Gem matching unsatisfying.

2

Comments

  • Rhasget
    Rhasget Posts: 412 Mover and Shaker
    Something about the carddraw RNG is dodgy.
    When I tested it a few months back, recording the draws from a normal deck, 4 creatures, 3 spells, 3 supports (no discard, draw, fetch or other tampering). I only tested 14 matches though.
    Drawing 40 cards and getting 4 of each was normal but the pattern on how the cards appeared was a clump pattern.
    If I started with 2 creatures and a spell, the next 20 draws would be 2 to 1 creature cards. 
    If I started with no creatures in hand though there was an average of three more draws before the next creature was drawn until half the deck had been drawn and the majority of cards left were creatures.

    And within the cardtypes it was clumped also.
    For example if I had Lannery Storm, Captivating Crew, Hostage Taker and Stormfleet Swashbuckler I drew two creatures x4 and the third x2 before drawing the forth, in all 14 games.

    If you record 1000 games like this it might show that the RNG indeed is random but within an event with 20-30 matches this behaviour can cause points dropped.

  • hawkyh1
    hawkyh1 Posts: 780 Critical Contributor
    we get more matches per move because the ai actively
    chooses to sacrifice a cascade to take out one of your
    lonely on colour matches. I would almost always take
    the cascade for the loyalty bonus.
    the rng is more extreme. this causes a disproportional
    numbers of games that that are played at the extremes.
    it's a combination of all the rng in the game eg card
    draw, gem matches, treasure bonus card, colourless
    support gem colour. it doesn't necessarily favour the ai.
    ie the number of extra hard games is balanced by the
    number of extra easy games. problem is players are
    grinding through more extra hard games where it feels
    like every time there is a dice roll it would go against
    them to the extreme. with so many possible combinations
    of the dice roll across so many different variables it then
    feels like the rng is cheating in favour of the ai. adding
    to this is the sense that gem board feels inflexible to
    variation. at times there are hardly any option but to take
    the single gem match possible.

    HH
  • babar3355
    babar3355 Posts: 1,128 Chairperson of the Boards
    Kinesia said:
    I thought I was pretty good at matching things (and definitely better than when I started!) but watching Babar's videos it's _very_ clear that better matching on the human side leads to more cascades.

    It's also a question of priorities... Sometimes you need to sacrifice the highest guaranteed mana for a higher chance of a cascade. Greg's priorities include mana-screwing the player so he doesn't match the same as most beginner players to start with, he's not aiming for the same things.
    Thanks for the kind words Kinesia, although sometimes its better to get lucky. 

    I will say that recording yourself and watching a game back, you often notice something that was a massive error.  Missing a match 5, leaving the opponent something ridiculous, etc.  It really helped me improve my game after shooting these videos to show off some really cool deck ideas.

    I also do think that people who focus on the "next level" will tend to have better outcomes.  When facing a choice of several moves, it helps to ask yourself "what happens if..."  Based on this, I almost always take a vertical match vs a horizontal match, all things being equal.  It just makes sense that disrupting gems deep into the gem board will have more potential for 2nd or 3rd level combinations.  Rarely will a top row connection lead to substantial 2nd, 3rd, etc level cascades.  

    Anyway, maybe this is obvious to everyone, or maybe I am completely wrong and believing in something that is  incorrect... but, it seems to be working for me.
  • MTG_Mage
    MTG_Mage Posts: 224 Tile Toppler
    Yes, thanks Volrak for the clear response to the question.

    However when it comes to cards drawn from packs I am very very convinced that it is not true RNG and there is something, either intentionally placed or not, that relies on the last card pulled. As I stated earlier, in a recent super pack I opened, which has 15 cards, there were 3 sets of 3 cards, thats 9 dupes of 3 cards. The odds of 2 of the same exact card in a pack with nearly 2000 cards in the game (1932 according to the card tab) should be extreeeeemly low and rare, like maybe in the 2.5 years since the game came out, a veteran player should only see this happen like a handful of times, not literally every other pack containing two+ of the same card. Even if considering that over 90% of the time each card pulled will be from the common/uncommon pool, those are each 600 card pools, so out of 1200 cards there shouldn't be the exact same card pulled 2 out of 5 times every other pack.

