No clue how I ended up here, never played MTG, but I agree with everything you said. The computer can create a shuffle that you would never experience during typical human play, so it's not exactly the same game. If flipping the deck is not valid, then neither should be any other inhumanly perfect shuffle.
Assigning random positions (akin to drawing balls out of a bag) is the perfect shuffle, but as you point out, a perfectly random shuffle can totally reorganize the deck - including putting all land at the top, albeit it is extremely unlikely. Humans could theoretically do a perfectly random shuffle too, if they did something akin to drawing random cards out of a bag one at a time. But what we instead tend to do are light, somewhat random, reorganizations by interleaving the stack a couple of times. If I understand you correctly, you are essentially asking for the shuffler to do something akin to distributing land equally and then doing a light random reorganization of the deck, thus closer emulating sloppy human shuffles and making total reorganizations impossible. I'm not going to judge whether that is good or bad, just clarifying some stuff.
th-cam.com/video/iRlZsmSxXUs/w-d-xo.html you and a bunch of people seem to be interpreting it that way so i will be sure to clarify in the supercut, my shuffle absolutely DOES NOT distribute the lands and then organize based on that. It literally mimics the way a person shuffles by doing a riffle, then chunking 3-5 chunks, then 3 more riffle shuffles, then it repeats that whole process, exactly like you would if you were playing a paper mtg game. doing this distributes the cards in a way that a real shuffle by a real person would. it has absolutely no pre determined distributions, and only asks for a number of cards, and a number of lands. because of this you can see that the MTGA shuffler produces results which are statistically outside the realm of what a human could do when they shuffle
@@DownSouthBeef I think I understood your shuffle part correctly - your 1+3+1+3-riffle is what I would call a light random reorganization, compared to a perfect shuffle where every card gets assigned a random position. The difference is that the riffle shuffle maintains some of the original order and just reorganizes "a bit", while the perfect shuffle determines a new complete order completely independently of how the cards were before. Since you mention what a human could do, I think it bears repeating that a human could absolutely, 100%, do what the MTGA shuffler does. The process is literally just numbering each card in your deck, then drawing a sequence of those numbers one at a time out of a bag, blindly, to determine the new ordering. Perfectly random complete new ordering, no bias. The point being that the MTGA shuffling is consistent with what a human would produce IF the human did a truly random, throrough, pull-from-a-bag shuffle. What really puzzles me is your clarification about NOT distributing lands evenly before the proposed riffle shuffles. Since the difference between riffle and perfect shuffle is that the riffle maintains a bit of the original order, the obvious question is what order do you propose that the cards are, or should be, in, prior to the riffle shuffles? Because the prior matters - it is the whole reason that (non-thorough) human riffle shuffling produces "pleasant" results.
@@DownSouthBeef You should try to check how different your human like shuffle is from actual random, you might find the difference is not that much. One way to do this is to label the cards 1-60 before the shuffle, then do your shuffle and record what numbered cards are in what positions and then repeat this a few thousand times and see if its distributed uniformly or not. If you find it is more and more uniform as you increase the sample size you likely have a shuffler that is effectively the same as just randomizing the cards, however if you find that the distribution is not uniform, (for instance the first few cards are consistently more likely to be shuffled to the middle or something to this effect), then your shuffler is actually achieving something and is maybe doing something that a human would do.
@@DownSouthBeef All of this is sort of based on the idea that people don't actually shuffle their cards very well in real life hahaha. That might be true because from what I have seen, you need approximately log_2 (x) or more bridge shuffles for a deck with x cards to make it effectively randomly ordered, (so for 100 cards we are talking 8 shuffles)
Respectfully, I think that having the arena shuffler give that sort of protection (preventing long chains of lands/nonlands in a row) would be a mistake from the perspective of responsible deckbuilding. If I know that the shuffler will do it's best to give me a land every so often, it becomes a lot easier to build decks with a lower landcount to abuse that mechanic. I cannot help but worry that your "let the players have fun" turns into "I'm building a 10 land mono-red deck because now I know that I'll likely still hit that 3rd land and I can have more burn spells." I know that there are people who are shooting angles on the ELO system for Arena already, especially in Brawl. Their decks are built to try and abuse the systems WOTC has in place on arena to keep games fun and competitive. I understand the frustration you're facing but I feel like it is better to have a responsible deck hit a failrate slightly higher than the human average than to make the card game we all know and love into a "trick the code" game where if you succeed you can build a deck that would never work in paper because of probabilities and achieve platinum status.
