"Distribution & High Cards"

Post Reply
User avatar
Joe Andrews
Active Poster
Posts: 412
Joined: Wed Jan 14, 2004 10:48 pm
Contact:

"Distribution & High Cards"

Post by Joe Andrews » Wed Mar 19, 2008 7:52 pm

#1 PET PEEVE -

"They had the better cards; that is why we lost"

Hardwood does a good job with the RDG; still, there are those unusual hands which do occur on occasion.....

The most common distribution patterns are:

4 - 4 - 3 - 2 (two 4 card suits, a 3 card suit, and a 2 card suit)

5 - 3 - 3 - 2

5 - 4- 3 - 1

5 - 4 - 2 - 2

4 - 3 - 3 - 3

These 5 distributions represent almost 75% of the possibilities
for the shape of a given hand. The other 25% allows for six card suits, voids, two five card suits, and all of the other "exotic" distributions.

Now let's take the HCP (High card points) If we use the Bridge system, then this "point" value is assigned: - Ace = 4; King = 3; Queen = 2; Jack =1 The deck always has 40 HCP....

Now, if our resident programmer could follow these parameters, then we would see a reduction in the number of complaints about "weird" hands, and / or one side getting all of the good cards:

-----------------------------------------------------------------------------------------------
1. Always use one of the distributions for a given hand, as listed above. (We would NEVER see 6, 7, or 8 card suits, 2 voids in the same hand, etc.)

2. Using the HCP chart, always assign a player at least 8 HCP and no more than 14 HCP in a given hand. No side would never have more than 24 COMBINED HCP - thus ensuring at least 16 HCP for either team (IN THAT SCENARIO) Some hands might feature all 4 players having ten points each, or two players with 11 points each, with the opponents sharing the other 18 points each etc. etc. There are millions of possible combinations! Remember the deck has only 40 HCP!

3. No suit would have more than two of the "honor" cards; (A,K,Q,J). Thus, you would not have the top 4 cards in one suit and garbage in the other three suits.

4. The computer would never allow a player to have more than two honor cards in the spade suit.

-------------------------------------------------------------------------------------------------
Is this programmable?

It would sure make for some more interesting, and closer matches... :roll:
Image

User avatar
Dust In The Wind
Guide
Posts: 5344
Joined: Wed Feb 09, 2005 1:29 pm
Location: North Ga Mts

Re: "Distribution & High Cards"

Post by Dust In The Wind » Thu Mar 20, 2008 6:40 am

Yes you can set parameters in a program by eliminating the some of the variables, basically they are thrown out if chosen in the universe and "next" is selected and again has to meet the criteria of set patterns. This called in some programs a set "universe" which is the patterns allowed in all the patterns available. We use it in the lottery program to select variables with some omissions in the total universe of possibilities because of the problems or patterns will create for various reasons. For instance how many times a losing ticket/hand may appear in succession and on the same premise how many times in a row a winner can appear back-to-back. Basically rules of the universe rather than totally random and left up to the generated to chose the ticket/hand. It gets quite complicated when you set rules/tests (basically a subroutine)for any given random generation. IE: if a 8x2x3x0 is chosen by the generator the test would determine this is not one of the patterns allowed and discard it. Another way to do the same is to omit those patterns from the universe but it basically uses the same routine but earlier in the program. I cannot go into detail as to why some patterns are not allowed in a lottery program.

In the same example of above you can also set an overall percentages that anyone given player may see a given pattern. Where this becomes a problem in a card game site is in the players. If I have the same four players sitting at the same table, playing the same variation of the game over a period of time they each would receive every possible variable over a period of time in relation to the overall probability of each hand. Where this would go wrong is we don't play the same players and we jump from table to table and have different players in games. So player #1 has seen their 8x2x3x0, player #3 is due as is player #4 and player #2 is close because of the number of games each player has played and now player #1 experiences in the same game 2 maybe 3 very unusual hands in the same sitting. If you take into consideration the 1000's of games being played each day in spades alone the universe can be used rather quickly and if all the tables are considered in the generation then the chances increase that your table can be selected to get the "oddball" hand. Looking at one table, with the same players is one variable using the same generator and now you add the rest of the tables in ALL the rooms. Basically I could go from table A and sit a table B and see the same hand dependent upon timing and the number of hands generated overall since it is not based on the individual and their own private universe. But such is life, we might feel we deserve our own private universe but that's not the way it is.

HOWEVER, would not this be considered "STACKING THE DECK".

