CardShark SpadesFord SoftwareContents:
Searching This File Rules Of Spades Rules Not Used In CardShark Spades Mechanics of the Game Options
AutoPlay Cash Out Winners New deal Replay deal Enter external deal Break spades before leading * Allow ruffing on first trick * Nil bidding allowed * Program Playing Styles * Scaled Scoring * Nil score Change game goal Show Scores Play a Tournament * Concede game
View Explanations View Suit Counts View/Email Deal Record or Screen Shot View Game stats * Hint Button CSS Cheated! The Program's Strategy
How the Program Plays No Cheating Ganging Up Defending Nil Bids When NOT To Ruff Searching This FileTo search this document, use the Search input box in the lower-right corner. The program will search for an exact match of the text entered (although the text match can be part of some other word, such as percent and percentage), so be as precise as possible and if no match is found, try searching for just one main word. To repeat the search, press the [&] button to the right of the search input box. IntroductionSpades is an exciting, challenging game which is similar to Bridge, but not nearly as complicated: Spades is always the trump suit and each player gets only one chance to bid. While Hearts and Spades are both fun, popular card games, Spades has a substantially smaller luck factor than does Hearts, which good players will appreciate. (But still, look for CardShark Hearts in 2012.) Both CardShark Hearts and Spades were originally written for Windows (around 1990) and distributed as shareware. CSSpades is being ported over to Android and improvements made. CardShark Hearts and Spades for Windows can still be found on some shareware download sites, but they may not work under Windows 7 or later and possibly not Vista either. (Only costs a download to find out.) You will still need to search for and download VBRUN300.DLL to get them to run. CSS plays four-handed cutthroat only, not partners. Some computer Spades game offer both options, but usually they play the same way for either version of Spades when, in fact, they are two completely different games requiring code to be written specifically for each version. If a Spades game plays partners but offers no options for choosing bidding and carding agreements, then it cannot be an expert-level game. Cutthroat was chosen for CSS because it is a better test of individual play. If you are an experienced partners player and have not played much cutthroat, keep in mind that the strategy for bidding and playing is completely different. For example, in partners you would generally not lead out your Aces right away, while in cutthroat, that is the best strategy. See The Program's Strategy. If there is enough demand for a partners version of CSS, we will release one, but it will take some time to do a complete rewrite of the cutthroat code as well as adding code for setting up and using the bidding and carding agreements. Obviously, this would be no small task, so we would need to hear from a lot of people. The Free VersionPorting CSSpades over from the Windows version has so far taken 8 months and over 10,000 lines of code just to get the basic interface and game play done. We started with the Expert style of play and will modify that to come up with the other styles. The features in the Contents section above which are NOT marked with an asterisk ( * ) are active in the Free Version; the rest will be in the Final Version. Our goal is to have the full game completed in time for Christmas of 2011. It is expected to sell for $4.99.The bad news is that it is very likely that you can NOT beat the free CSS. Even if you are just as good as CSS, then statistically you would only win once every 4 times, so even a top expert will lose to CSS 75% of the time. The good news is that there are several features, mainly the View options seen in the Contents section, which will help you learn to play better: You can turn on View All Hands, Explanations, Suit Counts and AutoPlay and either watch CSS play all four hands while providing explanations of non-obvious bids and plays or you play the South hand while seeing any or all of these options. While it may not make sense to play while viewing the other hands, the fact is that a novice will often not make the correct play even when seeing the other hands, so the Hint and View All options let you see if a play didn't work because you couldn't guess the other players' cards or if something else is involved (a nice way of saying you didn't understand what happened). The complete edition will have 7 playing styles, many of which should be beatable by the average player. The final version will also have the complete section on The Program's Strategy.
DefinitionsA few terms in Spades, borrowed mainly from Bridge, may not be familiar to all players. We were trying to avoid the bridge hand designations North, South, East, West in case they seemed strange to non-Bridge players, but using 1,2,3,4 was just too confusing.
Rules Of SpadesThere are many ways to play Spades. CardShark Spades uses the rules and playing options which minimize luck factors and maximize the test of the cardplayer's skill. Spades can be played with two pairs of partners or "cutthroat", where each player is on their own. The cutthroat variety could be played with 3-6 players, but four players is a standard number. Just as with partners Spades, 3-handed cutthroat requires programming code for bidding and playing which is completely different from 4-handed cutthroat. The preferable form for a computer Spades game is Cutthroat, since in Partners, your results are not based on your own skill versus CSS, but on your skill aided or detracted from (depending on the quality of the program) by a computer partner. In addition, playing partners only makes sense when the partners have an understanding of the meaning of certain types of bids and plays, which doesn't happen when playing with a random computer partner. CS-Spades plays four-handed cutthroat. You play one hand and CSS plays the other three hands individually. "Individually" means that CSS does not "cheat" -- it does not look into other hands or play any differently against the human opponent than it does against the other computer hands. Each player gets 13 cards. The order of bidding is clockwise and the turn to bid and play first also rotates clockwise. For example, player #1 starts the game by bidding and then leading the first card. On the next deal, player #2 will start the bidding and will lead the first card. Some people play that the person to the left of the dealer leads first, but this difference has no effect on the game. After a player leads out a card and the other players must play a card of the same suit or if a player has none of the suit, the player has the choice of trumping ("ruffing") with a spade or discarding from some other suit. If a trump is played on a trick it always wins the trick; otherwise, the high card played in the suit led wins the trick. If more than one trump is played on a trick, the high trump wins. The winner of the trick leads for the next trick. The default option is that spades cannot be led until they have been "broken" (i.e.: until someone has trumped a trick). Other than that, any suit or card can be led. The goal in Spades is to bid as many tricks as you think you can take. At the end of the hand, you get 10 points for every trick bid and made and only 1 point for every trick that you made but didn't bid. For example, if you bid 3 and take 5 tricks, you will get 3 x 10 + 2 over-tricks, for a total of 32 points. If you bid 5 and take 5, you get 50 points. If you do NOT take the number of tricks you bid, you lose 10 points for each trick bid. For example, if you bid 5 but take fewer than 5 tricks, you lose 50 points, whether you take 4, 3, 2, 1, or none. As an option, you can bid to take no tricks, called bidding Nil, and if you succeed in taking none, you get 100 points. If you fail, you lose 100 points. Keep in mind that if you DO bid Nil, the other players are going to play their lowest cards to try to FORCE you to win a trick. Unlike Bridge, each person gets only one turn to bid. The first player to reach a specified number of points wins the game. The default option is that the game goal is 500 points, but you can change it to whatever you want. The higher the number, the more opportunity there is to overcome bad luck. When the player whose turn it is to lead has all winning cards left, the default option is that the program will "cash out" and give that player the remaining tricks without having to play through them. This is highly preferable to getting to a point in a deal where the leader obviously cannot lose another trick, but you have to sit there playing one card after another to the end.
