Home → Help → Experts → MetaSpades

MetaSpades

by Phlexicon

metaSpades: The foundation and underpinnings behind the game of Spades.

First and foremost we must remember that there is no international standard version of “spades”, unlike many card games that have a worldwide recognized organizing body. Spades is in its infancy relative to the much older and established card games. The rules are as varied (actually, more varied) as there are spades playing sites. Some sites actually turn the rules over to the players to set for each game they host. Bobby Fisher became so bored with “standard chess” that he invented “fisher chess” and for the most part only played that after retiring and refusing to defend his world championship title. I’m mentioning this to open your mind to change (Obama liked this idea). We had a game that had been around for millennia and while it had gone through many variations (rule changes) and had been stable for hundreds of years, the person who was undoubtedly the best at that variation sought change to make it an even better game!

Secondly, we mustn’t forget that spades is a partnership game.

I want to divide these comments into 2 separate categories:

1. Changes that are relevant to the standard game, and
   
2. Changes that are more relevant to RepSpades™.

The Standard Game (Rubber Spades)

The 0 bid

The 0 bid is fundamentally no different than all the other even bids (2,4,6,8,10,12). We look at our hand and typically we say, “what do I think this hand is worth?”, all things being equal. Then we look at the game score. We look at the bids made prior to ours. We consider all the external factors including: who is our partner, who are our opponents, how hungry am I, do I need to pee, etc.. And then we choose what we think is the “best” bid for all the things that we considered. Please note that this is the “editorial we”, I am in no way implying that the 2 partners are conferring on what to bid. The first few years I played spades, the 0 bid was allowed. Only when I started playing at the zone was I rudely awakened to the fact that the 0 bid was no longer an option (how strange I thought). If I deem that my hand is worth a 2 bid I am allowed to bid it. I have the option of also bidding anything else I choose with it. But for some strange reason I’m not given this option with the 0 bid. I then considered what the game would be like if we extended that logic to all even bids! Me and my friends played games where no even bids (0,2,4..) were allowed. If you have a 4 bid you must choose an odd bid (typically 3 or 5). It seems a bit odd but this requires skill. I recommend to any of you who haven’t tried it to give it a go. I think you will find it entertaining and educational.

My point is this: Not all skills are equal! It takes skill to manage the double nil. It takes skill to play a good game of Mirrors or Suicide or Crawl etc. I believe people play them for the variation they provide, not for the skill level they require. I’ve played them all a few times just for the “fun” of it. To me, not allowing a 0 bid is fundamentally no different than not having a 4 bid. But I can assure you that not allowing any even numbered bids is even more fun and more challenging. And it requires skill. We call this variant “oddSpades” by the way.

I want to remind you that one of the critical modifications to spades was the introduction of the paired format. Originally it was a “cut-throat” game. Every woman for herself. The traditionalists argued that the game would be destroyed and its nature fundamentally changed. But all that happened was that it became better. Regular partnerships developed and the game became more social.

Bags

The concept of a bag penalty (bagging out) was added because the penalty for being set was too drastic. People were so concerned about being set that the bidding became very conservative and the game became rather dull. I personally believe a better solution to that particular problem would have been to make the penalty for a set to be that you get 0 points rather than -10n (where n represents the pair bid total). To get 0 for a 5 bid that is set makes the penalty only 1/2 of the current penalty of -50 (50 points loss vs 100). We should not forget the hidden penalty of -9 for every trick we don’t bid and take, and get 1 instead of 10 points! The idea of having a second penalty of -99 for every 10th bag taken was a bit extreme in my opinion and somewhat arbitrary.

One more critical idea before we move to the next topic. Since the extra (beyond the -9) penalty for bags was added to encourage more aggressive bidding (hopefully more 11 and 12 bids versus 9 and 10 bids), why should we penalize pairs when they set their opponents? Fundamentally spades is a battle to bid and make tricks and balance that against the risk of being set. Setting is one of the most important aspects of the game. To penalize a team (giving them bags) when they achieve a good outcome seems a bit strange to me. It’s an 11 bid and you choose to go for the set and you fail…penalty is that you get 2 bags. I have no problem with that. But when you succeed and take 3 extra tricks to set your opponents, it makes no sense to me that you should be punished for setting them!

