If you think matchmaking is rigged, read this

Multiplayer Discussion
21

To an extent, yes, clusters of decks do appear because by win rate they will have a tendency to clump up. That’s usually not what the rigged crowd is talking about.

They usually complain more that they’ll be running into a wall of counter decks, change decks, and immediately a new wall of different counter decks is suddenly waiting for them because the game wants them to lose more.

You can absolutely see some of the clustering you describe naturally occurring as you notice that deck distributions shift at each rank bracket. Some decks are super common in gold, but vanish by top legend. There are multiple reasons for this, but it is partially the effect you observed in your original post.

The matchmaking system is best described as pseudo-random.

The deck distributions per rank and what people are choosing to play that day are not particularly random.

What players are available the moment you hit play to get the fastest is somewhat random, so you end up over time having a matchmaking spread similar to the distribution of what’s being played at your rank bracket.

The game certainly isn’t looking to counter queue you, as you said, just the basic nature of a ranking system will lead to you being more likely to lose the more you win for various reasons (different decks, higher skilled players, etc.).

But you still expect that if you just keep queueing in the same rank range, that you will have your match spread look very much like what you observe in meta reports, as your matches are functionally close enough to a randomized distribution.

Yes, there will be streaks in there, but given enough games played, the argument that the system is seeking out control decks to counter your aggro deck or vice versa falls apart very quickly.

22

The data I provide actually predicts this in certain situations. For example, let’s say you’re playing Rock and you’re being countered by Paper. You have just lost a game, so unbeknownst to you you’ve just lowered your rating enough that your next opponent will be Rock. But because your last opponent was Paper, you switch to Scissors.

In hockey there’s a phrase: don’t go to where the puck is, but instead to where the puck is going to be. The complaint you’re describing comes from people not understanding that matchmaking IS by winrate, and that when they lose their previous opponent is where the puck was, not where it’s going to be next game. What these players should do instead is continue to play one more game with their current deck then switch decks after a win.

23

What you didn’t account for, which is the X factor in this, is the availability of players at that MMR. The longer the wait, the less likely you are to get MMR matched. While I do not think match making is rigged, their MMR structure is only viable in a time vacuum. So what might work in legend might only run into hard counters in diamond. I know the play style and deck variance differs depending on what rank you are.

24

I cant claim i understand everything you wrote here completely but my intuition tells me this does make a lot of sense.

This is i think what is happening with hearthstone rng.

25

The canned pseudoRNG algorithm in Unity will be sufficient to produce results indistinguishable from true random for every game of Hearthstone ever played from now till the end of the game’s lifespan.

Your pseudointellectual argument about pseudorandomness is totally irrelevant to the discussion.

3 Likes
26

The biggest argument against rigging is that there literally is no reason to rig a game when you have an MMR system. MMR has proven to be one of the best ways of matchmaking to keep people as close to 50% as possible by matching them with players of equal skill.

Any attempt to rig a game based on deck type is a waste of money and time that is better spent coming out with new content.

The problem is that people don’t want to admit their skill level is as low as it is and people think they are better than their opponents or better than they actually are. So instead of coming to terms with their skill level, they blame the system.

That’s absolutely what is going on and people don’t want to face the reality of the situation.

For crying out loud, they can barely balance the game, and these people think they have the coding intellect to rig games based on deck counters that require knowledge of what is perfectly balanced in order to rig a game properly and that they redo this code every time a new card is introduced and they monitor literally 10,000+ different decks over the history of Hearthstone and how each deck performs over the other 9,999 decks? Gimme a break.

5 Likes
27

Not to mention it tends to defeat the number one purpose of the matchmaking system:

Getting you into a reasonable game quickly.

Every added constraint you add just slows it down.

Otherwise the game would have to be like “hold up, you are playing a combo deck, going to wait for someone to queue with an aggro deck built to exploit your specific weakness.”

(And that assumes that their algorithm can correctly reduce your gameplan in the first place)

4 Likes
28

Psychologically speaking, rigging the game to promote negative feelings is not a waste of money. They know what they’re doing.

1 Like
29

I mean, it’s a terrible strategy for any product you want to make you money.

1 Like
30

How is that not more harmful to a company than positive?

To think that a company is purposefully trying to make your experience miserable is highly counterintuitive to what every game designer is trying to achieve - make an enjoyable game.

