The Problem with Sequels
Or, "Why Do(n't) Chess Variants Work?, 2". Continuing exploring the difficulties with variants.

The Problem with Sequels


It can be very difficult to develop a good idea. How do you build on something memorable, yet simple, in such a way that doesn't let it spiral out of control? Musicians, poets, novelists, and of course film directors all experience these difficulties. Technique, or craft, is what separates amateurs and professionals--amateurs have interesting ideas, but often don't know how to develop them. Perhaps this difficulty explains why some movies begin with a very promising installment, and when you go to watch the follow-up, all of a sudden Jar-Jar Binks appears on the screen, causing untold headaches. These difficulties also apply to Chess. If Chess is so good, why isn't there a Chess 2? 

Many chess players think that the reason there is no Chess 2 is because Chess is just fine the way it is. This is certainly plausible. Magnus Carlsen continues to play interesting games of Chess, and the game seems to offer him enough scope to show his superiority to his competition. Isn't Chess thriving? It's not so clear to me that it is. When you have top players like Vladimir Kramnik suggesting that Chess be played without castling in order to reduce the number of draws, it looks as if something is amiss. And Kramnik's suggestion itself, while symptomatic of a real problem, seems like a bizarre regression. 

Too Many Problems

Part of the problem seems to be that Chess players are not exactly sure what the problem with Chess in its current state is. It seems obvious on first glance, but then it turns out to be hard to put your finger on. Inexact statements of the problems with Chess may hinder the possibility of a solution. And not all proposed problems with Chess are necessarily real ones. For instance, Zied Haddad, quite an accomplished designer of variants (one of which is discussed below), had the following to say on my previous blog. 

In these modern days, having boring and endless opening theory is not interesting for the spectators. I regret the old romantic days where long sacrificing and attacking lines, so beautiful, so spectacular that for sure it will attract public because of all the drama and energy (this energy is reflected in the piece movements; sacrifices and material imbalance).

Here Zied refers to what I call the creeping opening theory problem. But I have to nitpick a little. Firstly, openings like the Sicilian Defense are actually very old. When exactly did they become too well understood, any why did they become too well understood? I think that creeping opening theory is only a problem in Chess because Chess itself is a problem--it's too small, and humans have outgrown it. So we cannot use creeping opening theory as an argument for the claim that Chess is stagnant, without committing the logical fallacy of arguing in a circle. 

What do I mean that Chess is too small? Well consider these two opening variations...

The fact that Chess even has these variations which are used as drawing weapons, is a huge sign that something has gone wrong. There is nothing equivalent to this in Go, for instance. There is nothing like this even in crazyhouse, which has amassed a lot of opening theory. But the opening theory in crazyhouse is interesting. It's a window into a dynamic and exciting game. 

Secondly, any variant without some element of randomness will amass opening theory. Since there are engines, databases, and continuous opportunities for games, variants will amass theory with increasing speed and accuracy. And so promoting a game based on the lack of opening theory is misleading. But the opening theory is not intrinsically a problem. Opening theory of a game can only be as problematic as the game itself. Opening theory can only creep up on games which lack depth, and which fail to reward advantages. (More on that below). 

I also disagree about the old romantic days. Why? I think that people are just better at defending, and will always be better at defending. This is just a byproduct of our increased knowledge. (And perhaps also, a product of Steinitz's thinking, much of which applies more generally than to just Chess. For instance, he was the first to explicitly formulate the idea that it is not possible to punish one's opponent unless they have made some oversight, however small. And so Steinitz helped players realize that many of the sacrificial attacks that were considered profound were actually just premature and refutable.) 

I suspect that the old romantic days are irrecoverable. I also suspect that we ought not want to recover those days--since so many of those games were predicated on such poor defense. It is important that variant players keep their eye on what matters here. Increased knowledge is a good thing, and nostalgia for the days of Anderssen and Morphy is probably misplaced. Or at least, creating variants on that basis is misplaced. And such nostalgia seems like bad philosophy--bad philosophy is not just wrong but it prevents the growth of knowledge. Even in highly tactical games like crazyhouse, materialism is a rewarding method, and strong players cannot routinely win in Anderssen-esque ways--they need to be positionally astute in order to get promising attacks. And this positional knowledge will never go away. 

