How to Theorize in Bughouse

In bughouse, relative mastery is rather straightforward to achieve. Whereas in standard chess, years of dedication may pay off very little, many bughouse players achieve a rating of 2300 and higher in less than a year. This is because bughouse is not as well understood as chess is. Bughouse is very poorly understood. This means that there is much less to learn in order to become "strong". Even though the engines can demolish any human opposition, they still require that humans micro-manage them to do it. They cannot, as of yet, understand when to sit in order to wait for, or to hold pieces, and they do not understand the game. We cannot really look to the engines for much guidance. That bughouse is still poorly understood is part of the fun, and something for us to explore as we go. 

A common misconception is that in order to become a strong bughouse player, you need to become strong in standard chess. This is not the case. It is refuted by the presence of many strong bughouse players who are not strong or even average chess players. There is also a great deal of theory in standard chess which is almost entirely inapplicable in bughouse. Some of this is not only unproductive to bughouse, but counterproductive, and much in standard chess has to be unlearned or at least reassessed or reintegrated. 

1. Mainstay openings in standard chess are either unplayable, harmless, or need to be relearned through the constraints and demands of bughouse.
2. Endgames are only applicable in rare circumstances.
3. Tactics which win material in standard chess don't win material in bughouse because the piece values are different.
4. Mates and attacks from a distance are greatly reduced in value because the dropped pieces always provide "blockers". 

There are a few analyses of bughouse opening lines, as well as collections of mates with dropped pieces. But there is very little about bughouse as a game in its entirety. By considering bughouse from the perspective of both boards, we can come up with a theory--a piece of software. And to run the software, all you need is average hardware. In other words, you don't need to be a "galaxy brain" or a prodigy in standard chess to become what might as be called a bughouse master. The usual emphasis on the greatest players, the top ten lists, and the like, become less interesting when you consider what those players know. It doesn't matter how fast their brains work as much as the ways in which they think. Strong players have theories--ideas about how the game should be played. These theories will consist of moves--openings, tactics, of expectations, and as I will explain later, even of aesthetic considerations. Because bughouse is played at such fast time controls, players bring the theory with them. The theory helps players deal with positions they have never encountered. No player's theory is complete. This not to say that there are not dominant players--the point is that much of what any player knows is implicit or vague. 

Some of the terminology in bughouse is multiply ambiguous. Take the term "sac-sitting" for instance. It's sometimes used descriptively, and at other times it's used as a pejorative, to imply a lack of skill. A sac-sitter sacrifices pieces, waits for more pieces to come from the partner board, and then probably sacrifices those pieces as well. There are at least four accounts of sac-sitting which vary according to what the idea doesn't include. 

1. It's not sac-sitting if you keep moving. You sacrifice material, but don't try to mate. You just continue playing against an opponent with a slightly exposed King.
2. It's not sac-sitting if you regain the material given. You use the exposed position of your opponent's King to recover the lost material and when all is said and done, the material is equal and your opponent is exposed, while you are not. 
3. It's not sac-sitting if you do it later in the game. You wait until the time is right, and then sacrifice material, probably with pieces already available to drop. 
4. It's not sac-sitting if you are up time.

When a player praises one player for their sac-sitting, and denounces another for sac-sitting, it may be in jest, or it may be inconsistency and hypocrisy. Or it may be that there are different meanings of the word. I myself favor the third understanding of sac-sitting. If you give your partner enough time to get you material, and you are able to set up an attack against your opponent's King, then you are not so much sac-sitting as attacking. (Or if you prefer, you are not attacking clumsily, but skillfully.) The vast majority of bughouse games feature material sacrifices; virtually none are decided by pure maneuvering. It therefore makes little sense to always use the term "sac-sitting" to refer to something objectionable. 

Not only is there no consensus on the definition of sac-sitting, there is no consensus on the value of king safety. How many pieces is it reasonable to part with in order to expose the opponent's King? I once partnered a strong bughouse grandmaster, whose opponent was sac-sitting. My partner said something to the effect of "my opponent has given two minor pieces. This should never work." I didn't think much of it at the time. Now I'm still not sure what to think of such a claim. I think that two minor pieces could be very cheap. If you get the opponent's King out far enough, and if you empower your partner to seek trades, two minor pieces is very little. To be fair to my partner in this match, I don't think he was saying that two minor pieces is worth any amount of king unsafety. But perhaps my partner didn't take into account the constraining effect on my board. My opponent was empowered to seek trades, whereas I was constrained to hold them. Given that my opponent in this series was already much stronger and faster than me, this was an additional advantage. There is no value to material which you cannot spend. (It frequently happens that players give the Queen, because it is the most expensive piece. When that piece reappears, and is taken, the player who parted with the Queen gets it back. I call this the paradox of the expensive piece.)

A curious thing happens in the following series. 