Rules Not UsedA large number of rules options are available for Spades. (See Wikipedia.) Spades is similar to Bridge in many ways, even when played without a partner. The biggest difference is in bidding, where Bridge is infinitely more complicated and requires a partner. Existing computer Bridge games do not allow anywhere near the variety of bidding systems and agreements which live partners may have, and because computer players are not as good as human players, they do not bid as well, thus computer bridge games are rarely satisfying for advanced Bridge players. In contrast, because of the simpler bidding system in Spades, playing cutthroat Spades (without a partner) is not only very possible, but it is much more fun for good players because you don't have weak computer players dragging you down. Another rules option in Spades is to use jokers. Using jokers doesn't really seem to make the game worse, but it would be hard to argue any way in which using Jokers makes the game of Spades better, so what's the point? Oddly enough, most of the Spades games for Android force you to play partners, and one of the couple that doesn't, actually forces you to play with jokers. One common rules option is called "Bags" or "Bagging", which is an anti-sandbagging rule. This will be discussed in depth below. Another option is to give a 100-point bonus to anyone bidding and making 7 or more tricks. This is an even worse variation of allowing Nil because the effect is even larger. The problem with this option is that hands which can make 7+ tricks are rare, and getting one is nothing but the luck of the deal. The rarity of it means that if someone scores 170+ points because of one lucky deal, it's unlikely that such luck will even out in the course of a game, meaning that the lucky player will most likely win the game based almost solely on the luck of one deal. It's bad enough for the equality of luck in the game if someone gets a 70-point or higher hand; adding 100 points to it just makes it worse. Every change in the rules of the game makes programming the source code of the game much more complex. A program cannot use the same code to play with and without jokers, or with or without partners, or with or without anti-sandbagging rules, or Nil, so essentially, different source code must be written for each of these alternatives. If a game is already not very good at playing the basic game, adding a lot of complexity to the code is just going to make the game play worse.
Some people and computer games play a rule ("Bags") which penalize sandbagging, which is when you purposely underbid to keep from going set. Typically, the penalty is -10 points for every trick taken and not bid, though the penalty is usually applied only after 10 overtricks have been accumulated over multiple deals for a one-time hit of minus 100 points. CardShark Spades does not offer this option for several reasons. The main reason is that the vast majority of underbidding comes not from intentional sandbagging but from the fact that on almost every deal when everyone bids correctly, one or more players will often end up taking tricks which no rational player could have counted on. In the deal below, south bid 3 and took 6. With only 2 Spades and no short suits to ruff, he somehow took a Spade trick. With the north hand holding only 2 Hearts and presumably ruffing the 3rd, south managed to get 3 Heart tricks. With Ten high in Clubs, he got a Club. Now this could be expected in a Bags game where people purposely throw away tricks, but in CSS, players try to take as many tricks as possible, bid or not, so it's just the way things play out. For example, on south's
Another reason not to use Bags is that the rule is ineffective. If a good player sees that he is going to make his bid or has already made it, the player can usually very easily get rid of other winning cards by discarding them. A decent player against any of the current Bags-playing computer games should virtually never have to take a Bag penalty. In addition, if someone consistently underbids, they are very unlikely to win the game against people who bid what they think they can make. If someone underbids when they are way ahead, then the sandbagging penalty is usually too little, too late. Instead, CSs offers an option to use Scaled Scoring to keep players with a big lead from underbidding. CSS also offers Duplicate play for competing against CSS playing the same cards. Another problem is that Bags completely changes the nature of the game. Even if you bid correctly, someone else may purposely underbid and then try to stick you with extra tricks which you did not bid. They can do this by throwing away what would normally be winning cards, such as playing a King when an Ace is led or trumping high when a low trump would win. To defend against this, you also have to change your bidding and play, in which case, you are not playing real Spades anymore; you are playing a game similar to Spades. Also, with this rule in effect and some other players pitching high cards, you cannot afford to try to set other players very often because it may cause you to take too many extra tricks and reduce your score. If you really want to play by this rule, there is a Spades games for Android which actually forces the use of the rule. However, the game's strategy does not handle the additional complexities well which, along with its tendency to frequently bid Nil with poor Nil holdings, makes it too easy to beat. Spades is already a good, interesting, challenging game. This rule, like all the other optional rules, doesn't make it a better game, just a different game.
Mechanics Of The GameCSS starts the first time by randomly selecting a dealer and dealing a random distribution of cards to the 3 computer players and you. CSS saves your selected options, scores, current deal number and dealer; the next time you start CSS, it will read in that data and proceed from there.When you start the game, the Options will be displayed so that you can changes options selected if you wish. See the Options section, below, for details. The dealer bids first and each player then bids in turn one time. Press Okay to advance to the next player. When it is your turn to bid, select the cards with which you expect to win tricks. This is just to aid you in bidding; all the program cares about is the total number of cards selected. For example, with KQ2 of Spades, you could play either the King or Queen to force out the Ace, then win a trick with the remaining face card; it doesn't matter if you select the K or Q or even the 2 as the card to bid. After you select the total number of cards you wish to bid, press Okay. When it is your turn to play, select a card by touching it. Whether bidding or playing, a card can be returned to the hand by touching it again. When satisfied with a selection, Press the Okay button. Since you have to first choose cards and then follow that by pressing Okay, you should rarely play a card by accident. However, if you do, just click Options and restart the deal. Don't forget to check off that you want the scores added. Press Hint to see what CSS would bid or play. Press Okay to accept CSS's card(s) or touch the card(s) to return them to your hand and select different cards. In normal playing mode, after each computer opponent plays, his card appears and you can press Enter to move to the next player or wait a fraction of a second and the others will play automatically. At the end of each trick (round of four cards played) and the end of each deal, the scoreboard in the center will be updated and a scoring detail for each player will show their bids by suit and the number of tricks taken by suit. When through looking at the scores, press the Okay button. In Options, you can choose to show scores all the time or only at the end of tricks. When it is your turn to play, the cards which you are NOT allowed to play are grayed out. For example, if Clubs are led and you have Clubs, all the other cards will be grayed out. If you have only one Club, it will be brought out automatically, but you still have to press Okay. At the end of each trick, press Okay to start the next trick. At the end of each deal, each player's hand and tricks bid and taken by suit are shown so that you can see how the players did. The scoreboard is updated for the results of the deal. Press Okay to start the next deal. Press Options and Replay if you want to replay the deal. Press Options, External, Okay and Rotate Hands if you want to play somebody else's hand.
Scores, program options, and deal and dealer number are saved to disk anytime you leave the program, including powering down your device, and are restored the next time you restart the program. If you are gone for a short time, such as pressing Home to run another app for a short while, CSS will normally resume at the same point when you return. If you put your device into a Sleep mode, CSS will also resume where you left it. If you have left it in Sleep long enough that you don't remember what has happened in the deal, you can press Options and start the deal over. If you press Back instead of Home, or if you press Home and leave CSS in the background long enough, Android will remove CSS from memory and when you restart CSS, it will start the deal over.
