Jump to content
CeladonMD

[GAME] General - There may be an issue with the shuffling system not being random

Recommended Posts

CeladonMD

Disclaimer: This is not a whining session. These comments are based on almost 3 years of observations and I am curious to see what the community thinks.

I have seen several times, that if I go on a win streak with a particular deck, certain patterns of shuffling emerge that provide evidence that a non-random variable(s) in shuffling is present. In other words, the shuffling creates a disadvantage that decreases your ability to perform well.

For example:
In a deck comprised of 3 basic Water Pokemon, and 9 basic Psychic, with all Psychic energies, it would be expected that 1/4 of first hand shuffles that result in only one basic Pokemon being drawn, would have only one water Pokemon. This seems to happen, in general, and I recorded how many times a Keldeo or Seismitoad popped up as the only basic Pokemon to start with: 28% over three days (pretty close to 1/4 so not bad). However, if I won 3-5 versus games in a row, or worse, 2 tournaments in a row, the percentage of Keldeo or Seismitoad jumped to 58% and wouldn't go down until I signed out (and stayed off for at least an hour) or lost 4-6 versus games or 2 consecutive tournaments (usually in the first round).

In addition, the drawing of water Pokemon was also matched with an absence of Supporter cards. The same deck mentioned above ran 16 support cards. In the 3-6 turns it would take to be defeated (given no real attack options aside from Seismitoad for 30 damage), not ONE Supporter card would arrive until at least turn 3. This means that out of 8 cards, not one would be a Supporter, further worsening the ability to adapt/attack/draw. If you consider that the chance of drawing a Supporter on the first turn was 7*(16/60)=186% and by turn 2, 213%, surely at least ONE Supporter should have been present, in fact, 2 should have been present. But, they were not. During periods when I was drawing Water basic Pokemon at higher rates, my Supporter draw rate on first hand was a measly 12%!

Obviously, being dealt your support/situational Pokemon at the wrong time is something that will happen to all of us. But, I have seen evidence of a clear jump in the likelihood that it will happen if a player goes on a win streak. This combined with poor Supporter card drawing is a crippling to any player, in almost any format.

I would like to hear from the community on their experience regarding non-random or biased shuffling experiences, especially with high winning streaks or higher ranking.

If anyone would like to see a copy of my numbers, please send me a message. If you have any data on this, please feel free to discuss it here as well.

The purpose of this thread is to expose any biased shuffling practices that are imposed. Though there is no assurance that what I observed is real or intended, I hope we can put the issue to rest by open discussion and study. After all, who wants to play a game where your own cards are stacked against you?

Edited by CeladonMD
  • Upvote 2

Share this post


Link to post
Share on other sites
awesome_guy

I personally have had a doubt on this once or twice..

 

one tournament I had 3 jirachi starts in a landobat deck with my basic cards being:

1 landorus-ex, 1 lucario ex, 3 hawlucha baby, 2 lando baby and 3 zubat (obviously + jirachi)

 

so now considering that I had 11 pokemon, odds of me starting jirachi should be 1/11 if we ignore the fact that I can start with multiple at once (which would make the odds even rarer than this)

 

odds of starting 2 pokemon would be 11/360 and removing those odds, our chances of starting jirachi would be 22/360 (around 1 in 18 times- i'm rounding to 18 and not 17 since we didn't count chances of starting 3 poke and so on)

 

considering that this happens 3 times straight, you get 1/5,832- the equivalent of flipping >12 tails straight (for those who like to calculate it that way)

 

seems a bit out of reach even for my luck?

 

this all may just be coincidence.. but as more people narrate their experience maybe we'll have more luck finding out

  • Upvote 4

Share this post


Link to post
Share on other sites
Aniketdianafury

Well this card shuffling system is suffering from a very savage disease , it is going against against many players now . In one game where my opponent was using raymin deck got shaymin on turn 1 and his hand was filled with energies, spirit links and seekers . But unfortunately i had a good hand and got night marchers so i knocked him out on turn 1 . And some times it is making players lose in 2-3 games in a row specially in tournament . This issue needs to be taken seriously and i completely support you @CeladonMD .

Edited by Aniketdianafury
  • Downvote 1

Share this post


Link to post
Share on other sites
harshu

I personally have had a doubt on this once or twice..

