Cristian

Deck Shuffling Clarification - It doesn't seem random?

570 posts in this topic

On 1/22/2017 at 0:56 PM, Fineberg83 said:

Can we get some input from one of the devs on this topic, please?

I think they are trying to avoid calling you all idiots. This is a tired topic, and method being used here lends itself to a crappy conconclusions. 

If you want to claim that the shuffle is juiced, you need numbers, not anecdotes. You need to compare the numbers the game produces with expected results. But no one here making the argument has a clue what the expected results are, let alone how they compare to the game. 

  • Like 3

Share this post


Link to post
Share on other sites

So, I searched around online and found this little gem at pokerlistngs on a blog post titled: Online Versus Live Play

Quote

1) When you are seeing over eight times as many hands in one session as you do live, you will be seeing over eight times as many bad beats and suck-outs.

2) The quality of players online is sub-par. You get far more huge calls with hands that are completely dominated. That is how a suck-out starts.

3) People remember bad beats; they don't remember that out of the last 20 sets they had, 13 of them got folded to (that's a win) and four of them won a big pot. But they do remember every detail of the last three that lost.People have too much ego. No one wants to admit they made a mistake. Everyone drives in the fast lane on the highway, because no one will admit that they are slower traffic. It's much easier to say that you lost because the site is rigged, than that it could have had something to do with how you played the hand.

Now on this, they are talking about playing 2 to 4 tables at a time to get to that "8 times as many hands" on number one. Still online play moves MUCH FASTER than standard live play. I'd guess that on PP, you're probably seeing 50-60 hands per hour. Live play, 30 per hour is high and probably closer to 10-15 per hour for most low stakes/friend's night type games.

Of course, 2 and 3 are the big ones here when comparing PP to anything else.

 

  • Like 5

Share this post


Link to post
Share on other sites
20 hours ago, ClearConscious said:

We ran a shuffle test on our deck. Here are the results after 2 billion hands:

  Mathematical Probability % of Hands % Difference
High Card 17.4119 % 17.4110 % 0.01 %
Pair 43.8225 % 43.8210 % 0.00 %
Two Pair 23.4955 % 23.4969 % 0.01 %
Three of a Kind 4.8299 % 4.8307 % 0.02 %
Straight 4.6194 % 4.6200 % 0.01 %
Flush 3.0255 % 3.0244 % 0.04 %
Full House 2.5961 % 2.5967 % 0.02 %
Four of a Kind 0.1681 % 0.1682 % 0.08 %
Straight Flush 0.0279 % 0.0279 % 0.22 %
Royal Flush 0.0032 % 0.0032 % 0.03 %

The question is:  Is PP using a physical phenomenon which is expected to be random to supplement their RNG calculation?

Do you have statistics from actual game play?  I'm not doubting your numbers in a test environment.  I'm just curious. 

It just seems to me that more people are hitting the winning card on the river than happen IRL.   (It's just an observation, not an indictment or criticism.)    To that end, could we get a river beat (last card loss) reputation award?  In my opinion, it seems that when people get the "Bad Beat" award, it should be more than 300 reputation.

Great game! Keep up the good work!

  • Like 1

Share this post


Link to post
Share on other sites
20 minutes ago, Zoltan6201 said:

It just seems to me that more people are hitting the winning card on the river than happen IRL.  

You simply can't compare Prominence to RL (pretend money betting versus real money betting). The amount of times I've seen under the gun bet 10xBB's and ALL 5 players call! This wouldn't happen in a million years of RL poker. The reason you are seeing more hands win on the river is simply because of the extremely low caliber of player on Prominence and their willingness to chase cards. It has nothing to do with the RNG.  

  • Like 5

Share this post


Link to post
Share on other sites
1 minute ago, Zanatoo said:

You simply can't compare Prominence to RL (pretend money betting versus real money betting). The amount of times I've seen under the gun bet 10xBB's and ALL 5 players call! This wouldn't happen in a million years of RL poker. The reason you are seeing more hands win on the river is simply because of the extremely low caliber of player on Prominence and their willingness to chase cards. It has nothing to do with the RNG.  

Okay, maybe not cash game RL, but tournament RL.  Especially low buy-in tournaments.  Even more comparable would be online poker for real money which is RL, but different.

I get your point, but it does not invalidate mine.

Share this post


Link to post
Share on other sites

I don't think it matters where it is, once you have more players at the river, you are more likely to see hands improve on the river. In PP (or any pretend money poker) you get more players, more often, still in the hand at the river than in any form of real money poker. Partly because it's pretend money but mostly because it's full of novices who have no idea about pot odds, implied odds, EV, or probability.