And this leads us into our next topic.

Boundaries and Values

In spades we have several boundary conditions. One is the finish lines were victory is obtained (typically 500 and -200). These are practical values used to obtain a game of reasonable length. But never-the-less they are arbitrary to some degree. Maybe 400 or 800 would be better. We shouldn’t be so dogmatic as to assume that these traditional values are carved in stone. Maybe getting -49 for every 5th bag would be better. Why did they choose 10 bags as the jump-off point? Why did they choose -99? Why not 8 bags and -64, or 10 bags and -49, or 7 bags and -111? I doubt if these particular values were well thought out. I’m not here to defend nor condemn them. I am merely questioning them. And I am inviting you to open your mind to new ideas. The boundary values we have for nils is another interesting topic. And this leads me to …

Nils

Nils were added to the standard (non-duplicated cards) game to help compensate for the uneven distribution of spades and high cards in a relative short game to 500. While their original intent was an attempt to make the game more competitive, it had a very interesting side effect. It created some new and exciting tactical components to the game! But. And this is a very big but (sp). Is the value of a nil (+100 or -100) the best value for the game? This is an interesting and important question. More on this later. Another key consideration is: How do we view the bid of a nil? To me it’s quite simple. I see it as a side bet that stands alone and doesn’t conflict with the basic game. I like what the nil has added to the old game but I also realize that it is way too over valued. Viewing the nil as a side bet makes it easier for me to see how to integrate it into the standard game. Since spades, in its current form is a partnership game, I think it is unwise to destroy this partnership aspect by making the nil a stand alone entity. To me the nil bid is a 0 bid with a side bet. If I bid nil and make it, my side gets a 100 (or whatever it’s worth) extra points. If we fail to make it we get -100 points. Now here is where I see a difference, and spades is played/scored this way at many sites. If I bid nil and my partner bids b then we are still a team and if I choose to sacrifice my nil to help us make our b bid, then I may do so. Also if my nil is set unintentionally from my perspective I still can help my partner make our bid and also try to set the opponents if we so choose. This now makes it quite easy to answer the question, “how do we count the trick that a nil takes?”. If my partner bids 3 and I nil and we both take 2 tricks, then we score +31 (1 bag) and -100 for the side bet of nil for a net score of -69. Currently here at GP the score would be -30-100+1 for -129 and 2 bags. I consider this a bit draconian. To have such a drastic penalty is overkill in my opinion. Spades is a partnership game. I see no reason to diminish that very important aspect! I suspect it is done this way here at GP merely because it was done that way at the zone. Not a very convincing argument to me.

Many things were done improperly at the zone and now is as good a time as any to fix them. Or maybe a better way to put it is: to follow the lead of other sites who may have found a better way to do it. Many here seem to think their particular knowledge an interpretation of the game is the only one. They are quite mistaken. Spades is in its infancy and needs to be nurtured to reach maturity.

RepSpades™

Scoring or weighting of individual hands

In repSpades we have a sxpades variant that is an attempt to come up with a more skilled version of the game in much the same way that Bridge expanded social or Rubber Bridge by introducing Duplicate Bridge (or just duplicate as it has come to be called). In the 1st variant of duplicate bridge there were essentially two tournaments going on concurrently and the scoring was radically different from the social game. Each hand was scored separately from all the others and a win on any hand was equal in value to a win on any other hand. Therefore a 1 point mistake, a 30 point mistake and a 450 point mistake were all equal. This had a tremendous impact on the bidding and playing strategies. And this was a very different game than the original. This same approach was tried with spades and failed. A long time ago someone in the bridge world got the clever idea to introduce a duplicate version that retained many of the game aspects of the original. The cards were duplicated as in the version called Duplicate but a new scoring scheme (International Match Points IMPs) was introduced. In this variation all hands were not equal and many of the things that players liked about a more normal game format were brought back in. This format was (and still is) called Team of 4.