 

one tournament I had 3 jirachi starts in a landobat deck with my basic cards being:

1 landorus-ex, 1 lucario ex, 3 hawlucha baby, 2 lando baby and 3 zubat (obviously + jirachi)

 

so now considering that I had 11 pokemon, odds of me starting jirachi should be 1/11 if we ignore the fact that I can start with multiple at once (which would make the odds even rarer than this)

 

odds of starting 2 pokemon would be 11/360 and removing those odds, our chances of starting jirachi would be 22/360 (around 1 in 18 times- i'm rounding to 18 and not 17 since we didn't count chances of starting 3 poke and so on)

 

considering that this happens 3 times straight, you get 1/5,832- the equivalent of flipping >12 tails straight (for those who like to calculate it that way)

 

seems a bit out of reach even for my luck?

 

this all may just be coincidence.. but as more people narrate their experience maybe we'll have more luck finding out

Well if you wanna see another experience, count mine. I got all tails, I was playing Malamar EX. No other pokemon was avialable and all I had was Energies. This continued for 3 turns. After second turn I Started attacking, 2 tails out of 2 flips, 3rd turn, attached DCE, 4 tails out of 4. 5th turn, got N still no pokemon. 5 Flips, 5 Tails and I lose...

 

The abive examples completely show how wrecked and broken is the system of Flips and Draws. You may also see other bugs of Drawing on this topic by azimuthal_ who has posted about cards being not shuffled which can be a cause of this Drawing system, which has relatively grown bad !

 

Just thought if this is related in any way and we can get help in :)

 

http://forums.pokemontcg.com/topic/31911-game-general-cards-do-not-shuffle-after-the-use-of-an-item-card-that-states-it/

  • Upvote 1
  • Downvote 1

Share this post


Link to post
Share on other sites
sterlingW

Disclaimer: This is not a whining session. These comments are based on almost 3 years of observations and I am curious to see what the community thinks.

 

I have seen several times, that if I go on a win streak with a particular deck, certain patterns of shuffling emerge that provide evidence that a non-random variable(s) in shuffling is present. In other words, the shuffling creates a disadvantage that decreases your ability to perform well.

 

For example:

In a deck comprised of 3 basic Water Pokemon, and 9 basic Psychic, with all Psychic energies, it would be expected that 1/4 of first hand shuffles that result in only one basic Pokemon being drawn, would have only one water Pokemon. This seems to happen, in general, and I recorded how many times a Keldeo or Seismitoad popped up as the only basic Pokemon to start with: 28% over three days (pretty close to 1/4 so not bad). However, if I won 3-5 versus games in a row, or worse, 2 tournaments in a row, the percentage of Keldeo or Seismitoad jumped to 58% and wouldn't go down until I signed out (and stayed off for at least an hour) or lost 4-6 versus games or 2 consecutive tournaments (usually in the first round).

 

In addition, the drawing of water Pokemon was also matched with an absence of Supporter cards. The same deck mentioned above ran 16 support cards. In the 3-6 turns it would take to be defeated (given no real attack options aside from Seismitoad for 30 damage), not ONE Supporter card would arrive until at least turn 3. This means that out of 8 cards, not one would be a Supporter, further worsening the ability to adapt/attack/draw. If you consider that the chance of drawing a Supporter on the first turn was 7*(16/60)=186% and by turn 2, 213%, surely at least ONE Supporter should have been present, in fact, 2 should have been present. But, they were not. During periods when I was drawing Water basic Pokemon at higher rates, my Supporter draw rate on first hand was a measly 12%!

 

Obviously, being dealt your support/situational Pokemon at the wrong time is something that will happen to all of us. But, I have seen evidence of a clear jump in the likelihood that it will happen if a player goes on a win streak. This combined with poor Supporter card drawing is a crippling to any player, in almost any format.

 

I would like to hear from the community on their experience regarding non-random or biased shuffling experiences, especially with high winning streaks or higher ranking.

 

If anyone would like to see a copy of my numbers, please send me a message. If you have any data on this, please feel free to discuss it here as well.

