one sided dealer

[X Barbiqute X]

Hardwood is the best site availabel on the net and its unfortunate that we cannot enjoy it 100%.

i know this subject has come up many times but again its frustrating that you lose game after game to 1 sided dealer. For the past few days, my pard and i have been bidding total of 2 or 3 and opp team is bidding either 7 and nil or total of 8 bids. Can some 1 tell me how one can enjoy a game like this? i was wondering will there ever be a change
in these preset random cards? i have played in few game sites and i have never encountered such cards as i have in hardwood. i remember games that i used to play that have been won by a bag or a point. Where are those games? We want fair deals.

ps: also seats should not matter, where else in hardwood we have to see red or blue seats are hot?

[Joe Andrews]

The RDG beat goes on, post after post, thread after thread!

The unbalanced deal situation is a familiarly recurring motif on many sites. It is not a problem which is limited to HW. Go to almost any Internet games site, play for a while, and you observe the same scenario.

And there is no easy solution. It is also a frustrating problem which affects many "fresh deal" card games, not just Spades. You clearly outplay the opponents, and then they get the big hands at the end of a game. Or you are on the short end of bad cards for a whole game, and get croaked by 300 + points! Then you "blow out" the opponents in the next game! As I said before, you are going to lose apprioximately 25% of the Spades games you play, and guess what? - you are going to win approximately 25% of the Spades games you, play - no matter what you do. The RDG will decide the outcome, one way or the other. Sometimes the opponents will have two or three Nils to your none or they will outbid you every hand. Then your turn will come to have the Nils and better hands. It is the other 50% of the games, where the cards are approximately equal for both sides, which will test the skill of the players.

And the problem occurs in "live" events, as well. For the past two years, my partner and I have participated in a major "live" Spades event, and made the playoffs, outplayed the ops -- only to lose to big Nil hands (Nil / 8, Nil 7) on the last hand of a game. Then there have been other times where "lady luck" allowed me to steal a lost game.

Is there a program that will deal balanced hands for both sides? Perhaps the "gurus" at HW can answer this. And I sure don't want to harp on the system that the ACBL (Bridge organization) utilizes at their events.

Go with the flow - its part of the game.

Footnote - Maybe HW could install a "surrender" or "resign" feature, where you could voluntarily quit a hopelessly lost game, and lose only half of the ratings points. (The opponents would still get credit for a win, and the appropraite # of points). You would still be charged with a loss, and not an "abandon". Any thoughts?

[B O W]

Nicely said Mr Andrews, at least you havent used any sarcasmn in your reply.

[Dead Presidents]

Its part of the game...it happens...its always going to happen...

I believe we get good cards more than bad...
We just dont notice it as much as getting crap...
Its even worst in live tournaments...
In live tournaments you dont get a chance to comeback as frequently as you do in hardwood...

Hardwood dealer makes more people than breaking them.

[Dust In The Wind]

The problem with RNG programs is percentages of the possible times a particular hand will be dealt and getting that set accurately to reflect a true deal of the cards. Also repeat percentage or seeing the same hand within a cirtant amount of time for any one or group of individuals.

The setup on this would be hugh because you would have to take into account each individual in a universe of a certain number of games AND combine that with each other individual within the game they now sit in a particular game room number. Which criteria would use you over another to determine which hand could be left in the selective pool for the individuals within the room to show "FAIRNESS".

I work with Lottery Universes and random generators with the games parameters designated by the Lottery Lawmakers of any particular game and each has their own specifications as to how "Winners" "Common" and "Etc" are to be distributed within a string of lottery tickets (also call a book).

You say how does this relate to Cards.... it is the exact same thing, you have so many of each card as you do of lottery winners and such and you create the possibilities or end result of the deal or layout within a ticket.

All I can say is we have "RULES" that govern how the tickets can be printed in succession IE: you would not see to $100,000 dollar winners printed back-to-back. I cannot go into detail as to how it is done.... sorry classified but the one example is a tip of the iceberg.

The problem of the RNG program is not the program itself but in it's memory of what has been selected for a random hand and the odds of the remaining hands to be dealt next FOR ANY GIVEN PLAYER.... this is key and why it fails.