And what, exactly, is the problem with draws in chess? Well it would seem to indicate that even though one player is stronger than the next, there isn't enough scope for this strength to make a difference much of the time. Suppose you outplay your opponent, nurturing a +3 advantage all game. But suddenly, all you have left are a Bishop and a Rook pawn, which happens to promote on the wrong square.

Now the game is a draw. This is a bit silly. We might summarize this problem as the inefficacy of material advantages. Can players still win, despite having all kinds of ways for huge material advantages to not go wrong? Of course, because they often do. But there is no good reason why they should be required to work around scenarios like the one in the diagram. 

Too Many Solutions?

The reason for the absence of Chess 2 is not that Chess is fine the way it is. The reason is that Chess 2 is extremely hard to make. I discuss this in the prequel to this blog post:

As I noted in that post, there are many parameters one can change. Here are just a few things that can be changed. 

  • The victory conditions
  • The size and shape of the board (not just whether or not the board is 8x8, but what happens on the edges. In some variants, pieces may bounce off the edges, like professional wrestlers. In other variants, the pieces may teleport through the edges of the board Pacman-style. 
  • The number of boards (Alice Chess and bughouse are played on two boards. There are also some fairly obscene looking variants like 3 and 4 dimensional chess). 
  • The cells (square, triangular, hexangonal?) (Some variants even do away with 1 cell per piece). 
  • The identity of pieces
  • The number of players
  • Mechanisms for playing (crazyhouse drops, Seirawan style piece gating)
  • The pieces (there are very many "Fairy Pieces", whose movements are either combinations of existing pieces, or simply stipulated by the designers). 

But these changes barely scratch the surface. There are changes which keep chess "chessy". There are variants with dice, which take away what for many is a key ingredient of chess--complete information. Chess is not supposed to (one imagines), be about one's motor reflexes, yet games like bughouse and judo chess, while undoubtedly having great scope for strategy and calculation, necessitate physical speed. This, to my mind makes them almost like video-games--not necessarily a bad thing.

There are variants like dark chess which take away visibility of the players. You can only "see" the squares your pieces can move to. There are variants where parts of the board vanish, or rotate. Even this page, which offers many of the possibilities, again only scratches the surface. 

And we should also consider that most of these options can be combined with one another, leading to hybrid variants

Too Many Aims

In my last post, I distinguished a kind of variant--Rival Variants. This is may appear weird because I am contrasting some games based on their characteristics like conditions of victory and piece kinds, next to the intention with which their creators made them. But of course, these categories may overlap. 

The many options for chess variants, but it is important to consider that the way we think about a problem may itself be part of the problem. I have referred to "Chess 2" and to "chess variants". There is an ambiguity here. Are we trying to replace chess, or are we not? The variants which have been most ambitious and grandiose in their aims have often failed the hardest. Variants with modest aims, which make simple adjustments or additions, often make very playable games. 

Some variants are like sequels, both others are more like fan-fiction. Sequels set themselves up for the risk of failure. Fan-fiction on the other hand, takes practically no risks. If someone wants to write a few pages about how Darth Vader and Jar-Jar Binks leave the turmoils of the failing Republic and go off to an obscure planet to  start a band which plays primarily space jazz, this does not aspire to be canonical, and it will never cost anyone much money. On the other hand, producing many physical sets, with the aim of usurping the hegemonic status of chess, could set the aspiring chess inventor back a few dollars, if the variant fails to catch on. (Especially if the variant is physically dissimilar). 

In any case, I would be happy to see Chess continue on its pluralistic path. in particular got me interested in variants ever since someone told me to try bughouse. But...why not more variants? 

Audience Expectations

Bird-Morphy, played in 1858, contains some flashy moves which may have been more for the benefit of the audience than the position. And of course, Zied mentioned what is good for the spectators. But how closely do the interests of the spectators and the interests of the game align? 