 

The purpose of this thread is to expose any biased shuffling practices that are imposed. Though there is no assurance that what I observed is real or intended, I hope we can put the issue to rest by open discussion and study. After all, who wants to play a game where your own cards are stacked against you?

Follows the real life one

Share this post


Link to post
Share on other sites
overcast_kid

I agree there is a formula for shuffling. ALSO the coin flips are rigged against players in the advantage. I've made multiple posts about this and without hesitation the posts get locked down. They always say its truly random but we know different. I support this post.

  • Upvote 1

Share this post


Link to post
Share on other sites
Aniketdianafury

Follows the real life one

 

Oh it does not follow's the real life one anymore for me  ;)  . 

 

Edited by Aniketdianafury

Share this post


Link to post
Share on other sites
TimeWhisp

Well with any randomizer there's going to be issues. I have noticed that decks seem to have their own randomization streaks too. There was a week where my Blastoise deck threw up every game. I drew unplayable hand after unplayable hand. Then magically one day I started drawing good again. But honestly I was getting an unplayable hand probably 50% of games and bad hands 25% of the time. Bad hands are part of any game with chance, but I feel like there are definitely streaks to it in this game.

 

The worst randomization for me is "flip a coin until you get tails." I can't tell you how many laughable situations this has caused for me. I once hit tails 7 times in a row in the same game with the same pokemon. I can count the times I've gotten more than 2 heads in a row and I play that deck a lot. Yet it seems when my opponent plays cards with this mechanic they always get at least 2 heads.

 

And then there's the counter to my theory. I play the Roserade that allows you to flip four coins and does 30 damage for each heads in one of my other decks. For some reason I'm unstoppable with this pokemon. I get 2 heads almost guaranteed and 3-4 heads frequently. Riddle me that.

Share this post


Link to post
Share on other sites
JamesPup

I also keep track of the odds I have in getting cards like when drawing at the beginning and using Corless. I would have to disagree though because I see that the odds seem to follow the way they should play out. I mean sometimes I get stuck with no energy when I should get at least two with ten cards drawn and things like that but that is expected sometimes. If there is something that is making certain cards being drawn more than others due to any winning streak or something like that than that needs to be fixed or at least it should be revealed to us players as it is important information. I still don't think that there is such a problem though.

  • Upvote 1

Share this post


Link to post
Share on other sites
awesome_guy

james.. how lucky do you have to be to get 10+ heads in a row? I've had 15 happen against me without a tails.. it could be a one man thing.. or it could be happening to multiple people..

 

if we say a ten thousand people play this game (fair estimate right?).. and each plays 100 battles- yes some play more but a lot play less, we have 1 million battles.. and 2^15 here (or 2^13+ up in my previous post) is in excess of 30,000- do I really have the kind of luck to get two different events where I have the luck of being one of 33 people in the world to have that happen to them?

 

 

this raises doubts to me about if the randomizer is truly random- which is why we're seeing if there are others who face this