Even if PP released their source code you would still have people complain about the software and how it's designed someway to manipulate the outcome. Just like I've seen people in RL card rooms accuse dealers of cheating because someone caught their "one outer" and they busted.   

For every player that loses, there's one who wins. Prominence have nothing to gain by rigging the cards. Plus, it's probably much more difficult to code than using a RNG!

Edited by Zanatoo

Share this post


Link to post
Share on other sites

I have always wondered why some players consistently check all the way thru the river when they are holding pairs, two pairs, or even trips, and then get upset when a donk  player gets a flush or straight on the river.   Whose fault is that!!  Make them challenge your hand!  persuade them that they're going to lose.  If a high card appears in the flop, and everyone calls then you raise to see if there is a trapper doing slow play or to tell everyone that you just tripped.

Go play the Underground games to learn how to win.  The chips you lose are never deducted from your bankroll, yet the winnings are added.

 

Share this post


Link to post
Share on other sites
2 minutes ago, Zanatoo said:

I don't think it matters where it is, once you have more players at the river, you are more likely to see hands improve on the river. In PP (or any pretend money poker) you get more players, more often, still in the hand at the river than in any form of real money poker. Partly because it's pretend money but mostly becasue it's full of novices who have no idea about pot odds, implied odds, EV, or probability.

Even if PP released their source code you would still have people complain about the software and how it's designed someway to manipulate the outcome. Just like I've seen people in RL card rooms accuse dealers of cheating becase someone caught their "one outer" and they busted.   

Maybe you're missing my point because I failed to specify what I was referring to on the river.

I meant that in one to one showdowns, players seem to hit more miraculous river cards than random cards would seem to allow.  Maybe it's luck.  Maybe it's really random.  I'm just saying that IMHO, it looks a bit off.

That is not complaining.  It's an observation.

As a customer, I have the right to inquire if the are"using a physical phenomenon which is expected to be random to supplement their RNG calculation".  Perhaps it's something they did not consider.  Maybe they'll read it and think that it might be a god idea.

They put these forums in place to improve their game. 

Constructive criticism is not "complaining". 

I'm not asking for source code.  I don't see anyone here asking for that.

10 minutes ago, Zanatoo said:

For every player that loses, there's one who wins. Prominence have nothing to gain my rigging the cards. Plus, it's probably much more difficult to code than using a RNG!

I'm not claiming that they are rigging the cards.  Although it's off-topic, rigging the cards would be simple.  Add an increased possibility percentage based upon level of the player.  Given player A is below rank 10, increase their RNG odds by 10%.  Easy.

  • Like 2

Share this post


Link to post
Share on other sites
3 minutes ago, Zoltan6201 said:

Maybe you're missing my point because I failed to specify what I was referring to on the river.

I meant that in one to one showdowns, players seem to hit more miraculous river cards than random cards would seem to allow.  Maybe it's luck.  Maybe it's really random.  I'm just saying that IMHO, it looks a bit off.

That is not complaining.  It's an observation.

As a customer, I have the right to inquire if the are"using a physical phenomenon which is expected to be random to supplement their RNG calculation".  Perhaps it's something they did not consider.  Maybe they'll read it and think that it might be a god idea.

They put these forums in place to improve their game. 

Constructive criticism is not "complaining". 

I'm not asking for source code.  I don't see anyone here asking for that.

I'm not claiming that they are rigging the cards.  Although it's off-topic, rigging the cards would be simple.  Add an increased possibility percentage based upon level of the player.  Given player A is below rank 10, increase their RNG odds by 10%.  Easy.

We do not shuffle the deck.  That is, we don't take an ordered deck of 52 cards and randomize the order of all 52 cards in the deck.  Instead, we randomly choose a card from the deck each time we need a card.

The game server has a RNG for choosing cards for the high card deal and the community cards, plus another six RNGs for the six seats for choosing player hole cards.  These seven RNGs are shared among all games running on the game server.  The main RNG is initialized at game server startup.  The seat RNGs are initialized each time a player is added to that seat number on any table on the server, using a random number generated with the main RNG.  Thus entropy is introduced by the timing of players being added to tables, and by the interaction of multiple tables drawing cards at different times throughout the hands.  We also don't start each hand with an ordered deck - we leave the table's deck in its unordered state from the previous hands, which adds more entropy.