It's easy to lose one's audience when trying to expand on the familiar. Successful musicians may be accused of repetitiveness, or they may be accused of "selling out". Film directors may want to challenge their audiences' expectations, but only end up giving them what they're already familiar with. Politicians try to convince people that they are progressive, when really they are merely a few steps behind the general population. What does this game suggest? In some sense, Morphy was revolutionary; in another, a traditionalist.

Of course, the execution of a game is distinct from the design of a game.

Many Chess players don't care about the theoretical status of the game, and have told me that it's still complicated enough for them as it is. This sounds hard to answer. I wonder what people said when Chess was on its way in, and its predecessor, Courier Chess, was on its way out. Were they unanimous in their boredom with Courier Chess--or did some Courier Chess players say that they would stick to that game as it was giving them plenty of intractable problems to deal with? 

One objection I have here is that people who struggle with one game may thrive in another. For instance, my chess was never very good, despite years of effort. But somehow, I manage a higher standard of play in crazyhouse and bughouse, despite having spent much less time on those games. So my response to players who say that standard chess is hard enough is simply to suggest that variants might suit your style of thinking more. And of course, the more deeply you understand a game, the more fun you can have with it. 

Someone once said to me that players will play the games that there is a demand for. But of course, this misses the fact that demand creates itself. People also play the games they have been taught and been exposed to. 

Time and Money

The test of any game is whether or not people want to play it. The tests of any game considered in relation to chess, are as follows. 

Can there be high profile matches, for large sums of money?

Are long time control matches interesting?

Are the players invested in the game?

Does the game have familiarity for chess players?

Chess 960 has been demonstrated to be of interest. It meets the first three criteria. You can read about Carlsen-So here, including a Magnus "tilt". (A tilt is when a player loses, gets mad, and plays bad). My interpretation is that a Magnus tilt is only possible because losing at 960 is so close to standard Chess. 

Interestingly, there was one game between Kasparov and Caruana where the latter got a kind of Benko-inspired b5 gambit. So I think that there is plenty of continuity with Chess in this variant, even in the opening. 

What would be ideal for variants would be longer games. Why not play crazyhouse at a long time control? It's a deep game, full of tactics. Why not take away the time scrambles and get serious? These games could be played either in person, or online. But the matches would be great steps forward for crazyhouse or nearly any other variant--maybe Seirawan Chess. Three minutes per player is a good way to experiment quickly and get a feel for the game, but it would be good to see games between strong players where one cannot use the clock to escape an inferior position. (I say this as someone who does a lot of flagging and running around with just a King.) When you are playing fast games online, losses don't mean much. You can just resign and start again. In the tournament hall, a loss has more finality and significance, and you need to be accurate. 

One More Variant as a Case Study

Since I used Zied's comment as a foil for my ideas here, I'll just end by looking at his variant. 

A strong and interesting Seirawan-looking variant is Musketeer Chess. White chooses a non-standard piece, then Black chooses one. Both players then use each of the two new pieces, which are then deployed behind the any of the 8 standard pieces before the game starts, and then introduced once that piece moves. 

In this position, I have a spider(!) on b1, and a Queen-Knight compound on behind the f1 Bishop. When the Bishop moves, the Queen-Knight enters the game via the f1 square. There are ten non-standard pieces to choose from. So this Seirawan-esque variant has a lot going for it. There are many possible pieces, but the player is not overwhelmed by all of them at once. And the continuity with Chess is clearly there. It manages to interweave all of the standard chess openings, like some of the variants I discussed in the previous posts. I will not detail all of the new pieces, but you can see for yourself here: 

Nonetheless, I suspect that the game might be more marketable if the non-standard pieces were cut back, and if there were clear and unarguable reasons for the pieces given. For instance, there is nothing arbitrary about the movement of the Knight-Bishop compound, or the Knight-Rook compound, or even a Bishop-King, Knight-King, or Rook-King, some of which occur in Shogi. But pieces like the spider seem a bit random, and the imagery is at odds with the standard pieces. 

The cleverness of the variant is that you do not always get your favorite pieces, but neither does your opponent. And so there is a lot of freedom and fruitfulness in the various combinations of pieces, interwoven with the usual openings. 

This has been a bit of messy post, but hopefully I have managed to tie up some loose ends. As always I enjoy comments and criticisms.