At least it feels that way, I personally think there is something wrong with the coding.

I will try to explain, let's pick a random summoning card like Netherwind Portal. There are in total 101 of 4 Mana Minions in standard to RNG from. So let's say 1 in 100 chance to calculate more easily. Last 2 games ago I had 2x Kargath Bladefist (not that I am complaining)

First time it is 1/100 but the second time it happening is again 1/100, but both for the second time it happening becomes 1 in 10000?

Just now I finished a game and I had 2 x Kargath again. Again I am not complaining it is a nice card to get, better then getting a desert obelisk with Jandice Barov :-) But it really doesn't feel like random, I most likely will get a lot of flamers again (like if Blizzard is their Family Company) but again this suited me but not my opponent I guess, I am not really complaining, I just want to point out that it feels fishy in my experience.

I believe this is called confirmation bias and that, counterintuitively, the odds aren't actually 1 in 10000 but only 1 in 100. I'm not a statistician (but I have studies some stats), but the first event "doesn't count" unless you're already looking for it in advance. The explaination here is that there will be lots of times that the opponent (or you) doesn't get a Kargath, but you don't register and take note of these occurrences because nothing intersting happened. Once you get the first Kargath you're now interested in what the next one is, but because you only took note when you got Kargath you can't count that first event in the probability.

I probably haven't explained that very well - like I said I'm not a statistician.

There was a really interesting article about Paul the psychic octopus (if you remember he was an octopus that predicted world cup scores alledgedly) on the website understanding uncertainty where this phenomenon is explained in more detail than I can understand!

I believe this is called confirmation bias and that, counterintuitively, the odds aren't actually 1 in 10000 but only 1 in 100. I'm not a statistician (but I have studies some stats), but the first event "doesn't count" unless you're already looking for it in advance. The explaination here is that there will be lots of times that the opponent (or you) doesn't get a Kargath, but you don't register and take note of these occurrences because nothing intersting happened. Once you get the first Kargath you're now interested in what the next one is, but because you only took note when you got Kargath you can't count that first event in the probability.

I probably haven't explained that very well - like I said I'm not a statistician.

There was a really interesting article about Paul the psychic octopus (if you remember he was an octopus that predicted world cup scores alledgedly) on the website understanding uncertainty where this phenomenon is explained in more detail than I can understand!

I get you, I have just read the article about the octopus, nice remark, nice read too. I am not a statistician as well, but I got too much Kargath's in a day, 3 in total and I played 12 games, I did not get Kargath 6 times.

So I played 9 times Netherwind Portal today, 3 times it gave me Kargath today. That just feels too much to be random, because for me it is 1 to 3 (33%), considering each time the odds this happening is 1/100 if you are waiting for it.

Nothing in this game is truly rng. It's all calculated and picked for you by algorithms. Gotta keep those numbers in check to keep the people playing and paying.

I believe this is called confirmation bias and that, counterintuitively, the odds aren't actually 1 in 10000 but only 1 in 100. I'm not a statistician (but I have studies some stats), but the first event "doesn't count" unless you're already looking for it in advance. The explaination here is that there will be lots of times that the opponent (or you) doesn't get a Kargath, but you don't register and take note of these occurrences because nothing intersting happened. Once you get the first Kargath you're now interested in what the next one is, but because you only took note when you got Kargath you can't count that first event in the probability.

I probably haven't explained that very well - like I said I'm not a statistician.

There was a really interesting article about Paul the psychic octopus (if you remember he was an octopus that predicted world cup scores alledgedly) on the website understanding uncertainty where this phenomenon is explained in more detail than I can understand!

I get you, I have just read the article about the octopus, nice remark, nice read too. I am not a statistician as well, but I got too much Kargath's in a day, 3 in total and I played 12 games, I did not get Kargath 6 times.

So I played 9 times Netherwind Portal today, 3 times it gave me Kargath today. That just feels too much to be random, because for me it is 1 to 3 (33%), considering each time the odds this happening is 1/100 if you are waiting for it.

I am buying a lottery ticket as we speak :-)

The guy who published it (David Spiegelhalter) has a really interesting take when it comes to stats. I saw him speak once at an epidemiology lecture (and he also does stuff for the BBC in the UK) and he has a really nice way of making things understandable while you're listening.