    So I do think that in game RNG is true RNG and peoples views are affected by the negativity bias, however I think that cards in packs are skewed toward getting dupes in order to allow the game to keep players with less new cards as that is advantageous for the devs. Assuming that is true, then there is a bug in how that card distribution RNG is modified and instead of looking at your pool of cards in collection, it merely looks at most recent cards received to ensure that some dupes are handed out but fails at seeing any further back than the current pack it is giving out.
  • DumasAG
    DumasAG Posts: 719 Critical Contributor
    MTG_Mage said:
    Yes, thanks Volrak for the clear response to the question.

    However when it comes to cards drawn from packs I am very very convinced that it is not true RNG and there is something, either intentionally placed or not, that relies on the last card pulled. As I stated earlier, in a recent super pack I opened, which has 15 cards, there were 3 sets of 3 cards, thats 9 dupes of 3 cards. The odds of 2 of the same exact card in a pack with nearly 2000 cards in the game (1932 according to the card tab) should be extreeeeemly low and rare, like maybe in the 2.5 years since the game came out, a veteran player should only see this happen like a handful of times, not literally every other pack containing two+ of the same card. Even if considering that over 90% of the time each card pulled will be from the common/uncommon pool, those are each 600 card pools, so out of 1200 cards there shouldn't be the exact same card pulled 2 out of 5 times every other pack.

    So I do think that in game RNG is true RNG and peoples views are affected by the negativity bias, however I think that cards in packs are skewed toward getting dupes in order to allow the game to keep players with less new cards as that is advantageous for the devs. Assuming that is true, then there is a bug in how that card distribution RNG is modified and instead of looking at your pool of cards in collection, it merely looks at most recent cards received to ensure that some dupes are handed out but fails at seeing any further back than the current pack it is giving out.
    So we've got two conspiracies here, then: Firstly, that the devs are intentionally trying to give us dupes, and second that the coding was done incompetently so that RNG isn't true RNG.
  • MTG_Mage
    MTG_Mage Posts: 224 Tile Toppler
    No. What I stated is that the in game RNG IS true RNG and all these threads claiming the AI is unfair are people who are affected by the negativity bias.
    I wanted to clearly put out the existence of and definition of the negativity bias since so many people jump to the conclusion of AI cheating and complain about it.

    However I am saying that the card pack distribution is in fact rigged against the players and is not random due to the overwhelming amount of times players get the exact same card in a pack.

    It is not in the best interest of the game for the devs to make in game RNG anything other than real as that can cause a bad player experience. However skewing the new content in their favor does effect the bottom line ($) so it is a realistic assumption, and they didn't get it right and seems to be bugged.
  • Unknown
    Unknown
    This content has been removed.
  • Volrak
    Volrak Posts: 732 Critical Contributor
    DumasAG said:
    I am constantly amazed by two things in your posts.
    Thanks for the compliments.  I do try to empathise because when using only text, without taking care it's too easy to write something that inadvertently comes across as an attack or a denigration (maybe it's the forums' own negativity bias!)  In this case I wasn't trying to correct anything (except perhaps hyperbole), but rather add some relevant info that might not be known.  I can't prove there is no RNG bug, although the data I've seen so far can be explained without a bug.

    Some games do deliberately skew the RNG.  For example in Battletech, the supposed odds to hit with a weapon are given, but rolls to hit are skewed for both you and the AI such that the more extreme an event outcome is (whether to hit or to miss), the odds are reported as being even less likely to happen than it actually is (e.g. 95% chance to hit may be reported as 90% chance to hit).  It seems to be an anticipation of negativity bias and an attempt to compensate for it, and I think it works to some degree (though some purists would rather be given the true odds).  Despite that, like mtgpq there have still been dozens of forum posts in that game claiming problems with the RNG, most of which again seem to come down to negativity bias.

