving discourse. Either it does nothing in which case it’s a waste of time or it does something in which case it’s cheating. There is not a third option

When at a "learning magic the gathering" night, everyone there myself included couldn't shuffle well, so the person who set the night up just told us to pile shuffle. It didn't feel right, it was very slow, and genuinely took more time than it should have. I've since found out a reason why that it was hard to riffle shuffle for us was because of the lack of card sleeves. When i got home i tried shuffling with two decks i have, one with cards sleeves that have more sharp edges(i think elestrals card sleeves), and one where the edges are flatter. The sharp edge one i could very easily riffle shuffle with, while the flat edge one was harder.  Long story short if you have trouble shuffling, start out with sharper edge sleeves.

In the digimon tcg tournament rules, there is a rule that states: you can pileshuffle between games but not between sets. For example: pairing is done, you search your seat, and then you CAN pileshuffle your deck. In the round its 1-0 and you want to go to set 2, you CANNOT pileshuffle for less timeconsumption. If someone pilesuffles their deck between sets, you can call judge for it.

they should implement that in every cardgame.

It's hilarious that the most efficient shuffle in the world is already the one that everyone does out the gate, and yet everyone becomes convinced that pile shuffling or some other form is better.

I think doing it that way is good after u make it then regular shuffle after that

To anyone that thinks the 10 pink card shuffle didn’t appear random:   If you ask a human to generate a sequence of 20 1s and 0s and compare it to a true randomization with the same parameters it’s actually pretty easy to tell which is the human and which is the computer. Humans think in patterns and are thus way more likely to space out 1s and 0s evenly whereas “true randomization” has clumps. It’s like putting 3 times in a row on a multiple choice test and thinking “there’s no way it’s 3 Cs in a row”

Thanks so much for this video.

I tried explaining this in a discord chat, but it's so nice to be backed by a well known judge.

It's so frustrating and it's either time wasting or cheating or ignorance. And ignorance shouldn't excuse the former 2