I believe this is called confirmation bias and that, counterintuitively, the odds aren't actually 1 in 10000 but only 1 in 100.

You are actually mixing two things. Of course every single occurrence of the card played is a 1/100 to get a specific result. But for having two of the same result twice in a row odds aren't actually 1/100, just because every roll is a separate one. Yes, you have always 50% chance to hit tails or heads with a coin flip and it's not like 5 tails in a row increase a chance to get heads next time. And if you hit Kargath on a first roll, you still have the same odds - 1/100 - to roll it for the second time in a row. You are right. But the actual odds of getting 5 tails in a row aren't 50% and neither having two Kargaths from the Portal is 1/100. It's actually much less.

Granted, this proves nothing, because unless chances are literal, definitive 0% everything can happen so judging entire algorithm of Hearthstone's RNG just by one unlikely occurrence is very inappropriate.

Nothing wrong with algorithms or randomness, it's just people who don't understand randomness.

Unless you personally checked the code I would be careful with such statement. For example number generator in Java language heavily (more than 10 times) favoured "0", when generating integers from any range (no idea if it does that still, I haven't used Java since I got my diploma).

As to whether hearthstone RNG is correctly coded - hard to say. Personally I doubt that Blizz uses any conditional branching ("if" statements) to make the game "more interesting" , a.k.a. Zephrys effect. They already have hard time to make this one card work correctly so implementing such mechanism for every random card, while not making it too obvious, likely outweights potential gains.

Much more likely is that their RNG doesn't have uniform distribution in which case some outcomes are more likely (for example lets assume our card generates 1 out of 100 minions at "random". In large enough sample all minions should be generated about equally. But if ther RNG is similar to Java's [begining of my post] than you will encouter first minion [most programming languages start numbering for "0"] way more often).

Just like the real world, hearthstone is filled with random chaos, but your attention is drawn to moments when stars align, this is what you remember, this is what you focus on, you forget/ignore the rest. People see patterns in random chaos as well, that doesn't mean there are any.

The human brain is shit at dealing with stuff like this, we want to understand and predict events for psychologuical security, but if we have no understanding we invent our own in order to try and grasp the situation and predict outcomes.

Superstition, victim blaming, religion, conspiracies, etc.

I believe this is called confirmation bias and that, counterintuitively, the odds aren't actually 1 in 10000 but only 1 in 100.

You are actually mixing two things. Of course every single occurrence of the card played is a 1/100 to get a specific result. But for having two of the same result twice in a row odds aren't actually 1/100, just because every roll is a separate one. Yes, you have always 50% chance to hit tails or heads with a coin flip and it's not like 5 tails in a row increase a chance to get heads next time. And if you hit Kargath on a first roll, you still have the same odds - 1/100 - to roll it for the second time in a row. You are right. But the actual odds of getting 5 tails in a row aren't 50% and neither having two Kargaths from the Portal is 1/100. It's actually much less.

Granted, this proves nothing, because unless chances are literal, definitive 0% everything can happen so judging entire algorithm of Hearthstone's RNG just by one unlikely occurrence is very inappropriate.

I'm not sure I am mixing two things (but as I said I'm not a statistician). My point about confirmation bias is that you only take note of the first event after it already occurs - in this example you or your opponent get a Kargath. Because that event has already happened (and you aren't noticing all the times when you didn't get Kargath) you cant include it in the overall calculation for two Kargaths to be drawn.

This is different from the idea of "inheriting" odds from the previous outcome (e.g. toin coss multiple times). It's true that each event is independant from those that came before (i.e tossing 6 heads in a row does not change the odds of the outcome for the 7th coin toss, which is still 50/50) however the odds of tossing 7 heads in a row before you throw the first coin can still be calculated.

To make a coin analogy of what I'm saying, if you randomly toss coins all the time but only really pay attention after you get 4 heads in a row, you might believe that your coin is baised when you end up flipping that 5th head in a row, because to you it would seem like it's happening more often that it should. In reality it's because you don't remember all those occassion when you didn't get 5 in a row.