OptionsAll the CSS options can be reached by clicking the Options button at the top of the screen. While the options window is open, the Help button moves down below it so that options can be looked up in this file.![]() Auto-play ~~~ Cash out winners ~~~ New deal ~~~ Replay deal ~~~ External deal Break spades ~~~ Allow ruffing 1st trick ~~~ Nil bidding ~~~ Program playing styles Scaled scoring ~~~ Nil score ~~~ Game goal ~~~ Show scores ~~~ Play tourn. ~~~ Concede game VIEW: All hands ~~~ Explanations ~~~ Suit counts ~~~ Deal record ~~~ Game stats Dupe Mode *Few people (if any) will be able to play as well as the Expert in CSS. But even if you are one of those few who plays exactly as well, your odds of winning a game are just 1 in 4, or 25% of the time. Even if you are a little better than CSS, you may win 30% or even 40% of the time, but that's still a lot of lost games to have to swallow if you are using to beating other games 90% of time or more. If only you could play CSS heads-up, your winning percentage would be higher -- at least 50% if you are equally good. Well, in the Dupe Mode, you can play heads-up, in a way. Dupe Mode (Duplicate Play) lets you compare your score to the score CSS gets when it plays the same cards. In the Duplicate Play mode, after you play each deal, CSS will replay the same deal with CSS playing your cards. This essentially removes all the luck from the game because you are competing against the same opponents and playing the same cards. The only difference is your skill versus CSS's. (In fact, you can even compete heads-up in the Duplicate mode even when using an external deal. After CSS is through playing the same deal you just played, your "Dupe Game" score will be increased or decreased by the difference in your score and CSS's score on the same hand. The scores are divided by 10 to keep the numbers more manageable. This means that if you and CSS make the same bid, but one of you takes an extra, un-bid trick, you will be tied after dividing your final scores by 10. Since a significant part of the game is not just making the most points with your cards, but setting other players (especially those who are ahead of you), CSS awards bonus points to the Duplicate score for setting others. You (or CSS) get 1 point for each 10 points by which the score of the player you set exceeds your score. For example, if the player set has 400 points and you have 300, you get 10 points on the Over/Under Dupe line. If the player set has FEWER points than you do, then you get 1 point on the Dupe line. At the end of the regular game, the "Dupe Total" score is increased by one if you have outscored the Shark in duplicate play or decreased by one if the Shark outscored you for the game. If you are using Scaled Scoring, the Dupe Total score is increased or decreased by the difference in the Dupe - Game points, not by just 1. If you are interested in learning from CSS's play of the same cards you just played, try to figure out what CSS will play before clicking Okay for each card. If instead you just want to generate CSS's score as quickly as possible to compare to your score, click on the Zip button. If you are curious about how CSS made its score, you could then click on Replay Deal before the scores are added in to see CSS play at a slower speed, or you could click on Deal Record (on the Options panel) to compare plays.
AutoPlayCSS plays your hand too. This is a good way to study the strategy. When Duplicate Play is chosen, AutoPlay is automatically toggled on for CSS's turn.
Cash Out WinnersToggles the Automatic Cash-Out option on and off. When checked, the option is on and the following takes place: At the end of each Trick, the program checks to see if the person with the lead has all winning cards left. If so, the program stops play, displays the remaining cards, and awards the remaining tricks to the person with the lead. After all, if the person with the lead doesn't have a card left that is lower than those held by the other hands and no trumps are out against him, there is no point in playing out the deal. One exception is that if only one trick is left, the program will not cash in, since it is faster just to play the last trick.
New dealAborts the current deal, zeros out the deal scores (but keeps the game scores), and prompts you for a deal number. The default number will be the next number in the current series of deal numbers and you can just click Okay; however, you can enter any number you wish. You can also change the dealer if you wish.Normally, the Add Scores checkbox will be checked. Uncheck it if you don't want the deal's scores added for some reason.
Replay dealEach deal has a deal number. This does not mean that there is a set of pre-existing, numbered deals. The deal number is used to "seed" the software's random number generator so that it will generate the same set of hands for the same deal/seed number. This allows you to replay a deal by entering its deal number without its having to be saved to disk. When you click Replay deal, CSS will display display the current deal number and dealer; however, you are not limited to replaying that deal. Enter any deal number and it will be recreated. When you replay a deal, you normally do NOT want its scores added to your regular game scores. If you do, just check the box which says to add the scores to the totals. Note that if you choose to add scores when replaying a deal, CSS will NOT remove the original scores. When you start to replay a deal, you can scramble the opponents' cards, as explained next. If you scramble the cards, play through the deal, and click Replay deal again, you will be playing the scrambled cards again. To get back the original card holdings for the deal number, press New deal and enter the deal number and dealer. Scramble CardsWhen a deal is set up to replay, you get the option of scrambling the opponents' cards. This lets you try your strategy against different distributions of the cards in the opponents' hands. For example, say that you bid 5 and don't make it. Was it a bad idea, or was it just a bad break in the distribution of the cards? Try replaying your same hand a couple of times with different distributions of the remaining cards in your opponents' hands. If you NEVER make it, then you can conclude that it was a bad idea. If you make it more often then not, then it was a good idea, but bad luck. Another alternative is to turn on Auto-play and see how CSS would have bid and played your hand. You can scramble the cards as many times as you wish before playing a deal. Press Okay when ready to play. To give specific cards to players, see External Deal below. Also see Run simulations in the next section.
Run simulationsIf you really, really want to see how many tricks a hand should take given different distributions of opponents cards, use the Run Sims features. If the hand you want to test is South's, just click Options, Replay, and Okay to get the Run Sims button. If the hand belongs to one of the other players, rotate the hands (see previous section) until South has it, then click Options, Replay, Okay. Scramble the cards if you want to before pressing Run Sims When Replay is clicked, the Run Sims button will appear; click it. To pause the Sim run at any time, press ZIP. To discontinue the Sim, press Stop Sims. When through with the simulation, press Options and New Deal. The number of tricks won by the hand are shown across the top half of the screen. The number of tricks won by each suit by the South hand are shown in the boxes around the screen normally used to show score details.
The simulator will stop after 100 deals. This is usually a large enough sample size to give meaningful results and it has the advantage that the total for each number of tricks taken is also its percentage. For example, if 4 tricks were taken on 54 of the deals, then 54% of the time, at least 4 tricks will be taken. Here is an example of a simulation run:
The results will be shown as follows:
From the above, we see that the hand will, on average, take 3 or more tricks 99% of the time (the other 1%, it will take 2). We could say that the hand's value is therefore 2.99, but that is so close to 3 that we round it up and see what it take to make 4+. Adding the numbers for 4-7 tricks, we get 92%. So we could say that the hand is worth 3.9. An aggressive player could take that one step further, rounding 3.92 to a base of 4, then adding .43+.25+.03 = .71, making the hand worth 4.7. Now lets see how that compares to the tricks taken by suit: The Spades box will show 1: 10, 2: 49, 3: 39, 4: 2, which could be equated to 1.9 or aggressively, 2,4. Hearts: 0: 1, 1: 99. A good solid 1. Clubs: 0: 36, 1: 64. Equates to .6. Diamonds: 0:26, 1: 52, 2: 21, 3: 1 is .74. Since the suit will take no tricks 26% of the time, it would probably be way too aggressive to round that up to get 1.24 for the next level. So 1.9 + 1 + .6 + .7 = 4.2 or midway between the 3.9 for the hand valued normally or 4.7 valued aggressively, which seems like a good compromise. The bottom line is that we sum suits to make a bid, so they are the most important valuation. The hand value just lets you see that the summed average suit values are reasonable.
