In that first exercise I chose my own first move based on what I considered to be the logical requirements of the SP, and I chose my opponent's first move by determining the least favorable move on his list. My hypothetical opponent was HarryO of the Chess960 Jungle blog, who has set up a second experiment similar to the first. He started by identifying an SP with a long list of first moves, SP395 beats SP491 in the competition for the most number of viable starts, then whittled the list down by applying another level of logic, What is a logical first move?.
If the position shown in the following diagram looks familiar, it's not your imagination playing tricks. It is similar to the SP used in the first 'Choosing' exercise; the only difference is that the King and Knight are switched. Just as in traditional chess, where a small difference in two positions can make a large difference in the evaluation of those positions, similar chess960 SPs can have large differences in their logical underpinnings.
For example, in SP491 I saw 'no reason to favor one of the castling options over the other'. In the current position, the players can castle O-O without any preparation, while castling O-O-O will take four moves: one Pawn move and three minor piece moves. This means that castling O-O is more likely to occur. When you consider that the players will also need to weaken the Queenside (a-side) slightly to let the corner Queen out, castling O-O looks even more likely.
Based on the principle of making the obvious moves first, 1.O-O is a good candidate move. Unlike many chess960 players, I prefer to hold off early castling even when possible. Castling, being a mainly defensive move, is somewhat passive and thereby risks losing White's natural initiative. On top of that, there is always a possibility that castling to the other side will prove to be appropriate. Are there any other good moves?
The move I chose in the SP491 exercise, 1.d4, suggests itself again. It plants a Pawn in the center, opens a diagonal for the Bishop on e1, and prepares a shelter for the Knight on d3. It also shields the Queen from the Bishop on the long a1-h8 diagonal. Also worth considering is 1.f4, which opens the other diagonal for the Be1, prepares an open file for the castled Rook on f1, and plans to build a Pawn center with e2-e4. The move 1.c4 also looks playable, hoping to play d4 on the next move and preparing to bring the Knight to c3.
All things considered, I would play the solid 1.O-O against a higher rated player, giving me one move to see how he assesses the position. I would play the looser 1.d4 against an equal or lower rated player. Nothing dictates that the same first move is best in all circumstances.
SP395 QRNNBKRB
To choose White's first move in the game where I am playing Black, let's look at the list in HarryO's 'Logical First Move' post. He leaves a choice of eight moves, including two that I also like for White, 1.c4 and 1.d4. Of the remaining moves, the two Knight moves, 1.Nc3 and 1.Ne3, aren't something I would play, but they both prepare Nd5 with an attack on the weak c-Pawn. If this happens, Black can defend adequately with ...Ne6, leaving a large number of other moves to play first. Both Knight moves leave White with the natural initiative of the first move, so they are good moves for White.
Are there any first moves for White that don't keep the initiative? Both 1.d3 and 1.e3 look suspicious, because they are passive and put no pressure on Black. The move 1.d3 at least opens a diagonal for a Bishop, but 1.e3 does nothing to further the development of any pieces. The move 1.e3 even allows Black the same range of options as White had on the first move. I could continue and choose a first move for Black, but that goes beyond the objective of this exercise.
To summarize, either 1.d4 or 1.O-O, according to circumstances, would be my first move, and 1.e3 would be the first move for my hypothetical opponent. I get good play in both games.
As interesting as HarryO's idea is, I believe it has a fatal flaw, which I stated in a comment to that SP395 post:-
I propose we play a two-game match with one of these SPs that you've flagged, where we each take White in one game. In the game where you are White, I will select your first move from your list of viable moves. In the game where I am White, I will choose my first move myself.
Let's return to SP491 and choose a pair of moves for the hypothetical match. I'll begin with the game where I'm playing White.
When I look at a new SP for the first time, I start with the castling choices. In SP491, I see no reason to favor one of the castling options over the other. Castling O-O-O will be possible after the Nc1 moves, while castling O-O requires the Be1 and the Nf1 to move off the back rank.
I adhere to classical opening principles, especially the one about paying attention to the center. For this reason, the moves 1.b4 and 1.g4 don't appeal to me, even though it is often attractive to advance by two ranks the Pawn directly in front of a Rook. On top of ignoring the center, both moves preclude future castling to that wing.
I also don't like developing Knights until I get some idea where they will be most active. In the given position it is too early to decide about the Knights. On top of that, the moves 1.Nd3 and 1.Ne3 leave the Knight exposed to harrassment by the Black center Pawns.
After a few minutes thought, I decided that 1.d4 is my preferred move in this position. It prepares Nc1-Nd3, where the Knight is both centralized and protected, and where castling O-O-O is already possible. As a bonus, the castled Rook would be placed on a partially open file. The move 1.d4 also opens a diagonal for the Bishop on e1, bringing White one step closer to castling O-O.