Edited by awesome_guy

Share this post


Link to post
Share on other sites
arklumpus
CeladonMD, on 03 Jul 2015 - 3:38 PM, said: Disclaimer: This is not a whining session. These comments are based on almost 3 years of observations and I am curious to see what the community thinks. I have seen several times, that if I go on a win streak with a particular deck, certain patterns of shuffling emerge that provide evidence that a non-random variable(s) in shuffling is present. In other words, the shuffling creates a disadvantage that decreases your ability to perform well. For example: In a deck comprised of 3 basic Water Pokemon, and 9 basic Psychic, with all Psychic energies, it would be expected that 1/4 of first hand shuffles that result in only one basic Pokemon being drawn, would have only one water Pokemon. This seems to happen, in general, and I recorded how many times a Keldeo or Seismitoad popped up as the only basic Pokemon to start with: 28% over three days (pretty close to 1/4 so not bad). However, if I won 3-5 versus games in a row, or worse, 2 tournaments in a row, the percentage of Keldeo or Seismitoad jumped to 58% and wouldn't go down until I signed out (and stayed off for at least an hour) or lost 4-6 versus games or 2 consecutive tournaments (usually in the first round). In addition, the drawing of water Pokemon was also matched with an absence of Supporter cards. The same deck mentioned above ran 16 support cards. In the 3-6 turns it would take to be defeated (given no real attack options aside from Seismitoad for 30 damage), not ONE Supporter card would arrive until at least turn 3. This means that out of 8 cards, not one would be a Supporter, further worsening the ability to adapt/attack/draw. If you consider that the chance of drawing a Supporter on the first turn was 7*(16/60)=186% and by turn 2, 213%, surely at least ONE Supporter should have been present, in fact, 2 should have been present. But, they were not. During periods when I was drawing Water basic Pokemon at higher rates, my Supporter draw rate on first hand was a measly 12%! Obviously, being dealt your support/situational Pokemon at the wrong time is something that will happen to all of us. But, I have seen evidence of a clear jump in the likelihood that it will happen if a player goes on a win streak. This combined with poor Supporter card drawing is a crippling to any player, in almost any format. I would like to hear from the community on their experience regarding non-random or biased shuffling experiences, especially with high winning streaks or higher ranking. If anyone would like to see a copy of my numbers, please send me a message. If you have any data on this, please feel free to discuss it here as well. The purpose of this thread is to expose any biased shuffling practices that are imposed. Though there is no assurance that what I observed is real or intended, I hope we can put the issue to rest by open discussion and study. After all, who wants to play a game where your own cards are stacked against you?
I think we ought to be more "scientific"... I.e., instead of relying on the "impression of bias", we can use statistical instruments to evaluate [why is ******** censored? o.O] the probability that the differences between the observed occurrences of the event and the expected (=random) ones are due to fluctuations. Percentages are not that useful to do this, thus, if you have the raw data (number of occurrences) and could post it, that would help a lot. If you have not done so already, it might also be a good idea to record all of the hands and do the skimming at the end. As a side note, I don't think that this bias (if a bias it really is) is intentional: it seems a nightmarish thing to implement (coin flips are easy, but an algorithm that gives a "bad" hand should not be trivial: for a start, how can a "bad" hand be defined?)... If real, it probably reflects some property of the rng, and the fact that only people who have been disadvantaged by this are reporting it is not strange, as the others likely just felt "lucky" and went on playing. Edited by arklumpus

Share this post


Link to post
Share on other sites
Stratadrake

If you consider that the chance of drawing a Supporter on the first turn was 7*(16/60)=186% and by turn 2, 213%, surely at least ONE Supporter should have been present, in fact, 2 should have been present. But, they were not. During periods when I was drawing Water basic Pokemon at higher rates, my Supporter draw rate on first hand was a measly 12%!

Actually, the probability of having no Supporters at all in your opening hand is the chance of drawing seven non-Supporter cards in a row from a full (randomly shuffled) deck:  (44/60) * (43/59) * (42/58) * ... * (38 / 54) = ~9.9%. Thus, you should be able to expect at least one Supporter in your opening hand 9 of 10 times, and if you're not, you do have a case to complain that something isn't right.

 

Of course, that's assuming a truly random shuffle, because a physical shuffle isn't actually a random process: a shuffle splits the deck between its rough top and bottom halves, and the cards are still in the same order relative to the others from their half. Which is, of course, why you're supposed to shuffle a deck repeatedly.

Edited by Stratadrake

Share this post


Link to post
Share on other sites
Cwarfare718

I think we ought to be more "scientific"... I.e., instead of relying on the "impression of bias", we can use statistical instruments to evaluate [why is ******** censored? o.O] the probability that the differences between the observed occurrences of the event and the expected (=random) ones are due to fluctuations. Percentages are not that useful to do this, thus, if you have the raw data (number of occurrences) and could post it, that would help a lot. If you have not done so already, it might also be a good idea to record all of the hands and do the skimming at the end. As a side note, I don't think that this bias (if a bias it really is) is intentional: it seems a nightmarish thing to implement (coin flips are easy, but an algorithm that gives a "bad" hand should not be trivial: for a start, how can a "bad" hand be defined?)... If real, it probably reflects some property of the rng, and the fact that only people who have been disadvantaged by this are reporting it is not strange, as the others likely just felt "lucky" and went on playing.