IF 4 players sit at a table and play with a deck of cards odds are that each will have the same opportunity with good and bad cards and the superior player skills will determine the outcome of the match.... Darwin and theory of probability and plain old mathematics.

You say OK.... problem identified, how do we fix it.... Standard RNG programs have been around for a long time, how can it be modified to reflect true "CHANCE"..... It will take a while to create, but would anyone buy it? and why? This is not a money game! At least here.

To fix the RNG...

1) Program must determines ODDS of each individual hand to be dealt and probability of that same hand or one simular to appear on the hand to be dealt to the individual at each position at the table
2) Program must remember each players X number of dealt hands (TBD)
3) Where does it reset to zero.... on each individual
4) It must also determine if Player Y is also Player Z under another ID or does it treat each individual name as a separate Player
5) How will when a conflict occurs between player "ODDS" will that conflict be determined

This is not to say randomness is no longer the rule it's to say as an "Individual" plays in his universe of possible deals certains ones will not be seen until the reset because of the overall odds of seeing that hand has been dealt and that the "Normal" hands they are seen are still within the pool of possible hands they may receive. To further explain this say that player 1 gets a "Hand of a Life Time" his remaining universe will no longer have that hand as a random set of cards dealt to him until he has seen a certain number of hands and a RESET of luck has occured.

There is much more involved here and ruleings as to how a card site would like it to be designed would also take part in creating it, but if you say just make it so it's fair........ I've never seen totally fair cards or a short period of time, someone will get the winning hand.

JUST DUST

PS - No I am not ready to lock myself up for a few weeks to create this, but it can be done and will be willing to advise or help another willing tackle such a project. Better be good with Math. No I can not use our program to create the same for cards.

[psunuke]

Whoa whoa whoa, hold on here, I will have to disagree STRONGLY here! I personally use something called the Monte Carlo method to use random numbers to simulate particle systems and have have enough probability and statistics background to believe you are barking up the wrong tree. Remembering who did what and such is exactly NOT what to do. The deal must be random, regardless of the past. Odds of somethign occuring on the 2nd deal must be the same as the first deal, there is no conditional probability. Probability remains the same. If not you will change the variance of the algorithm, and it will not be RANDOM. All that is crucial is the algorithm itself representing true random behavior, nothing else. If you decide to balance hands, you are not dealing a TRUE random deal anymore and it doesn't reflect reality!!

Now, that said, I believe the algorithm has been dealt with many times in the forum and I think I even recall Jonas giving out some info for people who want to test it. I have no idea if anyone looked more into it or if HW has, but I believe this is the ONLY possible issue. In order to test somethign like this you just must play out MANY hands and check each outcome to it's expected KNOWN probability, then you can you statistical tests to see whether your hypothesis is correct for what you predict to some statistical confidence.

As a little aside, one would also have to check how the algorithm handles multiple deals simultaneously and if this changes the actual algorithm then it must be tested under different dealing circumstances to be certain that it remains random.

Poker rooms often use player mouse entropy(usage of the mouse movements of players) combined w/digits of the cpu clock or some other ingenious scheme to ensure randomness.

And to address Joe's idea about half the rating points, for one, I never give up (not so important), and two, it is not practical as someone losing can just lose worse to lose less.

Mike

[gladius]

Psunuke hit the nail on the head with the hammer. I was nearly rolling on the floor in pain from laughter when I read Dust in the wind's response. I use RNGs to simulate reality all the time. Here is the simple low down. The program generates a random number, using a "seed". This seed can be based on many different things as psunuke said. This random number determines the card you get. The next random number determines the next card and so on. To ensure randomness, your RNG has to have a long enough stride to ensure there is no cyclic pattern that could be easily repeated. These random number lists are billions of numbers long and reset periodically. This is way overkill for cards. Your cards are random. The cards you are dealt the previous hand have absolutely nothing do the with the cards you get the current. End of story.

If they did, then it would not be random. Is it possible to have a string of great hands? Absolutely and also probable.