The RNGs are instances of .Net System.Random.

We have done testing with a game server playing 2 billion hands across several tables and found that the resulting data very precisely matches the mathematical probabilities for 7-card poker hands.

  • Like 3

Share this post


Link to post
Share on other sites
5 minutes ago, Farmer John40 said:

I have always wondered why some players consistently check all the way thru the river when they are holding pairs, two pairs, or even trips, and then get upset when a donk  player gets a flush or straight on the river.   Whose fault is that!!  Make them challenge your hand!  persuade them that they're going to lose.  If a high card appears in the flop, and everyone calls then you raise to see if there is a trapper doing slow play or to tell everyone that you just tripped.

Go play the Underground games to learn how to win.  The chips you lose are never deducted from your bankroll, yet the winnings are added.

 

You're making a false assumption. 

  • Like 1

Share this post


Link to post
Share on other sites
5 minutes ago, ClearConscious said:

We do not shuffle the deck.  That is, we don't take an ordered deck of 52 cards and randomize the order of all 52 cards in the deck.  Instead, we randomly choose a card from the deck each time we need a card.

 

The game server has a RNG for choosing cards for the high card deal and the community cards, plus another six RNGs for the six seats for choosing player hole cards.  These seven RNGs are shared among all games running on the game server.  The main RNG is initialized at game server startup.  The seat RNGs are initialized each time a player is added to that seat number on any table on the server, using a random number generated with the main RNG.  Thus entropy is introduced by the timing of players being added to tables, and by the interaction of multiple tables drawing cards at different times throughout the hands.  We also don't start each hand with an ordered deck - we leave the table's deck in its unordered state from the previous hands, which adds more entropy.

 

The RNGs are instances of .Net System.Random.

 

We have done testing with a game server playing 2 billion hands across several tables and found that the resulting data very precisely matches the mathematical probabilities for 7-card poker hands.

 

Can that post be stickied?? :D

Share this post


Link to post
Share on other sites
4 minutes ago, ClearConscious said:

We do not shuffle the deck.  That is, we don't take an ordered deck of 52 cards and randomize the order of all 52 cards in the deck.  Instead, we randomly choose a card from the deck each time we need a card.

 

The game server has a RNG for choosing cards for the high card deal and the community cards, plus another six RNGs for the six seats for choosing player hole cards.  These seven RNGs are shared among all games running on the game server.  The main RNG is initialized at game server startup.  The seat RNGs are initialized each time a player is added to that seat number on any table on the server, using a random number generated with the main RNG.  Thus entropy is introduced by the timing of players being added to tables, and by the interaction of multiple tables drawing cards at different times throughout the hands.  We also don't start each hand with an ordered deck - we leave the table's deck in its unordered state from the previous hands, which adds more entropy.

 

The RNGs are instances of .Net System.Random.

 

We have done testing with a game server playing 2 billion hands across several tables and found that the resulting data very precisely matches the mathematical probabilities for 7-card poker hands.

 

Yeah, I'm getting entropied.  (joke)

I understand that it's RNG.  I get that your test results of 2 billion approximate poker probability. 

I just wondered if you considered "using a physical phenomenon which is expected to be random to supplement their RNG calculation".

I'll take the answer as "No". 

Thanks.

  • Like 1

Share this post


Link to post
Share on other sites

Since some people have asked. Here's some additional details:

We do not shuffle the deck.  That is, we don't take an ordered deck of 52 cards and randomize the order of all 52 cards in the deck.  Instead, we randomly choose a card from the deck each time we need a card.

The game server has a RNG for choosing cards for the high card deal and the community cards, plus another six RNGs for the six seats for choosing player hole cards.  These seven RNGs are shared among all games running on the game server.  The main RNG is initialized at game server startup.  The seat RNGs are initialized each time a player is added to that seat number on any table on the server, using a random number generated with the main RNG.  Thus entropy is introduced by the timing of players being added to tables, and by the interaction of multiple tables drawing cards at different times throughout the hands.  We also don't start each hand with an ordered deck - we leave the table's deck in its unordered state from the previous hands, which adds more entropy.

The RNGs are instances of .Net System.Random.

We have done testing with a game server playing 2 billion hands across several tables and found that the resulting data very precisely matches the mathematical probabilities for 7-card poker hands.