Enter external dealEnter a deal from an external source or reuse a deal and change cards in one or more hands or rotate existing hands. Before you change cards, you are asked if you want modify the current hands. Modify any/all of the current hands: All hands in the current deal are shown. Touch a card in any hand to return it to the deck. To move cards from the deck to a specific player, pick the player, then touch the card(s) to move to that player. When done, press Okay. If you just want to play a hand dealt to another player without changing any cards, after saying Yes to modify all the current hands, press the Rotate Hands button until the desired hand moves to the South position, then press Okay. Modify just the South hand: Only South will retain his cards from the current deal and the rest are returned to the deck. Touch a card in South's hand to return it to the deck. Making sure "S" is selected, touch any card in the deck to put it in South's hand. When done with South, press Deal all to give the remaining cards to the other players, after which their hands are shuffled. Note that it doesn't matter much which cards are given to the other three players in a simulation because their cards will be scrambled after every deal anyway. Keep none of the hands: All cards are returned to the deck. Touch cards to give to the selected player. If any cards are left after entering the desired cards to deal, press Deal all to deal the rest. If you are entering a complete external deal, after entering 3 hands, press Deal all to give all the remaining cards to the 4th player. * (This option is not available in the Free version.) To play CSS against another computer game or even against three live opponents, select the cards dealt to CSS and press Okay, Options, Replay, and enter the Dealer. As the game progresses, you will be prompted for each player's bids and cards played. At the end of the deal, CSS will have a complete deal record which you can view if you wish, or you can have CSS bid and play the hands to compare it to how the other opponents bid and played. To change the dealer, after distributing the cards, press Options and Replay. Leave the deal number unchanged, but enter the desired dealer. You can either play against the West-North-East hands or let CSS play all four hands. Press Option, Replay, Okay, then if you wish, press Scramble to mix up the W-N-E cards to try the same South cards against various other distributions of opponents' cards. You can scramble the cards as many times as you wish until you press Okay. If you want to scramble the cards after pushing Okay, push Options, Replay, Okay, then Scramble.
Break spades before leading *The default is that you cannot lead a spade until a spade has been played on some other trick. By turning this option off, you can lead spades whether spades have been played or not. The main (though slight) advantage of being able to lead spades at any time is when you have a two-suited hand with the top spades. You can start by drawing in the outstanding spades, then running the second suit. Since a two-suited hand is going to be short in at least one suit, even without being able to lead spades at any time, it should be easy enough to trump the short suit and then lead spades, But this is a common option which doesn't require any special code to speak of, so here it is.
Allow ruffing on first trick *If this option is turned ON, a player who has no cards in the suit led may play a spade on the first trick if so desired. With this option turned OFF, spades may not be played on the first trick. Since there doesn't seem to be any rational reason why the first trick should be protected from ruffing as opposed to any other trick, the default is to allow ruffing on the first trick.
Nil bidding allowed *If a player thinks he can avoid taking any tricks, he can bid "Nil". If the other players force him to take a trick (by playing their low cards), then he loses 100 points. If he avoids taking any tricks, he scores 100 points. The "Nil Bid Score" menu option allows you to change the points won or lost on a Nil Bid to any desired value. Because the usual score from making or not making Nil is so large, it should not be bid unless you have a near-perfect Nil hand or unless you are way behind with one player about to win. If CSS bids Nil and is not way behind, you can almost forget about trying to make him take a trick, since CSS will not gamble 100 points without a near-perfect Nil hand. Playing Nil bids is a poor option for serious players. Among equal players, it can distort the results because every 40 or 50 deals, someone will get a can't-miss Nil hand. This is a hand with so many low cards that the player is almost certain to make the Nil bid and get 100 points. Like winning the lottery, getting a perfect Nil hand is nothing but luck, yet the Nil bidder gets 100 points, which obviously has a very significant impact on the score. Among players of unequal skill, the lesser player can gamble with Nil bids on hands that usually will not make Nil, but if that player is lucky, he can make the bid. The 100 points probably is not going to cause the less skilled player to win (more often than not, he will just drop farther and farther behind by bidding Nil and going set), but defending against the Nil bid can mess up the normal playing strategy and game results for the other players. Even worse are "blind" Nil bids where you gamble on a Nil bid before looking at your cards. This simply takes the already blind luck of the Nil bid to an even more exaggerated extreme. In addition, unless an opponent is really on the verge of winning the game and you are a couple of hundred points behind, you should not give up on making a comeback. It is not uncommon for the high scorer to make a big bid, get some bad breaks, and go set, allowing you to catch up. However, unlike some other rules variations, Nil bidding is so widely used that it has been added to this game despite its shortcomings. But if you find yourself always behind and having to bid Nil to catch up, a better solution for you might be to play some of the non-expert computer opponents until your game improves.
Program Playing StylesThere may be times when you would like a little variety or just don't want to play the strongest possible CSS opponent. For example, a beginner or a child may get discouraged at losing to CSS all the time and would prefer someone of their own skill level. Or you may just want to see how well other playing strategies work. In any case, you can select a different playing style for each player or the same style for each player. (Try playing against three Vicious players and see how you make out.) You can change the playing style at any time during play, but the effects may be bizarre. It is recommended that you NOT change styles during play. * Note: In the Free CSS game, only the Expert style is available.