    In mtgpq though, it's not clear to me that skewing RNG towards repeated events would be useful - to whatever degree it might compel some paying players to pay more, it would also demotivate them.  And at the same time it would also demotivate the non-paying players without any gain.
  • HarryMason
    HarryMason Posts: 136 Tile Toppler
    I can't say that I think the ai cheats. I used to , but I was a noob and was making excuses for losing. I will say that the ai has uncanny luck . Example : lucky me gets a sweet cascade. Unlucky me leaves Greg a 5er to murder my face with. Greg doesn't take the 5er . Takes random irrelevant match near the hixus corner . Stupid Greg! Oh, wait.... 6 mana...  18 mana... 24 mana.... Oh look, he just cascaded into that 5er he conspicuously skipped and was conspicuously unaffected by the 9 matches he just popped off . I've noticed that things like this don't happen a ton , but it'll happen more than once over the course of the event. Greg is also really good at dropping 3 same colored gems in that vertical slot at the top of the board multiple times in a row. I play koth and count 4 red gems on the board . Greg will drop 3 red matches in the same slot. These situations are frustrating and make me forget about the God cascade that let me beat 5.2 of boft on turn 2. So I know it goes both ways , but Greg's luck is over 9000
  • rafalele
    rafalele Posts: 876 Critical Contributor
    edited June 2018
    Have you ever seen your Hixus support in the bottom right corner? There is where IA's Hixus usually goes and gets a lot of reinforces.

    The most of the times my Hixus support gets matched and disappears the next IA's turn after the one I have casted it.

    Now Seal away is the new one that disappears in the next IA turn altough it has 2 shields (happened twice in the same match in last BOFT). 

    So if this is not cheating what it is?, oh, yes coded behavior. 
  • DumasAG
    DumasAG Posts: 719 Critical Contributor
    rafalele said:
    Have you ever seen your Hixus support in the bottom right corner? There is where IA's Hixus usually goes and gets a lot of reinforces.

    The most of the times my Hixus support gets matched and disappears the next IA's turn after the one I have casted it.

    Now Seal away is the new one that disappears in the next IA turn altough it has 2 shields (happened twice in the same match in last BOFT). 

    So if this is not cheating what it is?, oh, yes coded behavior. 
    Well, it's true that the AI does prioritize destroying your supports (at least, I think it still does). But this pretty much sounds like negativity bias, not cheating. How does coding the AI to cheat with hixus and seal away benefit literally anyone? It also doesn't make sense with the actions of the devs - dumb down the AI but program the game to cheat the player into losing...
  • babar3355
    babar3355 Posts: 1,128 Chairperson of the Boards
    MTG_Mage said:
    Yes, thanks Volrak for the clear response to the question.

    However when it comes to cards drawn from packs I am very very convinced that it is not true RNG and there is something, either intentionally placed or not, that relies on the last card pulled. As I stated earlier, in a recent super pack I opened, which has 15 cards, there were 3 sets of 3 cards, thats 9 dupes of 3 cards. The odds of 2 of the same exact card in a pack with nearly 2000 cards in the game (1932 according to the card tab) should be extreeeeemly low and rare, like maybe in the 2.5 years since the game came out, a veteran player should only see this happen like a handful of times, not literally every other pack containing two+ of the same card. Even if considering that over 90% of the time each card pulled will be from the common/uncommon pool, those are each 600 card pools, so out of 1200 cards there shouldn't be the exact same card pulled 2 out of 5 times every other pack.

    So I do think that in game RNG is true RNG and peoples views are affected by the negativity bias, however I think that cards in packs are skewed toward getting dupes in order to allow the game to keep players with less new cards as that is advantageous for the devs. Assuming that is true, then there is a bug in how that card distribution RNG is modified and instead of looking at your pool of cards in collection, it merely looks at most recent cards received to ensure that some dupes are handed out but fails at seeing any further back than the current pack it is giving out.
    Hmm, well I can't think of a super pack that could contain any of the 1932 cards. Dominaria super pack would only have access to roughly 192 cards... and of those the odds are dramatically skewed towards the 115 U/C cards with a 95%.  I don't know the odds of 3 sets of 3, Vollrak can probably get you there.  But the probability of getting 1 of 115 cards 3 times out of 15 isn't that crazy.  I mean its in the single digits or less, but someone will probably get that at least every day.. You were the lucky winner =)

    Or maybe its rigged.. I dont know.
  • wereotter
    wereotter Posts: 2,070 Chairperson of the Boards
    babar3355 said:
    MTG_Mage said:
    Yes, thanks Volrak for the clear response to the question.