This only makes sense if you have data that suggests the more a person hates your game, the more money they spend on it…

1 Like
31

Please explain why promoting negative feelings is a good idea for an entertainment product.

4 Likes
32

This movie sucks! Guess I have to buy 5…

1 Like
33

Haha I see the three of you in general replying to me, so I’ll just post this instead of trying to tag you all in it. This is my response to your questions though.

You ask why is promoting negative feelings a good idea for a product. It’s not promoting a bad experience, it’s just delivering one.

If we use Hearthstone as an example to what I mean, look like this. Just like people who have a good time winning, seeing a “victory” screen, the noise it’s with, is a lot like a casino. Makes you feel good.

You cannot tell me, they didn’t sink resources into the product to deliver the same punch when you take a loss. When a person is frustrated, they go off on a tangent of why.

The deck my opponent played is broken, or if they didn’t have X opening hand I would’ve won. The thing in common is the person deflects blame onto something else than themselves. “Most of the time”.

Now when you are successful at delivering a good reaction to winning, vs losing people want to chase winning again.

This in turn, would more than likely have the effect of people not leaving the game, but maybe instead spending money on the product. As perhaps a majority of humans will think if they had X deck or card combo what have you, they can go back to winning.

I would wager you guys are aware that game companies actually seek out individuals in the psychologically field to help with designs. An the awful truth is because of the example I listed.

I don’t want to use this as proof they rig the game lol. More of a statement that yes, giving negative feelings is indeed beneficial to games. Especially those like Hearthstone.

34

So in other words

plus “it isn’t my fault it sucks.”

Overall, I think the percentage of the general population with flow blown narcissism is exaggerated by social media and not actually large enough to make game design decisions around. And what you’re describing isn’t the way a sane person would think about Hearthstone, but the way a narcissist would think about it.

1 Like
35

Also, the most popular sports teams are obviously the ones that lose in boring games where they accomplish nothing of consequence.

36

Mand, you need to understand that a sports team getting beat up gives them a seething Sith Zenkai boost that makes them certain to win in the future. :laughing:

1 Like
37

There are reported issues with Unity’s random number generator [1][2], and besides that would be client based RNG, so very exploitable to hacking/cracking. Very unlikely that the RNG rolls are happening in unity, especially client-side.

The random rolls are almost certainly being performed server-side, and server-side code is unlikely to be running a Unity implementation, the backend is probably (I would hope) efficiently coded server that runs/manages headless hearthstone games for 1000’s to millions of users.

And FWIW, puns are the lowest form of humor :slight_smile: Not my quote BTW.

[1] forum.unity.com/threads/random-range-doesnt-look-random.954246/#:~:text=It%20happens%20always%2C%20every%20time,colors%20are%20not%20random%20usually.
[2] answers.unity.com/questions/985988/randomrange-not-random-enough.html

1 Like
38

I want everyone to know that regardless of whether I agree with a post or not, I am mandated by law to like any post containing multiple footnotes.

39

Also irrelevant to the discussion, because the likelihood that someone is hacking Blizzard’s RNG to make the entire system cheat people out of wins is even lower than that of Blizzard deliberately doing so.

The point was that pseudoRNG is a solved problem in computer science, and more than adequate for a low-call application like a card game, even a popular one.

Also, it wasn’t a pun, it was a description of your actual process posting here.

40

If RNG rolls happened clientside and were sent or communicated with the server that then relays them to the other players client (the only feasible way to implement what you describe as using Unity’s canned pseudo RNG algorithm) this would be a monumental security issue, since RNG rolls greatly determine the outcome of games.

Essentially, any decent cracker/hacker could alter the results of the rolls on their client’s memory in real-time and have it high-roll every single roll, or more subtly high-roll at a higher rate that may go undetected. This would of been done had it been possible. Since it has not been done, it is likely not possible and the most likely reason is because all rolls happen server-side (to prevent this very scenario.)

Please reread my original post, I write explicitly I would never expect blizzard to rig anything intentionally (for reasons that you probably don’t agree with.)

Please reread my original post, I link to resources that explain why pRNG might not be sufficient.

This is something that must be tested thoroughly and NOT assumed. pRNG algorithms are certainly mature, but for critical applications might not suffice. I’m not contesting that a pRNG implemented correctly is sufficient for Hearthstone, it most likely entirely is.

What I’m saying is, its very easy to get into trouble (i.e. biasing your randomness) when you use pRNG algorithms concurrently, this is a well known and established fact in computer science.