^^This^^

Beat me to it :)

Share this post


Link to post
Share on other sites
awesome_guy

^^This^^

Beat me to it :)

if you check cwarfare I run 15 supporter cards in my deck (draw supporter/ cards that can instantly get one)

 

if I can get any less than 8.5/10 normal hands, I kinda have a right to be unsatisfied- especially when a deck has a bad period that stretches for a month or more.. (don't even let me get started on topdecking- that seems to be like 40% supporter no matter what)

 

note: I play across practically every time zone- so skill doesn't exactly vary much- yet my deck's lean periods seem to stretch across weeks and its good periods seem to go quite long too.. just test 100 battles with any deck and i'm telling you you won't land within a percent of what you are supposed to get- meaning someone who is running a total of 7 supporter cards can land the same hands as me when I run 15

 

Either way- this is a theoretical concept where we are trying to gather information.. just repeating that you haven't seen a problem with it is okay, questioning our methods is just counterproductive

Edited by awesome_guy
  • Upvote 1

Share this post


Link to post
Share on other sites
Cwarfare718

@Awesome Guy, sorry I forgot to quote and don't know how to add one once I've posted.

 

I believe there are 2 potential issues here... The most likely of which is simply selective memory, and the other less plausible one which is RNG shuffling issues. However, if the latter happens to be the problem it is extremely unlikely that the RNG is built to sabotage successful decks, and more likely an unrelated technical flaw that appears to have some sort of "bias agenda". For the record, I have experienced terrible drawing streaks that initially appear to be intentionally implemented, but I understand it is a matter of probability and dismiss my emotional conclusion.

 

Selective Memory: an ability to remember some facts while apparently forgetting others, especially when they are inconvenient

How many successful decks have consistently drawn well, while not being recorded or recognized while doing such? It is convenient to point out negative outcomes and display the results, but essentially useless to post positive/normal experiences, therefore inaccurate data is presented. Although your personal experiences have brought you to these conclusions, it would take 100s if not 1000s of recorded games from a plethora of players to even debate this issue. (in my opinion)

 

As arklumpus addressed earlier, what is a bad hand? How can an RNG decide what is or is not beneficial for you to draw into? If anything there is an RNG shuffling issue that has no bias, but appears that way. I don't know....perhaps after using a deck for a certain period of time cards with lower counts in the deck show up in opening hands more often...or anything equally irrelevant to sabotage... But again...extremely unlikely.

Edited by Cwarfare718

Share this post


Link to post
Share on other sites
VancouverDad

I am horrible when it comes to statistics but I am very aware of the multitude of biases associated with things like this and try to be as non-biased as possible.

 

Note: The observations below are exactly that and I haven't kept stats to back them up.

 

I have found certain cards, such as pokemon catcher, to have an excessively low success rate, far beyond what I would expect.

 

I play Prof, Oak in many of my decks.  In a higher frequency than I would expect, I often get redealt a couple of key cards that I held in my hand before the shuffle.  I have found this to happen with N, also.

 

Few randomizers truly do it well.  I do wonder sometimes if there is a bit of an issue with flips and shuffles.

  • Upvote 1

Share this post


Link to post
Share on other sites
arklumpus

@awesome_guy: I was not really questioning your method, simply stating that, without clear data, we don't have much of a case... I think that helping to improve the game is something we all look forward to, but it must be done properly: going to someone saying "I always get tails" isn't going to get us anywhere, while "I have recorded 500 consecutive coin flips, and found that 350 of them was tails, which has a probability to happen that's lower than 10^-19: can you please investigate this?" just might.

 

Thus I was asking you for your data, so we can at least run a "quick and dirty" chi-squared test on it... Unfortunately, I have neither the time nor the cards to thoroughly test this at least for another week.

  • Upvote 1

Share this post


Link to post
Share on other sites
overcast_kid

I am horrible when it comes to statistics but I am very aware of the multitude of biases associated with things like this and try to be as non-biased as possible.

 

Note: The observations below are exactly that and I haven't kept stats to back them up.

 

I have found certain cards, such as pokemon catcher, to have an excessively low success rate, far beyond what I would expect.

 

I play Prof, Oak in many of my decks.  In a higher frequency than I would expect, I often get redealt a couple of key cards that I held in my hand before the shuffle.  I have found this to happen with N, also.

 

Few randomizers truly do it well.  I do wonder sometimes if there is a bit of an issue with flips and shuffles.