That’s absolutely not what I’m saying at all and really demonstrates the fact that wizards will be able to continue getting away with this for the foreseeable future. If you look at my newest video reaching mythic, I explained exactly this. It is already happening so the thing that you’re saying, does not apply to what I said. At this point, I’ve literally spent more than half an hour trying to explain to people that arena produces results which are not realistic. A 10 land deck would not work in paper, and it would not work with the shuffling that I used to demonstrate proper distributions either. Unless everything is just a one drop and you managed to Mulligan that one land. That’s outside the point, you are focused on the wrong issue.
That’s not what is happening. My interpretation is that when you shuffle a deck in paper, you aren’t literally assigning a random spot to each card, that just doesn’t happen. When you shuffle a deck of cards, riffle or pile, you splice them between each other
@@DownSouthBeef Contrary to what you might think, even though humans do not rely on a pseudo random generator, humans are really bad at emulating standard random processes (such as dice roll or coin tosses). For example, if asked to produce a random sequence of 4 dice rolls in their head, humans will be way more likely to pick, say, 3415 than 6666 because they see 6666 as very un-random, whereas real world dice do not have such a bias.
You keep commenting like you watched the whole video but it's obvious you didn't. MTGA is known to "smooth out" the selection you receive, and therefore none of the points being made apply, because they are CUTTING THE TAILS OFF OF THE NORMAL DISTRIBUTION until they decide you need to lose, which is why so many people recognize the enforced 50% win rate. Please stop talking about a topic which you are not a part of. Thank you for your contributions up to this point, but they are NOT on topic
No clue how I ended up here, never played MTG, but I agree with everything you said.
The computer can create a shuffle that you would never experience during typical human play, so it's not exactly the same game.
If flipping the deck is not valid, then neither should be any other inhumanly perfect shuffle.
Assigning random positions (akin to drawing balls out of a bag) is the perfect shuffle, but as you point out, a perfectly random shuffle can totally reorganize the deck - including putting all land at the top, albeit it is extremely unlikely.
Humans could theoretically do a perfectly random shuffle too, if they did something akin to drawing random cards out of a bag one at a time. But what we instead tend to do are
light, somewhat random, reorganizations by interleaving the stack a couple of times.
If I understand you correctly, you are essentially asking for the shuffler to do something akin to distributing land equally and then doing a light random reorganization of the deck, thus closer emulating sloppy human shuffles and making total reorganizations impossible. I'm not going to judge whether that is good or bad, just clarifying some stuff.
th-cam.com/video/iRlZsmSxXUs/w-d-xo.html
you and a bunch of people seem to be interpreting it that way so i will be sure to clarify in the supercut, my shuffle absolutely DOES NOT distribute the lands and then organize based on that.
It literally mimics the way a person shuffles by doing a riffle, then chunking 3-5 chunks, then 3 more riffle shuffles, then it repeats that whole process, exactly like you would if you were playing a paper mtg game. doing this distributes the cards in a way that a real shuffle by a real person would. it has absolutely no pre determined distributions, and only asks for a number of cards, and a number of lands.
because of this you can see that the MTGA shuffler produces results which are statistically outside the realm of what a human could do when they shuffle
@@DownSouthBeef I think I understood your shuffle part correctly - your 1+3+1+3-riffle is what I would call a light random reorganization, compared to a perfect shuffle where every card gets assigned a random position. The difference is that the riffle shuffle maintains some of the original order and just reorganizes "a bit", while the perfect shuffle determines a new complete order completely independently of how the cards were before.
Since you mention what a human could do, I think it bears repeating that a human could absolutely, 100%, do what the MTGA shuffler does. The process is literally just numbering each card in your deck, then drawing a sequence of those numbers one at a time out of a bag, blindly, to determine the new ordering. Perfectly random complete new ordering, no bias. The point being that the MTGA shuffling is consistent with what a human would produce IF the human did a truly random, throrough, pull-from-a-bag shuffle.