Here at GP we are trying to come up with an approach not unlike the Team of 4 (T4) concept that the bridge world conceived. We are trying not to make the same mistake that eSpades and other such duplicate spades sites made. Those of you that think this is the case, simply just don’t get it. We haven’t gone so far as to introduce IMPs, but that is definitely worthy of consideration. To ignore the hard work and mistakes overcome by decades of thought put into these types of formats by the bridge world would be a gross mistake. When others have broken the trail and done the heavy lifting, it’s foolish to ignore what they have discovered. The Bridge community is a sister nation not an enemy state. Will they learn from us? Time will tell. Should we learn from them? Without a doubt!

T4 is a form of duplicate and a form of Bridge. Is it different? Of course. Is that good? Absolutely! There is neither an attempt nor a desire to create a “bridge version of spades”. The desire is to create an equally challenging analog! To take the basic game of spades and transform it into a format that uses duplicated cards in a way that primarily minimizes luck and secondarily retains as much of the original game as possible. That is exactly what T4 did for Bridge. Hopefully we can do a similar thing here at GP for spades. Nothing more, nothing less.

Boundaries and values

There is one common element in all forms of duplicate games: The equalization of all approaches to the game thereby giving an equal weighting to all. What this implies is that we now have to have boundary conditions that make that possible. So far, the only solution that I know of consists of an equal number of hands being played by all competitors. This is the solution arrived at, for both standard forms of duplicate bridge and a third played at the collegiate level.
By playing a fixed number of hands, agreed upon before the beginning of a match, we still have a boundary and associated tactics involved with that boundary. They will be different than those in the standard game. Closely associated with the number of hands boundary are the boundaries and values associated with Bags and Nils.

Bags

To encourage more aggressive bidding and to give the game a push-pull aspect, an extra penalty (in addition to the obvious 9) for making more than what your pair bid was added. There are 2 components of this penalty. The first is the frequency (f) at which the penalty occurs. The second is the magnitude (m) of the penalty (the depth of the fall from grace). In the Standard game currently implemented at GP, f = 10 and m = 100. If we let b represent the bags taken we have the following general form of the bag penalty function:

P = b - m * trunc(b/f)

Therefore for GP the P = b - 100 * trunc(b/10)

This thread is expanded in detail below in Appendix A. To not distract from the flow of my argument I will not include it here.

I see no compelling reason for m = 10*f or for m = f^2. What I’m trying to demonstrate here is that there are many ways to bag a cat. I personally think f=1, 2 or 3 makes a lot more sense for repspades and I personally lean towards f = 1. Keeping the m = 10 that we currently have, keeps the flavor of the standard game. I think the key is to keep them the same. If we ever changed m to 8 (for example) for the standard game then I would recommend the same for repspades.

Nils

In addition to my earlier comments above on the nil, we now need to consider the value of that side bet. First should we allow the host to pick it? This might be a reasonable initial choice to facilitate experimentation. Or should we start at some specified upper bound, say 200, and work our way down by 25s? We could try a week at 200, then 175, all the way down to 25. Or maybe we could start with 100 and work down by 10s (100, 90, 80, 70, 60, 50). If we record all games by hand we can then compare how the value of the nil affects the outcome of our games. I’ve already done this on over 80 repSpades games that I personally have played myself.

General considerations

All these parameters need to be considered as a group. They are very much correlated. To consider their value out of context is very dangerous.

h := the number of hands played per game  
f := the frequency of the bag out  
m := the magnitude of the penalty for the bagout  
n := the value of the nil side bet

Just as the very nature of fairness dictates that the number of hands must be equal at all tables, other practical aspects of the nature of tourneys require limits on our parameters to keep things practical. The value of h might vary from 4 to 16 reasonably, depending on the number of rounds necessary to accommodate the number of teams (n) that are entered. Rounds = log(n) base 2 for a single elimination tournament. For example if we have 64 teams then R = 6. If we decide that 2 hours is a reasonable length of time for a tourney, then we have ~ 120/6 = 20 minutes per round. Allowing 6 minutes to get the scores, determine the winners and reseed for the next round, leaving only 14 minutes to play the match. For this extreme example, h will likely need to be 4. If we now examine the opposite extreme (2 teams), then we have: R =1, 120/1 = 120 and h could be as large as 30!