  Mathematical Probability % of Hands % Difference
High Card 17.4119 % 17.4110 % 0.01 %
Pair 43.8225 % 43.8210 % 0.00 %
Two Pair 23.4955 % 23.4969 % 0.01 %
Three of a Kind 4.8299 % 4.8307 % 0.02 %
Straight 4.6194 % 4.6200 % 0.01 %
Flush 3.0255 % 3.0244 % 0.04 %
Full House 2.5961 % 2.5967 % 0.02 %
Four of a Kind 0.1681 % 0.1682 % 0.08 %
Straight Flush 0.0279 % 0.0279 % 0.22 %
Royal Flush 0.0032 % 0.0032 % 0.03 %

 

  • Like 5

Share this post


Link to post
Share on other sites
22 minutes ago, Zoltan6201 said:

Yeah, I'm getting entropied.  (joke)

I understand that it's RNG.  I get that your test results of 2 billion approximate poker probability. 

I just wondered if you considered "using a physical phenomenon which is expected to be random to supplement their RNG calculation".

I'll take the answer as "No". 

Thanks.

I really wish we were recording the discussion we just had with our designer/engineers XD.

 

Share this post


Link to post
Share on other sites
1 hour ago, Zoltan6201 said:

You're making a false assumption. 

What the hell do you think an opinion is.?

If we are expected to read all your opinioned post, surly you can respect the opinion of others.

Share this post


Link to post
Share on other sites

Exactly one of the things i mean. I get 10s, board pairs 10s, enemy gets tens. Not jacks, not eights, not good hands in general, no exactly the same stuff. 3-5 boards in a row, and the next table might as well have "hot 10s". Its understood stuff happens within 20000 or a billion hands, but you can make a videomontage in just 200 hands here. Last table i wouldnt even bother playing without a 6 or 9. 999 once on board, 99 5 times, 6 6 3 times, 6 9 6 3 times. And its not like there were 1000s of boards played per game. Well at least everybodies darling 69 got some love xD " Unnatural boardtexture" sb else called it in a thread before and thats also my gripe as it makes leading out pretty pointless. Of course you only notice if the awareness is already present , but its really hard to ignore at the frequency.

Share this post


Link to post
Share on other sites
On January 24, 2017 at 5:21 AM, SevenSideSammy said:

The only "quirky" thing I notice with this algorithm is repeating pockets. For example, getting pocket 10's and then seeing a person get 'em  the very next hand.

But even then it's not the shuffle, it's me picking up on patterns that are really spread out and ignoring all the in-between.

Oh the joys of the human mind.

^ refers to this lol

Share this post


Link to post
Share on other sites

I said it once back in August. I have since learned from experience that the deck plays fine. Sure, I've vented about bad beats on threads where the topics were essentially "vent here about bad beats" but long ago gave up the idea that the deck was rigged and have in fact long argued against it. 

It's the players, not the deck. 

  • Like 3

Share this post


Link to post
Share on other sites
On January 24, 2017 at 6:01 AM, Jayhawker said:

I think they are trying to avoid calling you all idiots.

Glad we have you for that-_-

Share this post


Link to post
Share on other sites
18 hours ago, Farmer John40 said:

I have always wondered why some players consistently check all the way thru the river when they are holding pairs, two pairs, or even trips, and then get upset when a donk  player gets a flush or straight on the river.   Whose fault is that!!  Make them challenge your hand!  persuade them that they're going to lose.  If a high card appears in the flop, and everyone calls then you raise to see if there is a trapper doing slow play or to tell everyone that you just tripped.

Go play the Underground games to learn how to win.  The chips you lose are never deducted from your bankroll, yet the winnings are added.

 

No one in this thread forums is checking to the river, getting a bad beat and questioning the RNG.

These are intense battles that go back and forth leaving two players in a showdown.  The lower hand seems to win more often on a river card that seems impossible.   It's worth questioning.  Not complaining. 

Just asking if they considered "using a physical phenomenon which is expected to be random to supplement their RNG calculation".

It really is a good idea.  It's a way to offset the true lack of random number generation inherent in computational calculations.  It would be innovative.  It's not that hard either.  It could be an atomic clock digit.

Since we have now learned that they do not shuffle the cards at all, isn't anyone alarmed?

"Go to the underground"?  Really? 

"Surly".  Not at all.

Rumack: Can you fly this plane, and land it?

Ted Striker: Surely you can't be serious.

Rumack: I am serious... and don't call me Shirley.  -Airplane movie  from http://www.imdb.com/title/tt0080339/?ref_=ttqt_qt_tt

 

Edited by Zoltan6201
  • Like 2

Share this post


Link to post
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now