Normal - plays most like the average human player. The Normal player rarely, if ever, knows how many cards are out in a suit or who is likely to have them. This often results in his playing a high card on a trick where the next player is known to be ruffing, or ruffing and getting over-ruffed. His bids are usually not the best, either, sometimes overbidding or underbidding. Though the default playing style is Expert, there is no shame in changing the playing styles to Normal (just as in the typical first-person shooter video game). Timid Timid - will never bid a card which is not rated a solid "1" (as opposed to .7 or even .9). Timid will never bid Nil. Timid should be pretty easy to beat because he's always underbidding. He also doesn't like to trump if he's not the last player because he's afraid of getting overtrumped. He also doesn't remember the cards that have been played nor is he good at analyzing other players' bids and plays. Cautious Cautious - Doesn't take any unnecessary risks when bidding or playing. If a Cautious player bids Nil, the odds are extremely small that you can beat him. A Cautious player will only bid cards rated at .8, .9 or 1. He will try to take his winners quickly before they can get ruffed, such as leading Ace from AQ instead of trying to score the Queen first by waiting for someone else to lead the suit. Likewise, he will not ever finesse (play the Q from AQ when third to play and the King is out). If he's not last to play, he will trump with his highest trump if there's any chance of getting overtrumped. His memory and analysis of cards played is good. Unless using Scaled Scoring, ALL computer players become Cautious when way ahead and closing in on a game, in order to keep from blowing their lead. Expert Expert - the best overall playing style. The Expert player takes reasonable risks in bidding and play. and his memory and analysis of cards played is good. As a default, computer players are assigned the Expert style so that you can against play CSS's best game. Aggressive Aggressive - takes more chances. For example, an Aggressive player will always bid 2 with cards rated 1.6+ by CSS, whereas the Expert player would just bid 1. An Aggressive player is more likely to trump even at the risk of getting overtrumped. Vicious Vicious - is out to get YOU! The Vicious player is an Expert who will take a chance of not making his/her own bid for a chance to set you. On the positive side, this means that he goes set more frequently than the less psychotic players. Note: To maximize the effect of a Vicious player, put him on your left so that he plays immediately after you do. On your right, he can't do as much damage since you would often play after him. Every player becomes Vicious towards any player who gets way ahead and close to winning a game unless Scaled Scoring is on or a Tournament is being played, but a player with the actual Vicious style keeps after you no matter what. Note: the Vicious player does not cheat (i.e.: peek at other players' cards); but he usually saves his high cards and his trumps to try to play them when you play high cards. Cheater Cheater - peeks at everyone's cards! When a computer player cheats, He will virtually never overbid, such as counting an Ace when someone else is void in the same suit and able to ruff the Ace. During play, the Cheater would normally never lose a finesse or get over-ruffed when cheating. A Cheater might even cheat by not following suit ("revoking"), such as trumping a suit and later on, discarding on of the same suit. Since you would stand little chance against a player who always cheats, a Cheater doesn't cheat all the time. If you suspect a player with the Cheating playing style of cheating on a particular play, you can accuse him and win bonus points if you are right. If you think that a Cheater has played a card in a suit that he previously trumped, you must accuse him either as soon as he trumps or when he later plays a card from the same suit. For information about scoring bonus points against a Cheater, see CSS Cheated! Mystery Mystery - If you really like surprises, let CSS pick opponents for you by selecting the "Mystery" style. Just like in real life, it will be up to you to figure out each opponent's style and adjust your play accordingly.
Scaled Scoring *When one player approaches game, the other players will gang up to try to set him. The losing players may also start bidding and playing more wildly to try to catch up. At the same time, a player who is way ahead and approaching game will (or at least should) bid more conservatively to avoid losing a game that is essentially already won. All these strategies distort the normal bidding and playing of the hand. To counteract this, you can select "Scaled Scoring". Scaled Scoring gives each player points equal to the difference between that player's score and the scores of those players lower than him/her. For example, if the final scores for players 1-4 are 510, 450, 400, 350, respectively, then player 1 would get (510-450)+(510-400)+(510-350) for a total of 330 points instead of just 1 point for the win. Player 2 would get (450-400)+(450-350) for a total of 150 points. Player 3 would get (400-350) for a total of 50 points. And player 4 would get nothing. To keep these scores from getting unnecessarily large and hard to handle, we simply divide them by 10 to get smaller numbers with the same comparative value. In the example above, players 1-3, who got 330, 150, and 50, respectively, would instead get 33, 15, 5. Scaled Scoring Better Reflects Game ResultsTraditional scoring - in the Wins column, whoever won the game gets a single point, whether he was 400 points ahead of everyone else or 1 point ahead. The 2nd-place player gets nothing, even if he came within a single point of winning and was 200 points ahead of the 3rd and 4th players. Scaled scoring - the second-high and third-high scorers get points reflecting how high they scored. And the higher their game scores, the greater the number of points they get in the Wins column and the fewer points the high scorer gets. This means that even if it seems like you have no hope of winning, you have to continue to play your best game in order to hold down the number of "win" points the higher scorer(s) will get. If you are way ahead, instead of bidding conservatively and just high enough to finish out the game, you must continue to bid what you think you can make in order to maximize your points in the Wins column. As a way of keeping the high scorer from underbidding, this is a much better alternative to "bagging" (deducting points for overtricks), which is a penalty easily avoided by careful play. A good player should almost never get hit with a penalty for bagging in other computer Spades games. Nil score *Lets you change the points for making or going set on a Nil bid. Change game goalLets you change the score to which the game is played. Default is 500. The higher the game score, the longer it takes to complete a game, but this also allows more time for luck of the deal to balance out. In a game to 200, for example, the luck of the deal is a MUCH larger factor. Show ScoresThis option determines whether or not game scores are shown during play. If Show Scores is unchecked, scores are only shown at the end of each trick, deal, and game, but not during bidding or play.
Play a Tournament *When you are ready to play a SERIOUS game -- no peeking at cards, no aborting a deal that has turned sour for you, no asking for hints -- click on "Play a Tournament" to play against the CSS players. WARNING: Tournament play is not for people with weak egos!!! (But then, neither is CSS in general.) Just getting INTO the last round is an accomplishment. To win a tournament is something to really be proud of! In the Tournament mode, you play 10 sets of a specified number of deals per set. (See below for specifying the number of deals per set.) You can select the rules you want to play by before starting a tournament. At the end of each set (except the last one), the low scorer is out of the tournament and (if that isn't you) is replaced by another CSS player. Early in the tournament you will come up against some of the weaker players. As the tournament progresses, you will run into tougher opponents playing styles such as Vicious and even Cheating. (Don't forget to use the CSS Cheated! menu option if you think you have caught someone cheating.) In the last set, your opponents will be Mystery Guests and you have to figure out their playing styles and the best way to combat them. At the end of the last set, the high scorer is the winner of the tournament. If you are the low scorer at the end of any set, you are out of the tournament, so the tournament is effectively over. You will be told how much money you lost to each of the other players. Bring your (virtual) wallet! Before a tournament starts, you can select any rules you wish for the tounament. During the tournament, you cannot change the rules. You will be able to Concede the tournament if you wish. Two Tournament Rules are used to "encourage" each player, whether high scorer or low scorer, to continue to bid and play his or her cards to the hilt. The first rule is that the low scorer has to pay the other players $1 for each point by which they outscore him. This only takes place in your imagination, of course, but the idea is that the low scorer cannot afford to take foolish risks, such as bidding Nil without a good Nil hand just to try to get back in the game, since going set on risky bids increases the amount he (or you) has to pay the other players. The second rule is more tangible: the 2nd and 3rd place players in each set start the next set with a minus score based on the difference between their scores and the 1st place finishers. For example, at the end of a set of deals, player #3 has 200 points, #1 has 176, #2 has 80, and #4 has 60. (Remember, a set is a specific number of deals, not to a specific number of points.) Player #4 is low scorer and so is out of the tournament. Player #1 starts the next set with the difference between his score and #3's score divided by 2, which is (176 - 200) / 2 = -12. Player #2 starts the next set with (80 - 200) / 2 = -60. Both the high scorer and the new opponent start with 0. The reason for dividing the difference in the scores by 2 is that while we want to provide SOME incentive for players to keep bidding and playing the cards properly, we don't want the players to start the next round so far in the hole that they can't get out, just because they lost by a few tricks. Specifying the number of deals After clicking on "Play a Tournament", you will be asked how many deals you want per set. You can enter any number. Five or ten deals are good for a fast tournament, but even if you are a very good player, you have to overcome any luck of the deal in a very short set. Twenty to 30 gives you a better chance to overcome a temporary run of bad breaks and/or bad cards and thus make it through to the end of the tournament, but 10 sets of 30 deals is 300 deals, which will take quite a bit of time to finish (depending on how far you advance, of course). Duplicate mode in a tournament For an even greater challenge, you can combine Duplicate Play with a tournament. [End of Tournament section.]
