If you’ve never heard of Chess960, I can hardly blame you. It is a somewhat obscure variant. I won’t replicate the Wikipedia article here. Suffice it to say, the idea came from the famous grand master Bobby Fischer. In order to get to a high level of chess play, you have to memorize a large number of openings. Some find this rote memorization to have a negative effect on creativity and skill. “Fischer’s goal was to eliminate what he considered the complete dominance of openings preparation in classical chess, replacing it with creativity and talent.” To facilitate this, Fischer proposed randomizing the back row of pieces.
I am not a chess player, but I was immediately fascinated by the idea of Chess960 when I first read about it, and felt the concept would be very applicable to NXS. In fact, NXS is far more suited to a random starting setup than Chess (in fact I included the loose idea of a random opening position as an official variant from the beginning). The nature of Chess dictates that you must have a front row of Pawns, you should have one Bishop on each color square, the King should be between the Rooks so that he may castle, etc. NXS has far fewer constraints. In reality, NXS has no constraints as to piece placement. But as I began experimenting with the concept, I found that practicality dictates a few small concessions.
The following assumes we are discussing a medium game setup between two players. The rotation of the pieces are not randomized, the standard front facing is used in all cases.
There are 14 pieces in NXS:
- 3 @ Merchant (3-0)
- 3 @ Cutter (3-2)
- 2 @ Man-o-War (5-4)
- 2 @ Frigate (4-3)
- 2 @ Corvette (4-2)
- 2 @ Brigantine (4-2s)
As discussed previously, two-player games of NXS are best with a single Merchant per player. That brings us to 12 pieces. However, that one Merchant is critical to the unique character of a game of NXS, so I want to ensure that each player always gets one. We’ll set it aside. This bring us to 11 pieces, of which 9 are randomly chosen. The Merchant is added back in, and (initially) the positions of all 10 pieces (within the standard board layout pattern) were chosen at random.
The first game I played against an opponent resulted in an interesting conundrum. A single Man-o-War was (randomly) placed in the back rank. This resulted in a very unsatisfying situation. This piece is (by design) so very slow. Both my opponent and I hardly used the piece. In the standard medium game setup, the Man-o-War is purposely placed front and center (the b1 hex). Here, it is a very useful piece for area control.
And so, much like in Chess960, a little order in the randomness makes for a better result.
My current implementation is to remove all Merchant and Man-o-War pieces. Eight of the remaining nine pieces are chosen at random. A single Man-o-War is placed at b1. A single Merchant is added, and the positions of these nine pieces are randomized and placed into the standard medium game layout. Obviously, this is more easily accomplished with a computer, but you can also get the same effect at home by randomly drawing slips of paper with the pieces names written on them.
As the name implies, Chess960 has 960 possible starting positions. Calculating permutations is not exactly my strong suit, but if I have it correct, I believe that NXS (following the given parameters) has 68,040 possible starting positions.
But calling it NXS-68040 seems a step to far, no?
Addendum: I don’t plan to implement the short game random setup on Board Game Arena. But if you wanted to play at home, I would recommend taking out all but one Merchant, both Man-o-War pieces, and one Cutter. Then randomly select seven of the remaining pieces. This gives 10,080 possible starting setups. I’m not going to try to calculate the long game. I would recommend placing the two Man-o-War pieces at d15 and d5, taking out two Merchants, and randomizing the remaining pieces.
the NXS nexus 