The downside of 1.d4 is that the Pawn is exposed to attack by the Bishop on h8 and by ...Ne6. After playing with the position a few more minutes, I decided that both of these attacks can be defended in various ways.
SP491 QRNKBNRB
As for choosing a first move for my opponent, my eye immediately jumped to the 12th possibility on the list of viable SP491 moves: 1.h4. HarryO justified this move with the comment, 'it takes control of g5, threatens h5, and allows Nh2/f3', but none of those reasons is particularly compelling. The downside of 1.h4 is that it ignores the center, does nothing to develop a piece, and signals that White will not be castling O-O. Each of those disadvantages is enough to reject the move and taken together they brand 1.h4 as positionally weak.
A skilled chess960 player like GM Nakamura might whip up some serious threats with 1.h4, but he knows more about chess than the average player does. So 1.d4 for me and 1.h4 for my hypothetical opponent, and I'm off to a good start.
Discovered a start position SP today that has a ridiculous amount of viable first moves options! More than standard chess in fact, 15 in total. The absolute maximum number of first moves in some Chess960 starts is 21, but usually a large number of those can be [written] off because they are possible but theoretically too weak.
He then lists the 15 moves and reveals the methodology used to determine that number.
After a depth-24 search, Houdini-3 thinks that the variation in score between the best and worst first moves in that list is +/- 0.1 which is tiny. [...] In contrast, standard chess has at the very most 13 viable first moves but the variation in score is much bigger at +/- 0.43.
In other words, 1) Run your engine to a fixed depth on a certain SP, and 2) Record how many top moves have a similar evaluation. I imagine that different engines produce different results and that the results for one engine across all 960 SPs would probably tell us something new about the entire set of SPs.
SP491 QRNKBNRB
A few months ago I did a similar experiment that I recorded in a post titled First Move Diversity in Chess960. Working with a sample of real games between real players, most of them less than master strength, I calculated how many different first moves had been played with each SP. My sample was small by chess960 standards, so I'm not convinced the results showed anything conclusive. If the sample had been an order of magnitude larger, I could have also counted responses to the first moves.
For SP491 QRNKBNRB, I discovered that five different first moves had been played (shown here with the number of times each move was played):-
5: 1.g3
4: 1.g4
3: 1.d4
1: 1.Nd3
1: 1.Ng3
Note that the last move in this list, 1.Ng3, isn't even one of the 15 moves on HarryO's list. What move would I play in SP491? That's an exercise that I'll save for another post. In the meantime, it's also worth noting that SP491 was the position that appeared in The First Recorded Fischerandom Game, GM Bronstein - IM Douven 1996. Bronstein, who missed becoming World Champion by the narrowest of margins, played 1.d4.
The first column in the table, covering the a/h-files, represents one of the controversial aspects of chess960 : pieces that start in the corner squares. A corner square, having only three adjacent squares, offers fewer opportunities for development than the other back rank squares, which all have five adjacent squares.
A Rook in a corner square is not controversial. That is, after all, the setup used in traditional chess. Starting in the corner, a Rook doesn't interfere with the development of the minor pieces. It just waits patiently until an open file appears on the board, then heads for that file like a boat heads for open water. The first move of a Rook is nearly always along its back rank.
Once in a while a Rook is lifted to its second rank, where it gets active play laterally, but a lift to the third rank is rare. There is a stigma attached to the early two-step move of an a/h-Pawn, making that move psychologically difficult to play. Once in a great while, the blocking Pawn disappears on a capture or an exchange to its adjacent file, opening a file for the Rook immediately.
A Knight in the corner is not particularly controversial. It has only two choices for development, where the third rank on the adjacent file is often the natural move. Going to the second rank two files away requires another move to bring the Knight into full play. This is sometimes an attractive choice when the Knight continues to a central square, like the sequence Na1-c2-e3.
A Bishop in the corner is the most controversial corner placement. The first move of the Bishop can only be on its long diagonal, which requires moving the Pawn on the adjacent file. Even if this limitation can be considered a disadvantage, the Bishop is ready for action after a single Pawn move, which is a distinct advantage.
That leaves the case of a Queen in the corner. There are three ways to activate the strongest piece. I gave examples of diagonal development in a revent post, Opening Queen Swap in the Corner, although the early Queen exchange seen there is not typical.
The Queen can also be developed along its file, by moving the blocking Pawn one or two squares. Another recent post, Activating the Corner Queen, gave an example of a Queen developed vertically. This occurs more frequently than for the Rook, because the Queen obtains a new diagonal for its second move, thereby increasing its activity.