Concede gameIf you have fallen way behind and don't wish to continue the game, this button will clear the game scores and give a Win to the high scorer (assuming the high scorer is a computer player).
"View" OptionsIn addition to the View options listed below, notice that in the upper-right area, CSS also displays the time of day and battery life remaining. These are nice to know if you find CardShark Spades hard to turn off.Also shown is the current trick number, and above the Zip and Okay buttons, you will find the current deal number and dealer. These are handy in referring back to a particular play in a specific deal. Knowing who dealt and was thus first to bid and to play are also necessary. You can also use the deal and dealer numbers to replay a deal at any time by clicking Options and Replay. If you have a question about a particular deal, write down the deal and dealer numbers and the trick number as well.
View All HandsThis option lets you see the cards of the other players. During Duplicate Play, this option is automatically turned on, making Duplicate Play a good way to study how CSS plays. Also, all hands are also shown at the end of each deal even if this option is not selected so that you can see what you were up against. If a computer Spades game never lets you see the other hands, it's hard to know whether or not it bid and played reasonably. View Suit CountsAs described in The Program's Strategy section, each CSS player keeps track of all the cards outstanding (not in the player's hand), starting each player with the total number of cards in the current player's hand and adjusting those numbers based on deductions from the bidding and play of the deal. The Total outstanding, shown in the fourth column, is the exact number of cards still unplayed which are not in the current player's hand. The player usually doesn't know how the cards are divided among his opponents (though he can deduce that as play continues), but just knowing the possible counts gives CSS players a big advantage over most human players. The View Suit Counts option equalizes this advantage by letting the human player see the same type of information -- how many cards are outstanding and many each player may possibly have. In the image above, player #1 shows that any player could have up to all of the 7 total Hearts outstanding. When the next Heart is led and everyone follows suit, he will show that each could have 6, but when player #2 can't follow on the next round of Hearts, #1 (and the other two players) will adjust #2's possible count down from 6 to 0 while #3 and #4 will go from 6 to 5. In the image above, CSS is playing all four hands. If a person is playing and the other players' cards are not shown, the a list will come up of all the exact cards still unplayed in all three of the other hands combined. So if you have a Jack and can't remember if all the higher cards have been played, you can check that list. ExplanationsThis option displays the explanations for CSS players' bids and plays (except those which are obvious). The best way to study how CSS plays is to turn on AutoPlay, turn on Show All Hands, and turn on Explanation of Bids/Plays. In the screen shot above, player #1 (bottom of image) has decided to trump low. The explanation tells why he isn't worried about being overtrumped. The box containing the explanation is open towards the player whose bid or play is being explained. View/Email Deal RecordAt any time during a deal, you can use this option to view a record of the current deal. The deal record shows the bidding and the cards played on each trick. If the View Explanations option is checked, the explanations will also be shown in the Deal Record. * In the Duplicate Play mode, CSS's deal record is appended to the end of yours. You can view a deal record with "View Deal Record" up until Okay has been pressed after the scores for the deal have been shown. Touch the top line of the deal record to email it to Ford Software. The deal record is brought up in your email program and you can actually email it to anyone you wish. You can also get a screen shot of the screen at any time by pressing and holding the Hint/Zip button. When you let go, a window will pop up showing the name of the image file. The name is made up of the current date and time to keep screen shots from overwriting each other. Use a file manager to locate the file or use an email program attach it to an email. View Game Stats *The Game Stats option lets you see the degree to which the luck of the deal affected your game, plus how well each player bid and played his or her cards, to a certain extent. When you are through viewing the statistics, you are asked if you want to clear the stats file. You may want to clear the stats before a new game, but especially before a tournament begins. Average Tricks BID-MADE lets you see at a glance how much you underbid, on average, when you MAKE a bid. The next column shows the TOTAL UNDERBID since the last time the stats file was cleared. Average Tricks BID-SET shows you the average level at which you overbid and by how much. If the average you bid when going set is under 3, then that's not too bad, since bidding for that 1 extra trick on the 1 or 2 level doesn't hurt as bad as going for 1 more trick when you are bidding 5 or 6. The TOTAL BID-WON column shows who got the most winners dealt to them -- assuming that everyone bids and plays on about the same level of expertise. However, some of the players may underbid their hands and also play conservatively, so the total tricks won number is not indicative of the luck of the deal. Zip ButtonThe Zip Button replaces the Hint Button when in the AutoPlay mode. This option will cause the program to zip straight through a deal without any pausing until the end. It plays all four hands. This can be used to practice bidding. Leave AutoPlay checked, but when the cards are dealt, decide what you would bid for Player #1. Then continue until CSS bids for #1 and compare it to your bid. (If #1 is first to bid, CSS will bid before you.) Then click Zip to let CSS play through the deal and see how your bid compares to the actual results. Zip can also be used during CardShark's turn at Duplicate Play if you are not interested in studying the play of the CardShark and just want to generate its scores. This option is turned off at the end of each deal to prevent unintentional playing into the next deal. When Zip is turned on, the button caption appears in all caps ("ZIP"). When it is turned off, it appears as "Zip". Hint ButtonThe Hint Button replaces the Zip Button when not in the AutoPlay mode. If you would just like to see what The CardShark would play or bid in a particular situation, click the button at the bottom of the screen. When Hint is pressed, the card(s) selected by CSS will pop up. You can put the card(s) back in your hand by clicking on the card(s), then click on the card(s) you want. To accept CSS's suggestion, click the Okay button. CSS Cheated! *A player using the Cheating style is almost impossible to beat. So to make the game against a Cheater a little more interesting, he does not cheat on every play. In addition, many of a Cheater's plays will just naturally be the same as the non-cheater, such as when playing a high card to win a trick after everyone else has played, or when discarding a low card under someone else's Ace. This is similar to a real-life situation where a person would probably not be able to cheat on every play. Also as in real life, if you catch a player cheating, you can exact some punishment on him. When a player is using the Cheating style, a button will appear on screen which says "CSS Cheated!". Click on that button if you are convinced the player cheated. If the player DID cheat, you get twice your bid added to your score for having caught the scoundrel. If the player did NOT cheat, then twice your bid is deducted from your score for having insulted CSS player. (Obviously, the accusation cannot be made on the 13th trick.) CSS will not lie about the player cheating, but if you still don't believe CSS, you can always view the deal record to see the cards held in CSS player's hand, as well as the explanation for whatever play he made. Remember that none of the players will cheat unless you have assigned them the Cheating style of play.