Have a read of the psychic octopus paper I mentioned - that's the essence of what I understand is discussed in there (and he is a statistician).

Human brains will start seeing patterns in even the smallest coincidences – and in this particular game, it’s likely that certain minions (Kargath) will come next to one another. You just notice this coincidences. Because random events doesn't depend on previous event, but our brain whats to think that they do

“The problem is that to humans, truly random does not feel random,” Mattias Petter Johansson, a Spotify developer, wrote in an internet post earlier. “So we got tons of complaints from users about it not being random.

At least it feels that way, I personally think there is something wrong with the coding.

I will try to explain, let's pick a random summoning card like Netherwind Portal. There are in total 101 of 4 Mana Minions in standard to RNG from.

So let's say 1 in 100 chance to calculate more easily. Last 2 games ago I had 2x Kargath Bladefist (not that I am complaining)

First time it is 1/100 but the second time it happening is again 1/100, but both for the second time it happening becomes 1 in 10000?

Just now I finished a game and I had 2 x Kargath again. Again I am not complaining it is a nice card to get, better then getting a desert obelisk with Jandice Barov :-)

But it really doesn't feel like random, I most likely will get a lot of flamers again (like if Blizzard is their Family Company) but again this suited me but not my opponent I guess, I am not really complaining, I just want to point out that it feels fishy in my experience.

Maybe it is Blizzard's Kargath Prime day? Shit happens I guess? I indeed hate it when I get desert obelisk, that card can go in to the trash bin.

I believe this is called confirmation bias and that, counterintuitively, the odds aren't actually 1 in 10000 but only 1 in 100. I'm not a statistician (but I have studies some stats), but the first event "doesn't count" unless you're already looking for it in advance. The explaination here is that there will be lots of times that the opponent (or you) doesn't get a Kargath, but you don't register and take note of these occurrences because nothing intersting happened. Once you get the first Kargath you're now interested in what the next one is, but because you only took note when you got Kargath you can't count that first event in the probability.

I probably haven't explained that very well - like I said I'm not a statistician.

There was a really interesting article about Paul the psychic octopus (if you remember he was an octopus that predicted world cup scores alledgedly) on the website understanding uncertainty where this phenomenon is explained in more detail than I can understand!

Tech for tokens,... no token enemies.

Remove tech, get token enemies.

Tech for weapon... no weapon enemies.

Remove tech, get warriors all over.

Tech for mage. No mages.

Remove tech, get mages all the the time.

TECh fo SeCrets... AHHHHHHHHHHHHHHHHHHHHHHHHH INSANITY...

and note, how the "neutral matchmaker will count down all of the sudden, when you change your tech..."

there are two things in the world that smell like fish, and one of them is the matchmaker

I get you, I have just read the article about the octopus, nice remark, nice read too. I am not a statistician as well, but I got too much Kargath's in a day, 3 in total and I played 12 games, I did not get Kargath 6 times.

So I played 9 times Netherwind Portal today, 3 times it gave me Kargath today. That just feels too much to be random, because for me it is 1 to 3 (33%), considering each time the odds this happening is 1/100 if you are waiting for it.

I am buying a lottery ticket as we speak :-)

Nothing in this game is truly rng. It's all calculated and picked for you by algorithms. Gotta keep those numbers in check to keep the people playing and paying.

Nothing wrong with algorithms or randomness, it's just people who don't understand randomness.

The guy who published it (David Spiegelhalter) has a really interesting take when it comes to stats. I saw him speak once at an epidemiology lecture (and he also does stuff for the BBC in the UK) and he has a really nice way of making things understandable while you're listening.

You are actually mixing two things. Of course every single occurrence of the card played is a 1/100 to get a specific result. But for having two of the same result twice in a row odds aren't actually 1/100, just because every roll is a separate one. Yes, you have always 50% chance to hit tails or heads with a coin flip and it's not like 5 tails in a row increase a chance to get heads next time. And if you hit Kargath on a first roll, you still have the same odds - 1/100 - to roll it for the second time in a row. You are right. But the actual odds of getting 5 tails in a row aren't 50% and neither having two Kargaths from the Portal is 1/100. It's actually much less.