    However when it comes to cards drawn from packs I am very very convinced that it is not true RNG and there is something, either intentionally placed or not, that relies on the last card pulled. As I stated earlier, in a recent super pack I opened, which has 15 cards, there were 3 sets of 3 cards, thats 9 dupes of 3 cards. The odds of 2 of the same exact card in a pack with nearly 2000 cards in the game (1932 according to the card tab) should be extreeeeemly low and rare, like maybe in the 2.5 years since the game came out, a veteran player should only see this happen like a handful of times, not literally every other pack containing two+ of the same card. Even if considering that over 90% of the time each card pulled will be from the common/uncommon pool, those are each 600 card pools, so out of 1200 cards there shouldn't be the exact same card pulled 2 out of 5 times every other pack.

    So I do think that in game RNG is true RNG and peoples views are affected by the negativity bias, however I think that cards in packs are skewed toward getting dupes in order to allow the game to keep players with less new cards as that is advantageous for the devs. Assuming that is true, then there is a bug in how that card distribution RNG is modified and instead of looking at your pool of cards in collection, it merely looks at most recent cards received to ensure that some dupes are handed out but fails at seeing any further back than the current pack it is giving out.
    Hmm, well I can't think of a super pack that could contain any of the 1932 cards. Dominaria super pack would only have access to roughly 192 cards... and of those the odds are dramatically skewed towards the 115 U/C cards with a 95%.  I don't know the odds of 3 sets of 3, Vollrak can probably get you there.  But the probability of getting 1 of 115 cards 3 times out of 15 isn't that crazy.  I mean its in the single digits or less, but someone will probably get that at least every day.. You were the lucky winner =)

    Or maybe its rigged.. I dont know.
    Just opened a pack with three copies of Board the Weatherlight. So even if it isn't rigged, it certainly FEELS rigged.
  • DumasAG
    DumasAG Posts: 719 Critical Contributor
    wereotter said:
    babar3355 said:
    MTG_Mage said:
    Yes, thanks Volrak for the clear response to the question.

    However when it comes to cards drawn from packs I am very very convinced that it is not true RNG and there is something, either intentionally placed or not, that relies on the last card pulled. As I stated earlier, in a recent super pack I opened, which has 15 cards, there were 3 sets of 3 cards, thats 9 dupes of 3 cards. The odds of 2 of the same exact card in a pack with nearly 2000 cards in the game (1932 according to the card tab) should be extreeeeemly low and rare, like maybe in the 2.5 years since the game came out, a veteran player should only see this happen like a handful of times, not literally every other pack containing two+ of the same card. Even if considering that over 90% of the time each card pulled will be from the common/uncommon pool, those are each 600 card pools, so out of 1200 cards there shouldn't be the exact same card pulled 2 out of 5 times every other pack.

    So I do think that in game RNG is true RNG and peoples views are affected by the negativity bias, however I think that cards in packs are skewed toward getting dupes in order to allow the game to keep players with less new cards as that is advantageous for the devs. Assuming that is true, then there is a bug in how that card distribution RNG is modified and instead of looking at your pool of cards in collection, it merely looks at most recent cards received to ensure that some dupes are handed out but fails at seeing any further back than the current pack it is giving out.
    Hmm, well I can't think of a super pack that could contain any of the 1932 cards. Dominaria super pack would only have access to roughly 192 cards... and of those the odds are dramatically skewed towards the 115 U/C cards with a 95%.  I don't know the odds of 3 sets of 3, Vollrak can probably get you there.  But the probability of getting 1 of 115 cards 3 times out of 15 isn't that crazy.  I mean its in the single digits or less, but someone will probably get that at least every day.. You were the lucky winner =)