The Program's StrategyCSS has over 10,000 lines of code for analyzing its own hand and the opponents hands. These routines can vary depending on options chosen by the human player. Luck can effect the results of Spades. For example, if you are dealt AK4 of clubs and bid to take 2 clubs, it is possible, though not very likely, that someone else will have just one club and will trump the second trick, or even have NO clubs and will trump both the Ace and the King of clubs. But this situation would be relatively rare, and bad luck gets spread around to all players and usually evens out during a game, so the key to winning is to always make the plays with the highest odds of success, given the cards you have. How The Program BidsWhat would you bid with the hand In a regular game, bidding also depends on the score and possibly on other factors, but before you can make adjustments for these things, you have to understand what a hand is really worth on average. The first factor to consider is that AKQJ in any suit but Spades is virtually never going to win 4 tricks on its own. At best, the other 9 cards in the suit could be divided 3-3-3 (among the opponents) and since the J would be the 4th card played, everyone else will be out of the suit and the J will surely get trumped. And 3-3-3 is less likely then 4-3-2 or worse, meaning the Q will also get trumped. Even 5-4-1 or 6-3-0 are possible, in which case the K or even the A will not take a trick. As a general rule, you should not count beyond the 2nd round of a suit. Say you have KQJT. The Ace will take the 1st trick and your King, hopefully the 2nd. One of your cards winning the 3rd round is a long-shot. One advantage CSS has is being able to use more precise evaluation of cards. To arrive at the total bid for a hand, it will use tenths of a trick in valuing holdings. The valuations shown in the game (optionally) will be explained below. In CSS, you can click on the Show Explanations option to see why bids were made. (The box containing the explanation is open towards the player whose bid or play is being explained.) Although the average person cannot be expected to remember the fractional points each suit-holding plus distribution is worth, if you watch the explanations of CSS's bids enough times, you will begin to pick up on what various holdings are worth. Most of the bids made by CSS are based on simulations, in which 100 or more deals are played where South's cards are kept the same and the cards of the other three players are shuffled each time. Say that we want to know how many tricks might be won with AJ86 of Spades and no short suits. A run of 100 deals in which South has Running a simulation, we find that the hand will take 2 Spade tricks 41%, 3 tricks 46%, and 4 tricks 13% of the time, and never fewer than 2 tricks. To translate these numbers into fractional tricks, start with the number of tricks which contains some number great than 0, which in this case is 2 tricks with 41%, and the rest of the time it will take more than 2 tricks. So 100% of the time that hand will win 2+ Spades. The percentage of the time it will win more than 2 Spades is 100% minus the percent that it will take exactly 2 or 59% (100-41). So the Spade suit is worth 2.59 tricks. To be a little conservative, round down to the nearest tenth: 2.5. Some other complications must be considered, and they will be covered in the full version of CSSpades. Now let's say that we have another suit worth 1.2. Add that to the 2.5 and we get a hand value of 3.7. Should we round that up to bid 4 or just bid 3?
The payoff for bidding 1 more and making it is +10. The penalty for bidding 1 more and going set is minus the amount bid bid times 10. For example, to go from a bid of 1 to 2, you risk -20 to make +10. To break even, you have to make the bid two times for every 1 time you go set, and 2 out of 3 is 67%. We don't want to deal with hundredths of a point, so we round that to .7. To go from a bid of 4 to 5, you risk -50 to make +10. To break even, you have to make 5 six times for every time you go set, and 5 / 6 is .83, which we round to 9. The rounding up means that instead of just breaking even, you should make a little profit in the long run by following this chart. And it shouldn't be that hard to remember. It starts with .7 for going from a bid of 1 to 2, then .8 for next two steps up, and .9 for higher bids.
![]() While the average player may bid 1,2 or 3 for a holding of KQJ3 in a suit, CSS counts it at its true expected value of 0.7 (based on a number of simulation runs). These percentages for each suit total up to something which, even after rounding, may be different from what would be bid using only whole numbers. Nevertheless, it isn't necessary to try to remember all these percentages. It is enough to remember that KQJx in one suit and Kx in another are each worth a little less than 1, so the two combine to a fairly safe combined bid of 2. The three high Spades are good for at least 2 tricks while the 2 is worth a little something either for outlasting the other players' Spades or for trumping a Club. ![]() As you can see above, while on average the hand will win .7 tricks, in this deal it actually won 2 because each of the other hands had exactly 3 diamonds, meaning that the Ace, King, and Queen could each win a trick without getting trumped. Also, if East had the AQ of Spades instead of South, North's KJ of Spades would have been less likely to win tricks.
While A lone With a doubleton and only 2-3 small Spades, bidding a ruff is risky. If someone has Despite the previous paragraph, simulations have shown that a hand with 3 little Spades and no short suits (e.g.: 3-3-3-4 distribution) will, on average win 1 trick with a Spade.
With long trumps headed by the AKQ (but no Jack), the question becomes: what are the odds of being able to draw all the trumps without losing any? With 6 Spades, the flattest distribution you can get is 4-3-3, then 5-3-2, etc., where the 4-card suit is potentially a stopper on the 4th round of the suit (e.g.: Jxxx always stops AKQxxx). However, running several 100,000 deal simulations came up with a surprising result of AKQxxx being able to draw all the trump an average of about 70% of the time, meaning that you will, on average win 5.7 tricks with With 7 Spades to the AKQ, you can draw all the trump almost 90% of the time... With 7 Spades to the AK, 6 is definitely a safer bid. What would you bid with KQJT82 of Spades? You should bid 5... Even with the Ace of Spades and 6 small ones, 5 is a pretty safe bid...