The series begins with 12teen, who is by far the strongest bughouse player, asking caspiwins, "what sort of flow is good?" caspiwins replies "low". After some games, 12teen suggests that caspiwins needs to deal with high flow. Why would the strongest player switch gears? Especially since 12teen has also said that one should always adapt to one's partner. Perhaps due to the difficulty of holding pieces against a strong and fast player who is empowered to seek them. If the trades are good enough for caspiwins' opponent (HelmsKnight), 12teen's opponent (chickencrossroad) can afford to trade at a loss. If you cannot even make trades which win material, you will eventually be forced to make positional concessions. (Some players call this "being forced to pretzel".) There is such a thing as "reasonable flow", since trades are not only a premise of the game, but trading becomes increasingly likely over time. It also seems that stronger players trade more, and they trade faster. 

Holding flow is possible, although I've noticed that it helps to know before any moves are played that it will be necessary. There are some players who only attack, and if you see them playing your partner, you can play like a turtle and give only pawns. But even this comes at a significant cost. In standard chess, there is such a thing as the conversion of advantages. At a high level, this may consist of something like the following. A player uses the first move advantage to win a pawn, which later helps to win an exchange. This produces a more or less easily winning endgame. I'm sure that this exists in bughouse too, except that it may occur between the boards. It can occur in very many ways. But an extremely common one is that one player ruins their position by holding trades, and gets mated. Another one is that instead of mating, you simply "cash in", taking some material and abandoning mate on that board, but using the material on the other board. 

Is there anything which one player can reasonably ask of another? Some players ask their partners to be FAST and SAFE. Some demand unreasonably high flow, others demand unreasonably low flow. (They may not be demanding as such, but their position crumbles without their preferred unreasonably high or unreasonably low flow.) But isn't the idea of reasonable flow relative to the flow you want to give? How do we know what's really reasonable? Already, we've hinted at the answer. Trades occur naturally. But there is another way to consider this question. 

Suppose that you enter a game of bughouse, and you do not know any of the other players. How do you play? Anything from no trades to every possible trade could come from the other board. Not knowing in advance the sort of flow you are going to get, and not knowing what sort of flow will be required, you will want to play in a way where trades are not necessary, but not bad. And you will want to play in such a way that you can get your partner trades, or hold trades. Assume also that your partner and opponents don't know you or your usual way of playing. What follows from all of this? It will probably be wise to adopt a certain flexibility, neither early sac-sitting nor sac-accepting. If any sacrifice is made or accepted, the payoff should be proportional to the material given (or accepted). If you don't sac-sit, you can still apply pressure in various ways. And it is nearly always possible to prevent a sacrifice on your own King by dropping a Bishop, or moving the Queen. 

As I mentioned above, players have theories about what good play looks like. The theories consist not only of moves, but of aesthetic considerations too. Some players find certain closed systems boring. Others find sharp positions too volatile. Some methods of winning are showy and not parsimonious--converting the position to a chess position where there are strictly simpler ways to win. Some methods of play are rather unfulfilling, even when they happen to work. 

 The classic sac-sit. Suppose that this were objectively good for White, where would this leave the game? You could win just by making the most obvious moves with a single piece. The point of games like bughouse is presumably to think--and this sort of play is rather unthinking. Of course, the partner, who has to scramble to get trades, may not agree. It can be very hard to get the necessary trades for such play in a timely fashion. And if the trades are not coming, both boards are down material with nothing to show for it. 

It may sound like vanity to speak of aesthetic considerations, but we should not dismiss the idea of aesthetics entirely. Over time, strong players look for signs of intelligent play. Elegance and strength frequently converge. Players may be vain about the signs of strong play, and may try to apply them where they are inapplicable. It's always a good idea to be aware of your own theory about what good play is, so that it may be a guide rather than a formula. When I was rated around 2200, I was pretty much a sac-sitter. A strong player told me that I should try to learn to play without trades. This was helpful advice, and I realized that strong players often strive for efficiency. They use both the pieces they begin with, as well as the pieces they get from the other board. But it is quite possible to over-emphasize development. If a certain amount of flow comes, it's a bad idea to prioritize the development of the pieces on the back rank--it can get you into lost positions. The FM 1800_Strength I think coined the term board presence, which is a nice way of putting it. (I've never seen anyone else say it that way.) Once you have board presence, you are far less likely to suffer due to high flow. 

Somewhere in this video, in the bughouse part of the stream, the standard chess (and bughouse) Grandmaster Nils Grandelius says that aggression is better in bughouse.

I think this is right. Given that trades must eventually come, it is better, on average, to be attacking than to be defending. Assume also that you make at least one inaccuracy during a game. If you make an inaccuracy in attacking, you may fail to mate. If you make an inaccuracy in defending, you may fail to live. Attacking gives a player more scope for error. It's like walking over a bridge, rather than walking over a tightrope. 

It's important to understand losses with some objectivity. One of the most commonly used phrases in bughouse is "mb" (my bad). There is more to this than assigning blame. It's important to understand the real cause of a loss. Sometimes a team may co-create a loss, by embarking on two differing plans which are both feasible, if only the other player co-operates. Often the post-game chat looks like this...

Player A: You mated me! Sit! 

Player B: I got you knight because it mates. Keep check! 

