Shuffling and The Math Behind It

Frost Mage

Frost Mage

Pegasususus
Member
Does anyone else feel like the draw system in PTCGO is broken? Consistently I keep getting bad hands, whereas IRL when I play the exact same deck, it's nowhere near as inconsistent. It's like it's designed to give you a bad hand. Is this just me having bad luck or has anyone else experienced this?
 
RE: Terrible draws in PTCGO

Technically speaking it's all random, but real life is going to result in a lot less... random. (Maybe) The reason you get good hands when actually playing is that a lot of your better stuff is probably semi-clumped together, and unless you go through numerous different randomization shuffles it's going to be a lot more biased towards the better hands.

Or that's what I'm gonna' guess, anyways.
 
RE: Terrible draws in PTCGO

I'd argue against that. I can pile shuffle my deck for an hour and still get better hands than in PTCGO most of the time.
 
RE: Terrible draws in PTCGO

Well the randomization algorithms or whatever they use could never perfectly match how a shuffle goes in real life, so it is likely to get one of the many terrible prize/hand combos that would not happen if you were, say, pile shuffling (Which has somewhat of a pattern to it). I have also noticed that my decks consistently inconsistent online, be it PTCGO or PlayTCG, like getting 4 juniper in hand, 4 deoxys prized, playing juniper for supporter hands, or colress for 8 out of the 9 cards left in my deck and not getting the switch.
 
RE: Terrible draws in PTCGO

The random algorithms used in online games are just that -- random. Any good or bad luck is sheerly perception, and falls under the law of large numbers (random systems only behave uniformly for very large numbers of trials). You might find pile shuffling giving you better shuffles because pile shuffling is NOT random. It does not randomize your deck (which, technically, is the purpose of shuffling) and you should never solely rely on a pile shuffle. Even repeated pile shuffling can only serve to re-order your deck.

The result is that, by using solely or primarily pile shuffling to shuffle a deck, one will obtain a deck which has it's cards conveniently well distributed. If the deck was ordered (or even shuffled in the same way) before shuffling, then you are very unlikely to and up with repeated cards in your hand or prizes (so drawing 3 junipers is even less likely that it would be with a statistically randomized deck).

Many players do not realize that failing to actually randomize one's deck actually gives them an unfair advantage (and in harsher cases can be considered cheating). The proper way to randomize one's deck is to use repeated iterations of a non-patterned shuffle, such as a riffle shuffle, an indian shuffle, or a randomly organized pile shuffle, and preferably a combination of a few.

Remember that, if you've actually randomized your deck, you SHOULD get bad hands and prizes from time to time, and will almost never get perfect draws (mainly because there are far more unfavorable patterns than there are ideal ones).
 
Changed the name of the topic to what I'd like the discussion to veer towards. How do you shuffle? I sleeve shuffle, not riffle shuffle (it's virtually the same thing but without the risk of hurting your cards) usually around 7 times ("The Gilbert–Shannon–Reeds model provides a mathematical model of the random outcomes of riffling, that has been shown experimentally to be a good fit to human shuffling and that forms the basis for a recommendation that card decks be riffled seven times in order to thoroughly randomize them."), before pile shuffling in different orders each time (ex. for the first six cards I might just go straight across, but the next six cards I might go vertically up and down). I saw that Jason Klazincsky tends to riffle shuffle only a few times before letting his opponent cut the deck.
 
Btw, I totally agree. I played Blastoise on PTCGO and pretty much every hand I had was a couple junipers and at least 3 SER. Unfair. I both sleeve and riffle shuffle a couple of times each, unless I need to go really fast and in that case I sleeve shuffle like flash and hope time isn't called. I think 7 is a good number though. I may try sticking to that.
 
Assuming by sleeve shuffle you mean weaving a cut of the deck in from the side, it's ideally equivalent to riffle shuffling (and avoids bending the cards), so I would recommend that. 7 is absolutely the number to go for in the beginning of rounds, or after a long search (such as your first search, after checking which cards are prized). After your initial few searches, you should lower that number simply for the sake of game pace, but do not shuffle insufficiently.

The number of sleeve or riffle shuffles you do should be roughly proportional to the information you learn about the deck. In other words, during your first search you will spend a great deal of time simply counting and checking prizes, which means you will, both purposely and incidentally, learn many things about the ordering and contents of the deck. For that reason you really should do 7 shuffles to make sure you haven't gained any unfair advantage (e.g. noticing your top card is a supporter which you need, and thus not shuffling the top portion of the deck).