    Or maybe its rigged.. I dont know.
    Just opened a pack with three copies of Board the Weatherlight. So even if it isn't rigged, it certainly FEELS rigged.
    "Feels" being the key word there.
  • babar3355
    babar3355 Posts: 1,128 Chairperson of the Boards
    wereotter said:
    Just opened a pack with three copies of Board the Weatherlight. So even if it isn't rigged, it certainly FEELS rigged.
    Don't get me wrong, I think they should change the coding in 2 ways.

    1. Prevent a single opening from resulting in multiples of the same card.
    2. Implement a pity timer so that those people on a bad luck streak will have enhanced drop rates until they land a mythic+.

    I am just far from convinced that it is rigged.
  • Kinesia
    Kinesia Posts: 1,621 Chairperson of the Boards
    babar3355 said:
    wereotter said:
    Just opened a pack with three copies of Board the Weatherlight. So even if it isn't rigged, it certainly FEELS rigged.
    Don't get me wrong, I think they should change the coding in 2 ways.

    1. Prevent a single opening from resulting in multiples of the same card.
    2. Implement a pity timer so that those people on a bad luck streak will have enhanced drop rates until they land a mythic+.

    I am just far from convinced that it is rigged.

    I agree. I don't think it's rigged but there's a real reason for pity timers.
    Things need to be deliberately rigged _against_ bad feelings, noone wants to be the customer at the bad end of the bell curve.
  • HarryMason
    HarryMason Posts: 136 Tile Toppler
    That's sort of the sad part about rng based games. Random is random. I like to use the game diablo as an example. There's a specific rare item that you can get in the beginning of the game if rngesus loves you and/or you reload your save a few dozen times until it drops. From there , if rngesus loves you more, you can finish the game in about 13 minutes . Meaning overall, the game is beatable in about 17 minutes. That isn't a realistic expectation though. Why ? Rng doesn't often land on ideal circumstances. For every 40 mana turn I've seen Greg get, I've also seen him take a loyalty match over his own activation gem . We just don't notice when Greg is a nonce because we're too busy squishing him with our toys. The 'ai cheats' is a battle cry as old as vidya itself. Remember when double dragon cheated ? 
  • DumasAG
    DumasAG Posts: 719 Critical Contributor
    HarryMason said:
    That's sort of the sad part about rng based games. Random is random. I like to use the game diablo as an example. There's a specific rare item that you can get in the beginning of the game if rngesus loves you and/or you reload your save a few dozen times until it drops. From there , if rngesus loves you more, you can finish the game in about 13 minutes . Meaning overall, the game is beatable in about 17 minutes. That isn't a realistic expectation though. Why ? Rng doesn't often land on ideal circumstances. For every 40 mana turn I've seen Greg get, I've also seen him take a loyalty match over his own activation gem . We just don't notice when Greg is a nonce because we're too busy squishing him with our toys. The 'ai cheats' is a battle cry as old as vidya itself. Remember when double dragon cheated ? 
    Double Dragon DID cheat... Pretty sure Ninja Turtles on the NES did as well...
  • wereotter
    wereotter Posts: 2,070 Chairperson of the Boards
    babar3355 said:
    wereotter said:
    Just opened a pack with three copies of Board the Weatherlight. So even if it isn't rigged, it certainly FEELS rigged.
    Don't get me wrong, I think they should change the coding in 2 ways.

    1. Prevent a single opening from resulting in multiples of the same card.
    2. Implement a pity timer so that those people on a bad luck streak will have enhanced drop rates until they land a mythic+.

    I am just far from convinced that it is rigged.
    I agree. Duplicates within a pack shouldn't exist at all. Paper magic managed to figure out that was a bad thing and it only ever happens if one of the cards is a foil version of another card in the pack, in which case, it's technically a different card because of the finish.

    Triplicates of cards in a single pack are impossible in paper magic, and definitely should be here as well.

Categories