The truth is one should nearly always co-operate with the request to sit. And one should equally nearly always co-operate with the request to keep check. If your partner is asking you to keep check, it's probably because they can trade, or they are forced to trade.

Sometimes, a team cannot agree which of them should deliver the mate, and which player should take the mating piece. Usually, it is pointless to insist that the other player is the primary cause of the loss. One exception is in trading Queens. I've come to the view that it is the job of the player whose King is endangered by a Queen to say "no Queen" and not the job of the player trading Queens to look or to give a warning about the Queen. Given that bughouse is predicated on time, it is best to assume that you can give any piece until you hear otherwise. It's hard enough to play one board, without having to second-guess oneself because a partner's intentions are unclear. And so anyone who says "you should look before giving Queen!" is completely wrong. One should always have some awareness of the partner board, but you cannot always know what will mate in the future. You don't want to be forced to spend time typing "are you going to fix? or conversely, "Queen is fine". Ideally, there is a minimum of typing unless there is a mutual sit. 

Sometimes I find myself seeing that my partner is about to give a Queen for a Bishop, perhaps. And I want them to hurry up and do it so I get the Bishop first. The partner worries about my position, which may be less than ideal, but is far more playable if I at least get the lesser piece in a timely fashion.

Being forced to seek trades can be hard, just as being forced to hold trades can be. Things get harder when partners ask you to get some pieces while at the same time holding others. But sometimes this is necessary, or just good because the payoff is so high. If you can get me diags (pawns and bishops) without giving a knight, I can ignore the threats and play aggressively. And communication is necessary here.

Let's return to an earlier question--how do we determine the real cause of a loss? You will often receive the blame for losses, often absurdly and illogically. If you are an improving player, you want to know what, if anything, you could have done differently. In order to distinguish legitimate feedback from nonsensical trolling and derailing, it helps to consider the following. 

1. Did I give excessive material? 
2. Did I fail to send reasonable flow?
3. Did I make unreasonable requests that material be held? 
4. Did I play slowly? (I also think there is such a thing as playing too quickly).
5. Did I fail to accommodate my partner's strategy? 

Given everything said above, it is a good idea to reflect on the last point here. When I proposed that we assume that there is mutual ignorance between players, this was a thought experiment. In actual play, you will become familiar with the players. You will be able to spot an alt at a glace. (And very few players exhibit much degree of variance. In their choice of openings and overall strategy, most players stick to what they know and enjoy, even though variance can yield great advantages). Even if someone sac-sits unsoundly, or if they do the reverse (accepting sacrifices and defending weakly afterwards), you should make every effort to accommodate them. If you are in a difficult match situation, with a partner who clearly needs to change something about their play, it's a good idea to propose the smallest possible adjustment to their play. If you lose games where you win material, never ask your partner to get you anything, never ask your partner to sit, play with reasonable speed, and so on, there was nothing you could have done. It's important to know when it wasn't your fault, so you don't change a way of playing that in a fairer matchup, would produce wins. Clarity will help you avoid making bad adjustments to your game. Sometimes it helps to go into the partner's game and see what happened. (Obviously seeing clearly the real causes of a loss is distinct from explaining the cause to your partner, and you can at least in principle, see the cause and not share the knowledge with a partner who prefers not to hear it.) 

What about improvements in your own play? Any player can stagnate. It's a well-established law that when you reach a new highest rating, you must then lose three hundred points. If it feels like you are playing worse, this may be a sign that you have learned something, which you have to integrate into your game. If you have stagnated for more than 6 months, switch your opening. Don't play to be a better player; play to discover or prove something which interests you. Many players have found that playing the bughouse Sicilian has helped their overall game. As I mentioned before, you don't need standard chess. If you are 1000 rated in standard chess, you can be proficient in bughouse. If you know what a fork is, you'll do fine. Patience and self-criticism always pay off. 

Bughouse is interesting. It would be even more interesting if the variant became a parent, and there were some variants of bughouse. Chess players resist change, but they have no good arguments for it. Every argument I have heard against changing standard chess could have been made against standard chess itself with equal cogency. There is a massive industry in continuing to split hairs in the sale of DVDs about the Caro-Kann defense and other sleepy variations. The grandmasters who produce these make obscene amounts of money, so it's natural that they would be unwilling to accept any changes to chess. Bughouse is still very interesting, but it could be made more interesting--a good first step would be to shuffle the pieces. How much of what's been said in this post would still apply to bughouse 960? If it's true of bughouse, it ought to be true of bughouse 960. 

This post has deliberately not contained any openings or tactics. My other blogs have opening and tactical advice, which is sound enough, if you are below 2000. "Commonly missed opportunities" contains a variety of tactical motifs. I also have blogs on the London System for White and the French and Sicilian defenses for Black. I may rewrite those blogs to reflect some refinements I've found since I wrote them. But part of the reason I've been so non-specific is that many players want to play their own opening lines. So this post has been more about providing some tools for self-criticism. It would also be interesting to be pushed back on about some of the claims made. Feel free to let me know what you think sac-sitting is, or ought to be, because I don't really know.