To allow for this rather large range of the value of h (the number of hands per match) we must now ask ourselves, “how does this effect the other 3 parameters in our group?”. The 1st thing I notice is that if f is any value other than 1 then we have to adjust m and n as we vary h. This is not acceptable due to the complexities that this introduces. Therefore I think to allow for variable values of h we have to set f = 1 to simplify the format.

By examining approx.80 real games I’ve found that the average value per bag is approx. -5. Therefore the range of m that we might want to initially experiment with is 6 to 10 (for f = 1). We obviously need to gain some experience with these values (m = 6, 7, 8, 9 and 10).

The value of n has been already dealt with above.

Specific considerations

Ideas like the 0 bid seem like no-brainers to me. To give us more options merely increases our freedom. If we have the option we then can call a game with no DN and/or no 0 bid or no even bids or whatever we prefer. Sometimes we call a game with no nils. By including the option, we then have the choice whether to use it or not.

Any changes made by GP that give us more options, allowing us to experiment and try to find changes, give the game the best opportunity to evolve. With this in mind I encourage GP to allow the host to set the game boundaries. If h, f, m & n are data driven parameters that allow the host to define the game at the table then the maximum freedom is obtained. I believe these options might be nice for the standard game but are a necessity for repSpades.

Conclusions

To create a fair game with duplicated cards at both tables, it’s essentially a given that the number of hands played at each table must be the same. With this foundation we need to adjust the games parameters such that bagging, setting, nilling and playing are properly balanced to maintain as much flavor of the standard game as possible. To use parameters that skew the balance of the game to favor one aspect over the other is to detract from the symmetries that make “spades” such a joy to play. To arrive at those parameters that give us the optimum balance will take more skill than it takes to play the game. It will take more patience than any one of us has. It will take cooperation and experimentation. But most of all it will take an open mind. We must not assume that due to our limited experience we can assume the particular version of spades that we know best, is “spades”. The game of spades is not well defined and is still growing. We must not be static in our thinking.

Appendix A

P(m,f,b) = b - m * trunc(b/f)

Let’s run some numbers to get a feel for this function.

m = 100, f = 10 b = 1,2,3... 20 P = 1,2, ..,8,9,-90,-89,-88,....,-81,-180,... this is GP (m=10f)  
m = 50, f = 5 b = 1,2,3... 20 P = 1,2,3,4,-45,...,-41,-90,-89....,-180... m = 10f  
m = 10, f = 1 b = 1,2,3... 20 P = -9,-18,-27,-36,-45,..-90,.......,-180... m = 10f

m = 50, f = 10 b = 1,2,3... 20 P = 1,2,3...,8,9,-40,-39,-38,...,-31,-80,... m = 5f  
m = 100, f = 5 b = 1,2,3... 20 P = 1,2,..4,-95,-94...,-91,-190,..., -380,... m = 20f  
m = 6, f = 1 b = 1,2,3... 20 P = -5,-10,-15,.......................,-100... m = 6f  
m = 25, f =5 b = 1,2,3... 20 P = 1,2,..,-20,........................, -80... m = f^2

Notice that examples 1,2,4,5, and 7 all start out 1,2,3..

At bag #5 the respective values are 5, -45, -45, 5, -95, -25, and -20. At bag 10 the respective values are -90, -90, -90,-40,-190, -50, and -40. At bag 20 the respective values are -180,-180,-180,-80,-380, -100, and -80.

Let’s now take a closer look and some other variations:

P( 6,1,b) => -5, -10, -15, -20, -25, -30..  
P(12,2,b) => 1, -10, -9, -20, -19, -30..  
P(18,3,b) => 1, 2, -15 ,-14, -13, -30..

P(**-4**,1,b) => 5, 10, 15, 20, 25, 30.. here bags are worth 1/2 of a bid trick (**interesting**)

P(**-14**,1,b) => 15, 30, 45, 60, 75, 90, 105...here bags are worth 3/2 that of a bid trick (**very interesting**)