A proof that card draw is RIGGED

General Discussion
21

There was a thread some weeks ago where I wrote about probability of independent events. You participated in it with… well… theoretical knowledge instead of gibberish. :slight_smile:

1 Like
22

blizzard benefits from it because it prevents players from wining too much ((wins=better rewards=less (or 0) real money spent on packs/arenas entry fee ))

23

…but in arena your opponent is at the same record as you.

1 Like
24

I remember you missing the point in that thread as well. I see a theme developing…

25

I see your point, but the way that Arena works, there’s always the same number of people with a certain record. The result of rigging or not rigging is exactly the same for Blizzard.

EDIT: you can find the percentages here: https://hearthstone.gamepedia.com/Arena#Exact_sequence_of_matches

26

Ugh, looking at that table makes me cringe at it’s mathematical absurdity. It suggests there is a statistical likelihood of you achieving any given final result without taking into account any of the multitude of variables that affect any given outcome. This is what we call ‘lying with numbers’ where you can make statistics say pretty much anything you want if you lay them out correctly.

I mean, just take the first row as an example. It suggests you have a 50% chance of winning regardless of what cards you or your opponent have picked…that’s just stupid

1 Like
27

I just read it as the % of people that get each result and in that regard it’s just statistical fact as arena is effectively a massive swiss style tournament in disguise.

Doesn’t say anything about individual deck/ability but because every game your matched by record and someone will always lose (or both a get a do-over in a draw), that is the distribution of results (It’s a little off as at high wins you can get matched to slightly different records, plus you’ve got some people who hold x-2 runs to cash in on the ‘free’ pack, every time arena is changed).

28

Someone bring the OP some aloe vera. He just got burned!

1 Like
29

@Rayven, either I didn’t get what you said, or you didn’t get the table. Think of it this way: on any given pool of players, disregarding ties and assuming all finish their arena runs, there’s a path all of them will follow through that table, considering that players are matched according to their current arena score of wins and losses and that at each “cell” half win the match, half lose. Hence there’s no likelihood, it’s all deterministic, given the assumptions. It’s a way to find the distribution of players according to their final win/loss score.

Does that make more sense to you?

30

You actually picked op on this. Now when I make my troll post with the facade “Why I stopped playing arena” I can use your words. <3
You really got me excited now. :laughing:

1 Like
31

Math and conspiracy right there :stuck_out_tongue:

Those two can be combined for solving some general issues in every aspect of life to, but that comment just did not make any sense.

Yeah micromanagement with probally results that cant even get implemented if they wanted to since (( if one loses , another one need to win right ? ))

32

Irregarding the veribility, how does it not make sense?

33

It could be solved by implementing BOTS that cant even get identified by humans as playing against a bot so they can reduce the gold output with bot win rigging conditions ! :scream:

But i doubt it …

34

So, proof comes later or…

2 Likes
35

Never preorder!

(6 characters)

36

Too many nerdy blizzard fanboys here…well keep wasting ur money …joke’s on u :slight_smile:

2 Likes
37

Doesnt need it since the monetisation model for arena is already finetuned, more people lose then win and to break even you need 3 wins. OP asking for this kind of proof is like talking to someone who believes the earth is flat, you literal need to take him on a rocket into space to ‘‘disprove’’ it.

And Vlad posted the % arena model already , but altering the win conditions for players in this model cant be done.
They even upgraded the quest rewards in the past a little bit.

Never spend a dime on hearthstone, and setting your view straight makes me a fanboy… XD

And for your information, you responded with the orange one.
Imgur

1 Like
38

That table is not a probability distribution for a single draft, but a statistical distribution of all drafts overall. It is built on two assumptions, though.
The matchmaking process for Arena is known. It tries to find an opponent that is at the same number of wins and losses as you. If it fails to find one in a short enough time, it will start looking for opponents that are “close” in wins and losses. New players (I believe up to 5 arena runs) go into a separate pool, all others go into the main pool.

The assumptions underpinning the table are:

  1. Blizzard is not lying about the matchmaking and this is the actual process.
  2. The matchmaking always gives you an opponent at the exact same level.

I personally believe the first assumption to be true, and the second to be true at the lower tiers but not true at the higher end of scores (and possibly also not true for the new-player matchmaking pool, since this pool will be much smaller).

Let’s look at the first row. Under the above assumptions, every player at 0-0 will find an opponent at 0-0. In the unlikely event of a tie, the game is effectively discarded (both players will remain at 0-0) so we can ignore ties for the purpose of this table. That means that every match at 0-0 will have one winner (advancing to 1-0) and one loser (advancing to 0-1). That is what the table says. And the same goes for every row.
As mentioned, I do not believe matchmaking to always be perfect at the hig end. If you are at 10-1, chances are your opponent is at 10-2, or 11-1, or so. They will still be close. But this would impact the table a bit - I just don’t know how exactly and I also don’t think it would be a big impact.

But like I said, this is all built on two assumptions. I think that most people would agree that if you buy into the first assumption, then the second assumption is very credible at least for the lower scores. The matchmaking pool obviously diminishes with each next win, so at the higher scores it becomes harder.

But what if you don’t believe the first assumption? What if Arena actually does not try to pair you with someone at the same score, but uses a completely different matchmaking? Of course, if 0-0 players regularly are paired with 8-1 or 10-0 players, then the table falls apart at the seams.
Well, I don’t think Blizzard would lie about their matchmaking. But if you, or anyone else, does think that matchmaking doesn’t work as promised; if you, or anyone else, believes that 0-0 and 1-1 players are regularly playing vs 8-1 players, then there is an easy way to prove that. Play some arenas. After each game, send a friend request to your opponent and if they accept it ask what their arena score was at the start of your game. Compare that to yours.
If in a set of 10 or 20 observations, you get more than 3 where the opponent and you were 3 or more wins apart, then I will accept that as proof that matchmaking is not as Blizzard tells us it is.
If in that same set of 10-20 observations, you get zero where there is a 3-or-more win difference, then that is of course still not solid proof - but it does add a lot of credibility to the assumption that Blizzard’s matchmaking algorithm is in fact exactly the way Blizzard tells us it is.

39

I fully understand all that. What I’m trying, obviously with difficulty, to say is that that table tells you absolutely NOTHING. “Your first match up in arena has a 50% chance of finishing 1-0 or 0-1”…no fcking sht!!!

It’s completely useless information that will mislead the majority of people who don’t understand the table at all to look at it and think that it shows the likelihood of them getting a particular score by the end of their run.

It takes nothing into consideration at all and is not any kind of representation of true results of arena runs

40

Unless they changed it, they actually go into the same pool but are given a “fake loss” to in theory make there match-ups easier. So game 1 is actually against someone with a 0-1 record, game 2 will be against either a 1-1 or 0-2 record. Think it only lasts for the first 2 runs, not 5.