Later in the game, if you for example play a professor's letter and simply grab the first two energy you see, then you really haven't learned much and, assuming the deck was random (and thus well shuffled) to begin with, you don't need to do too much additional shuffling (but I would not do less than 3 iterations even for minimal searches).

When shuffling cards into your deck, because the cards you insert will be in order, or relatively close together and in order, you must shuffle a number of times roughly proportional to the number of cards shuffled in. So if you play N with a 7 card hand, you really should do 4-7 shuffles to randomize the deck sufficiently (note: randomizing does NOT mean evenly distributing cards. It means putting the deck into a state with equal likelihood as any other possible ordering of the cards).

And as I said before with pile shuffling, it's practically useless unless some random aspect is added somewhere. This can be anything from placing cards on piles in random order (without regard to sequence, pattern, or stack size) to sleeve or riffle shuffling each pile together when finished.
 
I definitely feel that the PTCGO shuffling algorithm is a little weird. I run 3/3 Virizion EX/Genesect EX with 2-2 Raichu and Jirachi EX, and I've drawn opening hands with Pikachu and/or Jirachi way more often than Virizion or Genesect+Ultra Ball :(
 
As far as Pile Shuffling goes, the only time that I even use it is at the beginning of my games when both players are still getting set up. When playing a match, I often Sleeve Shuffle as well as Riffle Shuffle once or twice. Not sure which works best in terms of randomization, but Riffle, Pile, and Sleeve Shuffling have given me both dead hands (for example, a game that I played today) and astoundingly great hands (unike today)
 
I think the thing about pile shuffling after a game are these two things:

1. It's mainly there to show your opponent that you're not cheating.

2. Pile shuffling alone should never happen. Before I even pile shuffle, I sleeve shuffle a good amount of times, then pile shuffle, after which I sleeve shuffle again, before giving my opponent a chance to cut the deck.
 
There is definitely SOMETHING with the TPCGO. I have a particular "One-of" cards that I throw into my decks for late game. And I end up with this card in opening hand about 75% of the time.
 
optimal_max said:
There is definitely SOMETHING with the TPCGO. I have a particular "One-of" cards that I throw into my decks for late game. And I end up with this card in opening hand about 75% of the time.
Click to expand...

Gotta love starting with G-Booster and Juniper in the same hand.
 
I was just talking to the guy who runs the league around here, and he said this basically: "It's fine to de-clump your energies after a game, as long as you shuffle afterward." So then why would you de-clump? What's the point of de-clumping if it is going to be sufficiently (hopefully) randomized after you do so? It really bugs me, I'm gonna talk to him about that today actually when I go to league.
 
Frost Mage said:
I was just talking to the guy who runs the league around here, and he said this basically: "It's fine to de-clump your energies after a game, as long as you shuffle afterward." So then why would you de-clump? What's the point of de-clumping if it is going to be sufficiently (hopefully) randomized after you do so? It really bugs me, I'm gonna talk to him about that today actually when I go to league.
Click to expand...

There is genuinely no point to manipulating your deck in any way whatsoever prior to doing a proper, randomized shuffle, as the result should still be (more or less) completely random either way. Therefore it is a complete waste of time (both yours and your opponent's) to do so. Time spent doing, say, an ordered 5 pile shuffle, would be better spent doing more iterations of a sleeve shuffle or a randomized pile shuffle.

That someone would vehemently object to this would imply that they do not intend to fully randomize their deck prior to play, and thus are stacking their deck. This is cheating.

Obviously, very few players genuinely mean to cheat when doing pile shuffles and things like that. They typically just see other players do the motions and do the same. Often times they do also proceed to do a randomized shuffle afterwards. Besides (as I said before) wasting time, there's nothing inherently wrong with this.

If you feel someone is spending inordinate amounts of time manipulating their deck, simply kindly inform them of the actual point of shuffling, make sure they do a randomized shuffle, and go on with play. There's no point in getting in an argument over this if the context is trivial (e.g. at a league or something).

I should not that there is at least one actual point to pile shuffling, and that is to count the number of cards in your deck. If you don't end up with even piles after doing a 5- (or any other factor of 60) pile shuffle, then you can quickly ascertain that you are missing a card (a serious matter in competitive contexts especially). And solve the situation before gameplay begins.
 
You must log in or register to reply here.
Back
Top