Granted, this proves nothing, because unless chances are literal, definitive 0% everything can happen so judging entire algorithm of Hearthstone's RNG just by one unlikely occurrence is very inappropriate.

What if yogg saron box , cast 10 same spells in two different games in a row, that its normal to happen i mean its not that fishy?

What if yogg saron box , cast 10 same spells in two different games in a row, that its normal to happen i mean its not that fishy?

Unless you personally checked the code I would be careful with such statement. For example number generator in Java language heavily (more than 10 times) favoured "0", when generating integers from any range (no idea if it does that still, I haven't used Java since I got my diploma).

As to whether hearthstone RNG is correctly coded - hard to say. Personally I doubt that Blizz uses any conditional branching ("if" statements) to make the game "more interesting" , a.k.a. Zephrys effect. They already have hard time to make this one card work correctly so implementing such mechanism for every random card, while not making it too obvious, likely outweights potential gains.

Much more likely is that their RNG doesn't have uniform distribution in which case some outcomes are more likely (for example lets assume our card generates 1 out of 100 minions at "random". In large enough sample all minions should be generated about equally. But if ther RNG is similar to Java's [begining of my post] than you will encouter first minion [most programming languages start numbering for "0"] way more often).

Anything to back it up?

And no, not normal, but POSSIBLE. People win lotteries, get struck by a lightning and so on, right?

Here is the full versions of 2 consecutive matches, look closely and compare:https://clips.twitch.tv/GracefulFancyDuckSuperVinlin

https://clips.twitch.tv/AmericanSpicyApeBudStar

In both games Alex gave him Deathwing and Kalecgos in the same order.

In both games Kalecgos discovered Luna's, Cinderstorm, Puzzle Box in the same order.

And the Box also casted 8 same spells and in the same positions:

Topsy Turvy

Fresh Scent

Shadow Bolt

Landscaping

Rising Winds

Mindgames in #1, Polymorph in #2

Earth Shock in #1, Rocket Boots in #2

Pick Pocket

Hellfire

Aeroponics

Just like the real world, hearthstone is filled with random chaos, but your attention is drawn to moments when stars align, this is what you remember, this is what you focus on, you forget/ignore the rest. People see patterns in random chaos as well, that doesn't mean there are any.

The human brain is shit at dealing with stuff like this, we want to understand and predict events for psychologuical security, but if we have no understanding we invent our own in order to try and grasp the situation and predict outcomes.

Superstition, victim blaming, religion, conspiracies, etc.

What are the odds? Maybe there is a deterministic RNG, meaning that certain input range has a certain output range?

I'm not sure I am mixing two things (but as I said I'm not a statistician). My point about confirmation bias is that you only take note of the first event after it already occurs - in this example you or your opponent get a Kargath. Because that event has already happened (and you aren't noticing all the times when you didn't get Kargath) you cant include it in the overall calculation for two Kargaths to be drawn.

This is different from the idea of "inheriting" odds from the previous outcome (e.g. toin coss multiple times). It's true that each event is independant from those that came before (i.e tossing 6 heads in a row does not change the odds of the outcome for the 7th coin toss, which is still 50/50) however the odds of tossing 7 heads in a row before you throw the first coin can still be calculated.

To make a coin analogy of what I'm saying, if you randomly toss coins all the time but only really pay attention after you get 4 heads in a row, you might believe that your coin is baised when you end up flipping that 5th head in a row, because to you it would seem like it's happening more often that it should. In reality it's because you don't remember all those occassion when you didn't get 5 in a row.

Have a read of the psychic octopus paper I mentioned - that's the essence of what I understand is discussed in there (and he

isa statistician).What?!? No.

Human brains will start seeing patterns in even the smallest coincidences – and in this particular game, it’s likely that certain minions (Kargath) will come next to one another. You just notice this coincidences. Because random events doesn't depend on previous event, but our brain whats to think that they do

“The problem is that to humans, truly random does not feel random,” Mattias Petter Johansson, a Spotify developer, wrote in an internet post earlier. “So we got tons of complaints from users about it not being random.

May be the RNG is intended to be dramatic :P~

:P~~~