The main problem rests in the way peoples minds are put together. You are way more likely to remember abnormalities, than the truely random behavior. As a result, you starting thinking things are not random and dealer is not dealing random hands. If you track it, as Jonas once pointed out, you will see the cards are truely random and pass all the statistical tests. The problem with randomness is in your mind.

PS.
The math is elementary enough an eight grade student should be able to do it. I am sorry you have to lock yourself away to perform addition subtraction and multiplication.

[Dust In The Wind]

Does it simulate the shuffle of cards from the perspective of cards thrown in after a hand and randomly dealt or from the deck as it would be from a fresh pack each time.

I personally don't have a problem with the deals but can see where some people feel that it is not fair to them in particular.

If 4 players on for different routes to a particular game played at a specific time will the deal be the same as if they met an hour later at a different table? In all probability yes, but it is likely they would not if they could be in the 2 places at the same time.

It is a true RNG to take all hands as possible on each hand and from what perspective a fresh deck or the randomous of the order of the cards.

I personally use the "CUT-DECK" option to change my luck of the deal and do so in real cards too by tapping the top card when I do not want to cut or take the cut otherwise.

If you want a more even split of cards and let skill dictate the winner then the randomous of the RNG has to change in it's formulation on the odds of getting certain cards on any given hand. The way it is done here is very simple each hand had the possibility of pulling 13 spades as the next one, is it likely 52 cards 4 suits and figure your odds.

What some universes or odds hold within their pattern are the percentages/likelyhood that a certain combination would occur and take that into consideration in the RNG. This is from each individuals perspective.

JUST DUST

PS - Gladuis, personal attacks are not necessary in the forum, I happen to be very good at math. I was offering a solution to what is deemed unfair and have gone through the RNG calculations and how cut-deck and many other features here effect the outcome of the deal. As Psnuke put it in how many games being played in how many rooms with how many players how is the RNG set to work and this is across the board on not only spades but the other games as well. If each individual room uses it's individual RNG no matter who sits within the reset.

You don't see the same dealer dealing all the cards at each table from the same deck at any Casino or tournament. This is were I see the possible fault.

[psunuke]

Dust, I think that what you say in your first statement here is the problem, although I think what needs to be made is a clarification for us all to understand properly.

I think what you intended on is a way to "rig" for lack of a better word, to see more of an even distribution, this would be biasing and wouldn't reflect a true natural deal, although it is a valid thing to use if everyone wanted to use such a thing. But what i think is necessary to understand that after you deal a hand, that outcome has NO impact on the probability of the next hand, in reality.

It's like if you flip a coin and it comes up heads, that doesn't skew the probability of a tails coming up on the next flip to greater than 50% (fair coin). Now if we're counting the number of streaks of say 5 heads in a row or whatever kind of combination we wanted out of so many set flips, we can use basic probabliity theory to find out what we should EXPECT from such an experiment, and then use some statistical anaylsis to validate whether any observed differences are out of the "ordinary". Same goes for the cards. Every deal shoudl be 100% completely random, and the RNG should NOT be affected by other rooms w/other players. I have no clue as to the acutal algorithmic structure used in HW, so I won't really comment on that.

I'm quite certain that the game play would be different if we did skew the deal to be a more "average" deal instead of actual deal, and I woudl suspect, at first that the better players will make out better, but perhaps not, as lets take shooting for example. A very good player may make an easy shoot out of those far from average hands, while a bad player still gets stopped by the good player. If in fact things are more average it may actually be a detriment to the good player, but there will be things on the other side of the fence too, and some compensatory effect will take place resulting in the final outcome of which I don't think any of us can predict unless we did it. Well, I'm sick of writing so I'm off.

Mike

[gladius]

Does it simulate the shuffle of cards from the perspective of cards thrown in after a hand and randomly dealt or from the deck as it would be from a fresh pack each time.

Each shuffle is brand new set of random numbers, seeded randomly from some event. i.e. the previous hand has no effect on the currrent hand.

I personally don't have a problem with the deals but can see where some people feel that it is not fair to them in particular.

HW even post the probability distribution of cards. You can look and test. It is very random.