JUST DUST

PS - Straight up you can create a program to do just anything you can imagine... the trick is seeing all the loopholes that create their own set of problems. I think I would not be happy knowing I will never be dealt a "Straight Flush" at some point in my time playing... why even have that hand if it's not possible to get.

PS - It is possible to set a program to use percentages in any given random generation but where would you place the percentages base on... the individual player or the overall site. The only way you could tell if your percentages of any given hand like 13 spades is to look at each hand played at Hardwood and the total universe. I have never seen 13 spades, ever (even real cards) but have seen plenty of times all 13 tricks go.

PS - By the way this is not all inclusive of what I know about generation of the random or fixed random, just trying to answer the question. Which is Yes.
TO BE OR NOT TO BE..... NOW WHAT KIND OF QUESTION IS THAT??? TO BE OF COURSE!!!!!

User avatar
Joe Andrews
Active Poster
Posts: 412
Joined: Wed Jan 14, 2004 10:48 pm
Contact:

Re: "Distribution & High Cards"

Post by Joe Andrews » Thu Mar 20, 2008 7:56 am

Dust - Thanks for your excellent input on this topic.

I was exploring some ideas for more equitable distributions / hands for players. Maybe the random deals are in the only way to go. After all, that parallels what happens in real life. - Or does it? :mrgreen:

Nothing can be more frustrating than to be in a close game, with scores in the low - mid 400's for both sides, and THEN to have the computer serve up a powerhouse hand for your opponents, while you and your partner cough up 1 or 2 bids! - And out you go - frustrated and irritated. :roll:

This can be especially brutal if you are in the finals of a tournament, and get blown away with a big hand by the ops!

I suppose the answer is Duplicate (or Comparison) Spades, a test of pure skill. However, the devotees on the Standard 500 point game may not like the idea of Duplicate. After all, "we've always done it that way"! :lol:

And the beauty (is that the right word?) of fresh or random deal Spades is that, on occasion, a inferior team can defeat a superior team. Maybe that is not such a bad thing, after all.

Carry on!
Image

_S_X_Eian_
Active Poster
Posts: 75
Joined: Tue Apr 25, 2006 6:33 am

Re: "Distribution & High Cards"

Post by _S_X_Eian_ » Thu Mar 27, 2008 11:57 am

I don't think I would like to know I can never have more then a certain number of cards. I do agree it does suck when your in a close game just to be blown away in the last hand or those occasions as I am sure Joe remember from Nashville when a team gets a Nil/8 bid in the last hand to win the game when you are winning the whole game. But I do think this does add to the interest in cards. I prefer the live game and will take any hand given to me. Like with any card game it is the unknown that makes these games intriguing. I can think back to a few of those rare hand and think it is interesting when they can randomly occur. Like the time my nil was set when I had the lone 5 of hearts and my partner had the lone 4 and the opps having the 3 and 2 respectively. Yes this sucked but I still found it interesting knowing at any one time I can have any 13 cards. The time I took 13 tricks with the other 3 players going nil. (I bid 11 had 11 spades and the K,J of hearts and was first to bid) The time I was dealt 13 cards with my highest being a 6. If you take out those random rare hands and know each time you are going to get a couple of high cards then the nil might be a little less effective. I have had hands where I have had 15-20 HCP and went nil. I do see your point Joe and yes it can be frustrating for sure but I do think those frustrations are part of the game. That is what allows new players with no knowledge of strategy to play the game and have a chance to win against a seasoned pair.

A better way in a live environment I think is to either play duplicate as was suggested where everyone plays the same cards (geared more toward higher skilled players who want to remove luck) or by playing a 2/3 or 3/5 format. The main problem with this is it is time consuming but I think we could all assume that if we played several games with the same players the more skilled pair would probably win 7/10 games at the least leaving 3 for luck of the cards. 70% is close to the 2/3 (66.66%) and the 3/5 (60%) formats where the team the plays the best game of spades with the fewest errors will win the 2/3 of 3/5 battle. They have flexibility for that "game of luck". I have been victim to this many times in the late stages of live tournaments where I make it to the semi finals only to hit a rut and lose out "because of the cards".

One thing that could work is to maybe make the instances where the rare hands occur fewer and far between. If those 5 hands show up 75% of the time or account for 75% of the combinations then assuming a live deal they will show up 3 out of every 4 hands. Maybe increase the frequency of those hands still allowing the random hands to show up just not as often.

Just some thoughts from me

Todd

Post Reply