^ this 

you dont want to know my crazy conspiracy theory but ill tell you anyway. Im pretty sure that there is rubberbanding in pokemon tcgo. Rubberbanding is when the game purposely hurts you or helps your opponent to make the game seem more even when in reality you should be stomping them. For example what is the highest damage anyone has gotten with attacks required by flips?. I have a hunch that it wont go beyond a certain amount of damage.  Pokemon catcher works like never. Not since they were errata. I think pokemon tcgo alters flip success on certain cards to steer us away from old cards to promote new ones .I think the game pulls cards out for example dependent on how many cards we have in hand vs our opponent and that will determine a supporters drawn. I get  us needing to be anti bias, Im just pontificating out loud to see where we all stand. 

Share this post


Link to post
Share on other sites
awesome_guy

let's actually do that.. 500 flips with each of 1. hammer 2. scoop 3. catcher 4. laser

 

if any of them land more than 30-40 off the mean we can probably conclude a biased flip rate

 

I can try this out over the weekend or such but would request anyone to help for any one of these.. another data I will record is longest streak of heads/tails

  • Upvote 1

Share this post


Link to post
Share on other sites
Patriarch

A user on my board is saying the same thing, Celadon. I find the theory interesting, though having played Pokemon TCG on a variety of clients over the years, it's probably just the RNG that's acting up. Either that, or as another one of my users posited, players IRL are more used to even distribution than they are to "pure random."

 

However, your idea of a rubberbanding code is concerning. While I won't assume the worst of PTCGO's staff unless proven otherwise, It would still be pretty reassuring if a dev team member came in and said on the record "No no no, we do not have a rubberbanding algorithm and never will -- quit being ridiculous."

Share this post


Link to post
Share on other sites
overcast_kid

 

A user on my board is saying the same thing, Celadon. I find the theory interesting, though having played Pokemon TCG on a variety of clients over the years, it's probably just the RNG that's acting up. Either that, or as another one of my users posited, players IRL are more used to even distribution than they are to "pure random."

 

 

 

However, your idea of a rubberbanding code is concerning. While I won't assume the worst of PTCGO's staff unless proven otherwise, It would still be pretty reassuring if a dev team member came in and said on the record "No no no, we do not have a rubberbanding algorithm and never will -- quit being ridiculous."

careful about getting someone from staff to comment on this because ive brought this up before and they shut my post down and locked it real quick. To me it seemed like i was being swept under a rug. Why listen to me though im a skeptic by nature. To me though rubberbanding algorithms wouldnt be to hard to implement to a game like this to promote superficial difficulty. I mean they can do it in trainer challenges easy medium and hard modes so why not PvP?

Edited by overcast_kid
  • Upvote 1

Share this post


Link to post
Share on other sites
tarif15

i too face this problem . i mean .... i get jirachi ex / trubbish 80 % of the time when i play with my yveltal deck :/ which has 3 yveltal ex , 2 yveltal baby , 2 toad , 1 mewtwo and 1 darkrai :/ thats 3 against 9 and the "3" wins the hand 80 % of the  time . really irritating

Share this post


Link to post
Share on other sites
Patriarch

careful about getting someone from staff to comment on this...

Hah, I'm not gonna be careful about jack -- if I get censored for bringing up serious topics and serious concerns, then so be it. I've been a player for years, have nothing but positive contributions and constructive criticism, and generally do a good job of giving the developers the benefit of the doubt. I enjoy this game very much, and if I can't speak my mind on a whole forum with the explicit goal of user feedback, then something is seriously broken.

  • Upvote 1

Share this post


Link to post
Share on other sites
overcast_kid

I share your enthusiasm, im merely saying be aware that this topic has been brought up and shut down before. I mean if we trade buy and play this game and secretly is rigged then that worries me. We should be able to talk about it openly.

Edited by overcast_kid

Share this post


Link to post
Share on other sites
Patriarch

I'd say an untouched thread consisting of 20+ posts counts as "open discussion."

At any rate, I'd doubt they'd ever do it. Considering how skilled at programming some members of the playing community are, it would be caught pretty quickly.

Share this post


Link to post
Share on other sites
Guest
This topic is now closed to further replies.

×
×
  • Create New...