What really puzzles me is your clarification about NOT distributing lands evenly before the proposed riffle shuffles. Since the difference between riffle and perfect shuffle is that the riffle maintains a bit of the original order, the obvious question is what order do you propose that the cards are, or should be, in, prior to the riffle shuffles? Because the prior matters - it is the whole reason that (non-thorough) human riffle shuffling produces "pleasant" results.
th-cam.com/video/oczs_7DEWyQ/w-d-xo.html
@@DownSouthBeef You should try to check how different your human like shuffle is from actual random, you might find the difference is not that much. One way to do this is to label the cards 1-60 before the shuffle, then do your shuffle and record what numbered cards are in what positions and then repeat this a few thousand times and see if its distributed uniformly or not. If you find it is more and more uniform as you increase the sample size you likely have a shuffler that is effectively the same as just randomizing the cards, however if you find that the distribution is not uniform, (for instance the first few cards are consistently more likely to be shuffled to the middle or something to this effect), then your shuffler is actually achieving something and is maybe doing something that a human would do.
@@DownSouthBeef All of this is sort of based on the idea that people don't actually shuffle their cards very well in real life hahaha. That might be true because from what I have seen, you need approximately log_2 (x) or more bridge shuffles for a deck with x cards to make it effectively randomly ordered, (so for 100 cards we are talking 8 shuffles)
Respectfully, I think that having the arena shuffler give that sort of protection (preventing long chains of lands/nonlands in a row) would be a mistake from the perspective of responsible deckbuilding. If I know that the shuffler will do it's best to give me a land every so often, it becomes a lot easier to build decks with a lower landcount to abuse that mechanic. I cannot help but worry that your "let the players have fun" turns into "I'm building a 10 land mono-red deck because now I know that I'll likely still hit that 3rd land and I can have more burn spells." I know that there are people who are shooting angles on the ELO system for Arena already, especially in Brawl. Their decks are built to try and abuse the systems WOTC has in place on arena to keep games fun and competitive. I understand the frustration you're facing but I feel like it is better to have a responsible deck hit a failrate slightly higher than the human average than to make the card game we all know and love into a "trick the code" game where if you succeed you can build a deck that would never work in paper because of probabilities and achieve platinum status.
That’s absolutely not what I’m saying at all and really demonstrates the fact that wizards will be able to continue getting away with this for the foreseeable future. If you look at my newest video reaching mythic, I explained exactly this. It is already happening so the thing that you’re saying, does not apply to what I said.
At this point, I’ve literally spent more than half an hour trying to explain to people that arena produces results which are not realistic. A 10 land deck would not work in paper, and it would not work with the shuffling that I used to demonstrate proper distributions either. Unless everything is just a one drop and you managed to Mulligan that one land. That’s outside the point, you are focused on the wrong issue.
That’s not what is happening. My interpretation is that when you shuffle a deck in paper, you aren’t literally assigning a random spot to each card, that just doesn’t happen. When you shuffle a deck of cards, riffle or pile, you splice them between each other
@@NattyD_Fergulicious a computer's pesudo RNG is inherently less random than an actual shuffle
@@DownSouthBeef Contrary to what you might think, even though humans do not rely on a pseudo random generator, humans are really bad at emulating standard random processes (such as dice roll or coin tosses). For example, if asked to produce a random sequence of 4 dice rolls in their head, humans will be way more likely to pick, say, 3415 than 6666 because they see 6666 as very un-random, whereas real world dice do not have such a bias.
@@nowinnablewars4480 th-cam.com/video/Xiq_hLrscdE/w-d-xo.html
This may be the exact phenomenon of human shuffling you are asking for: th-cam.com/video/RIGJH12vVCY/w-d-xo.html
You keep commenting like you watched the whole video but it's obvious you didn't.
MTGA is known to "smooth out" the selection you receive, and therefore none of the points being made apply, because they are CUTTING THE TAILS OFF OF THE NORMAL DISTRIBUTION until they decide you need to lose, which is why so many people recognize the enforced 50% win rate. Please stop talking about a topic which you are not a part of. Thank you for your contributions up to this point, but they are NOT on topic