If 4 players on for different routes to a particular game played at a specific time will the deal be the same as if they met an hour later at a different table? In all probability yes, but it is likely they would not if they could be in the 2 places at the same time.

No, they would get different hands, the random number generator seeds would be different.

It is a true RNG to take all hands as possible on each hand and from what perspective a fresh deck or the randomous of the order of the cards.

As cards are dealt, a random number is selected. You are given a card out of the deck based on this random number.

I personally use the "CUT-DECK" option to change my luck of the deal and do so in real cards too by tapping the top card when I do not want to cut or take the cut otherwise.

I don't know the exact coding but the random number sequence is determine at the deal. Tapping the top cards as is, cutting the deck initiates a new seed and consequently a new set of random numbers.

If you want a more even split of cards and let skill dictate the winner then the randomous of the RNG has to change in it's formulation on the odds of getting certain cards on any given hand. The way it is done here is very simple each hand had the possibility of pulling 13 spades as the next one, is it likely 52 cards 4 suits and figure your odds.

No, absolutely not. You are randomly given cards each hand. Just like shuffling a real deck of cards. Each time the changes are shuffled the odds are reset. The previous hand in real life or simulated with RNGs has no impact on what you are going to get the next hand. For example if you are getting the first card of the hand, the odds of you getting the ace of hearts is 1 out of 52 with a standard deck of cards. When the deck is shuffle and next hand begins, the odds of you getting the ace of hearts is exactly the same, regardless of whether or not you had it the previous hand.

What some universes or odds hold within their pattern are the percentages/likelyhood that a certain combination would occur and take that into consideration in the RNG. This is from each individuals perspective.

Not really. You have a probability distribution functions (pdf) that determines the odds of getting certain cards. For cards the pdf is simple, the odds of getting any cards in the equal. There are several different ways to pick the function to deal out the cards, but in order for probability and statics to hold, the odds never changes.

Hope this helps.

[p0is0ned_fl0wer]

I shuffle and drop cards..does it simulate that?

[Dead Presidents]

I just look at it as an inability to play with not so good cards and losing...hence...calling it a one-sided deal.

We complain when we lose.
Never when we win.

[Just_Ice]

When people play with real cards, it's fair.

When people play on-line with a RNG determining card location, it's fair.

As long as there is no manipulation or pre-determining which card goes where, and each is as random as it can be in it own sphere, then both are fair.

However, to say that both methods acheive the same results is wrong. All you have to do is look at the methods for card mixing and you can see they are not the same, nor will they produce the same type of hands.

With real cards and a normal shuffle, even a series of normal shuffles, it is impossible to move the top card to the bottom of the deck. This is an extreme example, but it shows that with real cards there are limitations to how well the cards can be mixed. In fact, with real cards, many cards don't move very far from their original location, even if they change order.

On the other hand, when a RNG mixes a deck of virtual cards, it is not limited by physics as real cards are. In one RNG mixing of the cards it can move the top card to the bottom of the deck, something that's impossible to do with real cards. RNG mixing can be much more severe and much more complete.

The result is a different "feel" for how the cards are distributed. If you really look at it the RNG has a much better chance of making deals "random" because it is not restricted by the laws of physics. As I said earlier, as long as no prior manipulation is done, then both methods are perfectly fair. But, I will never say they are the same, and I will never laugh at someone who notices the difference.

These people (including myself) who notice a difference in the way the cards are dealt are seeing a REAL thing. Laughing at them only means you do not understand that difference. There most certainly is a difference. What we all have to do is understand that difference; that RNG shuffles ARE different, and that they WILL BE different until on-line games perfectly duplicate how real cards are shuffled. That most certainly can be done. I have posted how it can be done several times now.

The question is, do we duplicate real card suffles just because RNG shuffles are different? Each game site will have to weigh the value that has verse the time and effort it will take to accomplish it.

In short, the shuffles made by a RNG are perfectly fair and unbiased, but they are different than the shuffles made with real cards. Maybe they are better, and it can be argued that they are... but they are not the same as what we see when we play with real cards. I believe that is the real issue whenever RNG shuffles are brought up.