The most popular development of the Rook -- laterally along the first rank -- is perhaps the least popular for the Queen. It requires too many moves to get the Queen into the game, although it might be interesting to collect examples of this to see the circumstances under which it occurs.
There is a choice for any player who has trouble with these many possibilities for the corner square. Stay with the traditional start position. The corner Rook maneuvers are well known and there is little need for further thinking.
I touched on the topic of the corner Queen a few years ago in 'Fianchetto the Light Squared Bishop', where I quoted GM Seirawan saying, 'Unfortunately, because the Queens are buried in the corner, I don't see real tactics breaking out in the first dozen moves or so.' I don't believe that such positions are 'unfortunate'. There is no reason why fireworks need to occur immediately and good results often come to the player who is patient.
My previous post, Opening Queen Swap in the Corner, discussed a treatment of the corner Queen which doesn't happen often. I recently played another game which had a more normal treatment. It started with SP180 NBBRNKRQ and I had the Black pieces. Note that castling O-O is possible on the first move.
The first moves were 1.c4 c5 2.Nb3 Nb6 3.Nxc5 Nxc4 4.b3, reaching the position in the top diagram. Retreating the Knight didn't appeal to me, so I decided to play 4...b6, continuing the symmetry and offering a Pawn sacrifice. In my notes to the game, I wrote, 'Not sure if this is sound, but it's interesting!', a sentiment which arises frequently when I evaluate chess960 opening moves.
The critical line is the desperado continuation 5.Nxd7+ Bxd7 6.bxc4 Nd6 7.d3 Bc6, threatening the c-Pawn and keeping the initiative with Black. My opponent played instead 5.bxc4, and the game continued 5...bxc5 6.Bc2 O-O 7.Bb2 Bb7 8.O-O, reaching the bottom diagram.
In this position I felt that Black had a slight initiative. How is this possible? In fact, the move 6.Bc2 wasted a tempo because the Bishop is not posted any better on c2 than on b1. The b1-h7 diagonal is far more important than the a4-d1 diagonal. This observation helped to answer the key question: how should I develop the Queen. Since the Queen has no aggressive prospects on the a1-h8 diagonal, it can only come into immediate play via h6.
With this in mind I played 8...h5, reasoning that if White played similarly, Black's activity would occur first. The weakening of Black's Kingside (h-side for the purists) is secondary, because the h-Pawn can't be attacked. It might even play an aggressive role by advancing to h4.
Instead of 9.h4, White shielded the Queen on the long diagonal with 9.Nf3. I continued with the plan from the previous move and played 9...Qh6. After 10.d3, the time was right for 10...d5, opening the game. Black's Queen can transfer in one move to the Queenside (a-side), thereby supporting any tactical variations with the most powerful piece on the board. Meanwhile, White's Queen will languish in the corner.
Although the game lasted another 50 moves, Black was always in command and eventually won. The early activation of Black's Queen was the deciding factor.
The game was my first playing for a team, conducted under the system described in Parallel Games, two games against the same opponent with the same start position. Much to my dismay, the team was already losing 0-2 a short time into the match when our last board lost both his games by blitzing his moves. 'Overconfidence' was his excuse.
I won my game with White while the other team members played their games to an even result, so the team was one point down with only my game as Black remaining. The game started as SP861 RKNBRNBQ, a Queen-in-the-corner position. The first moves were 1.f4 f5 2.Nd3, reaching the position shown in the top diagram. Here I played 2...g5, planning the continuation 3.fxg5 e5, where Black has excellent play for the Pawn.
My opponent avoided this and the game continued 3.g3 gxf4 4.gxf4 c6 5.e4 fxe4 6.Qxe4 Nd6 7.Qe5, reaching the bottom diagram. White is hoping for an exchange of Queens on e5, repairing his Pawn structure and retaining the initiative.
Black can't avoid the Queen swap, but he can dictate the terms. I played the unusual idea 7...Ng6, when White has nothing better than 8.Qxh8 Nxh8. Black is not worse and has good play against the isolated f-Pawn. The game continued 9.Bg4 e6 10.Ng3 Kc7.
White launched a Queenside attack with the a-Pawn and castled O-O. Black repulsed the attack and launched a counterattack on White's weak Kingside, eventually winning a Bishop for a Knight. A long period of maneuvering set in and I finally converted my positional advantages into an extra Pawn. I wrote about the endgame in a post on my main blog, Simple Positions, Pretty Geometries.
The league's organizer, in a summary of the match, noted,
The cliffhanger of Section B delivered the teams New England [my team] and Team XLink who shared the victory with a result of 5-5. Funnily enough no draws occurred. Absolutely well worth seeing was the clash of two chess960 heavyweights at the first board. Mark with Black managed to defeat Carl in a model correspondence game.
It's always nice to win, but it's even nicer when the game has special value.