Bidding and playing 2-suited hands: The kind of 2-suited hand referred to here consists of Spades and a side suit. The simplest to bid and play would be something like You will very rarely, if ever, get such a perfect hand. More common would be a hand like A hand like A Two-Suiter Suddenly Appears: Say that you start with a hand like S:AKxx, H:xx, C:Txxxx, D:xx. Your Spades are not long enough to control trumps and your Clubs are not good enough to control that suit to consider this a two-suiter. But say that after a few tricks have been played, you are down to S:AKxx, H:-, C:Txx, D:- and opponents have played a couple of trumps. This hand has evolved into a two-suiter. The question is how to play it... Bidding when you are behind: When you are way behind (50-100 or more points), you should bid normally. The best way to make up ground when you are behind is by setting the player(s) ahead of you. If you stretch to bid one more trick, even if you make it, all you get is another 10 points. But if you set a player ahead of you, you shorten the gap by 30, 40, even 60 or more points, depending on his bid. When everyone "underbids": You may see deals like the one below where only a total of 9 tricks were bid. What happened to the other 4 tricks? Well, #1's Clubs are worth a little less than 2 because of having 5 in the suit. The AK are 2 sure tricks and the 3 may be worth something by trumping a Heart. #2's Hearts are almost worthless because of their length and odds of getting trumped (still, a bid of 1 is preferable). #3 has the Ace of Diamonds and counting 3 of the 5 Spades is actually pretty aggressive. #4 can count 1 for the King and 1 for trumping value. Obviously, SOMEBODY is going to be making overtrick(s), but it's not really possible to say when looking at an individual hand who it will be. This is yet another example of why "Bags" is not a good rule: everyone has bid correctly, yet someone would be penalized for taking overtrick(s).
How The Program PlaysCS-Spades' main strength is that it never forgets what has been played. It usually knows (statistically) the best time to play a high card or a low card and when to risk an over-ruff by ruffing low. Here are things the program normally will NOT do: ("..." indicates more discussion is available in the complete app's documentation.)
2. Lead a K from Kxx or even lead small from Kxx... 3. Lead the top trump out unless (1) there is only one other trump out, or (2) the player also has the next high trump(s), or (3) the player has winners in other suits and wants to reduce the odds of those cards getting trumped... 4. Lead any Spade from K9862, unless trying to clear out trumps to make a long side suit good... Here are some things the program normally WILL do:
2. Finesse: When the 3rd to play has a suit of 4 or less and two small cards have been played, CSS player, holding AQ... will play the Q and hope that one of the first two players has the K. In Bridge, this is called a "finesse"... A finesse can also be made in a suit other than the one led. It's usually in Spades. For example, with two cards left, #1 leads a Heart; #2 plays the J of Spades and #3 has the AQ and the King is still out. #3 should play the Queen because if he plays the Ace and #4 has the King, he can just discard (or trump low if he has 2 trumps) and the King will take the Queen on the last trick. If #1 or #2 has the King, they will take the Queen on the last trick. If #3 plays the Queen on the Jack, the only way he can lose is if #4 has the King and then his Ace will win the last trick anyway... Another situation is when there are 3 cards left and you have 3. Force the Ace: With KQJx, the program will play the K to force out the Ace so that he can take tricks with the Q and maybe even the J before someone has a chance to discard in the suit and ruff it. 4. Keep top Spades with lowers: With, say, S:A93 and no good side suit, CSS will not lead the Ace of Spades if he hopes to ruff with the 9 and 3... 5. Draw trumps with tops: Lead Spades when only top ones are left... The program can also draw conclusions/inferences from the cards played... Following are more extensive examples... Examples of Expert Analysis... Nil Bidding... No CheatingThe program plays each hand individually (unless you have given a player the "cheating" playing style, obviously). Each computer player keeps track of its own personal knowledge of what's going on based on what that player has passed, played, and seen others play, as explained above. When deciding what to bid or play for one hand, CSS never "peeks" into the other hands. It uses only its own knowledge. In the deal record saved to disk, reasons for each play, as well as inferences made from the plays of others, are shown. You can see exactly why CSS has made each bid and play. For more information about how CSS plays, check the Strategy section which shows some of the complex analysis the program uses to figure out what other players have. If you really want to test CSS, use an External Deal. Defending Nil BidsDefending against Nil bids is difficult because while you are trying to play low and force the Nil bidder to win a trick, you also have to be sure to win enough tricks to make your bid... Ganging UpOne basic strategy is that if one player gets way ahead (by 100+ points) and is getting within striking distance of winning (over 350 points in a 500 point game), the other players will try to set him. (NOTE: This strategy is not used in Scaled Scoring or in the Tournament Mode.)... Bottom line: if you get way ahead and think that CSS players are ganging up on you, you're not paranoid, they ARE ganging up on you, but they will also gang up on any computer player that gets way ahead. When NOT To RuffSome computer Spades games (and some people) play that you MUST ruff when you can't follow. This is a very poor rule, since it takes away what is clearly the single most significant strategic decision to be made in the play of most hands: whether to ruff or to discard. Even when NOT required by a rule, most other computer Spades games (and most people, including the Normal playing style in CSS) simply ruff every time they get the chance, without worrying about getting over-ruffed. Again, they are ignoring one of the most important aspects of the game. CardShark Spades uses its memory and analysis of the cards that its opponents have played and of the cards they could possibly have left. CSS players will not ruff if there is a good chance of getting over-ruffed, because that is just a waste of trump. You have to be particularly careful about ruffing if you have something like K-4 of Spades and have bid to make the King. If you ruff with the 4 and get over-ruffed, then someone leads the Ace and draws your King, you will get no Spades tricks at all. Another situation is if your are 2nd or 3rd to play with something like AK7 of Spades. If you play the 7, you may get overruffed, but if you discard another suit, the AK will protect the 7 from being led out of your hand and you may get a better chance to make the 7 later... If a computer player can ruff in one suit, but with a risk of getting overruffed, and he is short in another suit, the player may decide to discard from the second suit and hope to get a safer ruffing opportunity in that suit later on rather than to risk getting over-ruffed now. Note the deal below where #2 has led the Heart Queen. As you can see in #3's Suit Count box, he already knows that #1 is out of Hearts (probably because #1 has already trumped Hearts once) and will likely be trumping, so it would be silly for #2 to trump and have #1 overtrump. As the cards sit, #2 could trump with the Ten and #1 would not be able to overtrump, but what if #1 had the Queen or the Jack or even the Ace? It's safer to get short in Diamonds and trump that suit later.
On the flip side, if you have a long suit, such as AKQx, and the player on your left plays the 2 on the Ace and the J on the King, he is likely out of the suit now and able to trump if you lead the Q, but at the same time, he may be afraid of ruffing and getting overruffed, so there is a chance that if you lead the Q, he will discard instead of ruffing and maybe the next two players will have to follow suit... Implications of not ruffing: The major effect that players not ruffing has is that they discard from other suits instead. This can mean that if you bid 2 on a suit containing A-K-3, you may not make the Ace and King if you do not get around to playing them before the other players discard that suit and start ruffing it. So when faced with opponents who discard rather than ruff at every opportunity, you should take your winners as soon as possible. The exceptions are...
See Much More Strategy In The Final VersionIf you are interested in learning how to play expert cutthroat Spades, you should be aware that books currently available on playing Spades do not discuss Expert level topics (such as fractional point hand evaluation) and in addition, they only cover partners Spades for which the bidding and playing techniques are completely different. The Strategy section in the final version of CardShark Spades covers more topics than shown here. Look for it at the end of 2011.
Problems?CSS has been tested extensively. However, here's what one journalist says about Android:
If Android forces you to close CSS, try to note the deal number and dealer so that when you restart, you can do so from the same point (press Options and New Deal).
|