Last night I was thinking: why should I continue to play or try to invent chess variants? Aside from entertainment value, and to escape from standard chess, while still being involved with board game(s) like it, the answer seemed to boil down to the quest to find or invent the Next Chess. First, I will speculate that such a golden variant will not replace chess for 100-500 years, i.e. not in any of our lifetimes. So, finding/inventing the Next Chess would be a generational project, realistically it seems. Here at CVP website, alone, we may leave our own contributions (our games played and/or invented, and opinions about what the Next Chess should be, or include/exclude in the way of characteristics) and just hope one day when a Next Chess is sought for, due to real need in the far future, our two cents worth might help our decendants decide on the beast, rather than make a less informed choice, perhaps in haste.

After trying to define the parameters of what the Next Chess ought to include (or exclude), I must say that I concluded tentatively that there is not much left to invent that fits in well with the boundaries I came up with - perhaps not so shocking when one considers there's 5000+ variants on this website, with 1200+ presets on Game Courier alone (of the latter I even remotely like around 10%, which fits in with Sturgeon's Law, i.e. that 90% of anything is crap).

What are the parameters for the Next Chess I came up with? Well, I didn't need to convince myself of too many before it looked like the possibilities left might be rather confined. First of all, chess, the most successful variant, is played widely and most seriously in physical tournament halls. You cannot expect it to go mainly online anytime soon, if ever. So, 3D, 4D+ variants are all ruled out, as a physical board needs to be convenient. Circular boards aren't too bad for the size of the cells, for decent physical size pieces as well, if 4x16 or even 5x16 is used. The problem is, another conclusion I drew, that most everyone (especially if one wants to include children) would want to include rooks and/or bishops as the primary slider pieces to have in any sort of variant, and on circular boards, a single rook, powerful as it is, is incapable of mating a lone K normally, which is hardly a step forward for anyone graduating from normal chess - looking ahead, this is an argument against e.g. Omega Chess, with its extra corner squares [edit: which also effectively act to make the variant practically 12x12 in terms of taking up physical space, and for the physical sizes of cells and pieces]. Hexagonal boards so far have not proved popular, as Glinski's never took off, and at 91 cells the size of the cells and pieces is rather reduced, too, for a typical set I'd think.

That leaves 2D square or rectangular boards, which must be at least 8x8 in size, or at most 12x10, if a board is to fit fairly nicely on a coffeetable or on a table in a tournament hall. Any bigger, and the physical squares and pieces become unattractively small, arguably (it's hardly a step forward for those graduating from standard chess sets, at any rate). Also, well played games could prove too many moves long on average, on such big boards. Anything smaller than 8x8 will likely prove to make for games that are too short in length.

For games that aren't too long, 32-48 pieces in total would be ideal, IMO, both to pack up the pieces, and so that after all the exchanges in a game, it finishes at a reasonable move number (say pieces are traded at the rate of a pair every 5 moves, as in chess, and an average game of that ends when about 16 units are left on the board). 60-72 pieces in total might be tolerable, and 56 for a total has at least two historic precedents.

A nice pieces/empty cells ratio for the setup is 50% as in chess, though it's hard to prove this, and this is a more flexible guideline I'd say. 10x10  is an otherwise nice board size that has trouble meeting this ratio smoothly. That's aside from the fact that board-lengths greater than 8 cause pawns to take longer to promote, if no special rules or pawns are used. Betza has expressed respect for the ancients who invented the chess pawn type, and I'd consider it the best pawn type as a rule, IMHO. Pawns being on the 2nd rank in a setup are ideal, but on the 3rd rank is not so bad for the pawn-line either, except we lose the sweet possibilities of smothered and back rank mates that come with chess' setup and rules.

As for the pieces, for the sake of the children's participation alone, symmetric moving pieces are best as in chess, though divergence might be allowed (i.e. capture differently than move). For a given variant, I agree with Fergus that at least some of the piece types should be close in value, so as to make for more interesting trades, and perhaps a greater range of opening possibilities. With Rs and or Bs being desirable, along with at least one piece with queen-like strength, all the above rather limits the piece types that can be included in the Next Chess, no matter the board size (if rectangular or square).

Other parameters would be that the Next Chess should not be too drawish, and that the pawns all be guarded in the setup, to try to minimize White's early initiative. I'd say drops are double-edged, if allowed, making a game more tactical, and less of a blend of strategy and tactics like western chess. The pieces should continue to be nice figurines, and there should be a fixed setup, for people's coffeetables, merchandizing and movie scenes alone.

As for computer resistence as a feature for the Next Chess, I'd pretty well give up on that for any board game; AlphaGo put in the last nail as far as my hopes went, especially should AI keep going forward in big strides in our future.

[edit: 18 April 2018: Other parameters would be if there's queens in a game, start with one per side. Also, have no bishop's adjustment rule or multi-path piece types or Ultima-style ones, for simplicity for the sake of children's participation. Having the king as the one royal piece and pawns that promote as in chess (i.e. last rank, any piece type) also seems tradditional, logical and wise. Lame pieces don't seem particular attractive, with chessplayers graduading from nimble knights. Any element of chance should be avoided, as per a strength of chess. Neutral piece types seem best to avoid. If compound pieces are used, ones with R power should not be placed in the setup so that they can develop symmetrically then swap quickly. A near-symmetrical setup of some sort is always a good idea. Also good is to avoid extra off-board squares and/or pieces, for the sake of streamlining, e.g. for decorative sets on coffee tables or perhaps even clubs and tournament halls. Ideally all the ranks in each army should have no empty cells, although a minor deviation from this that looks rather nice may be okay.]

[edit: 18 April 2018: Based on all the above criteria being applied stringently, I came up with just 5 'Next Chess' candidates (after an initial pick of 41 from my massive but fairly choosey favorite games list) - disappointing since so few, and since only one isn't one of my own inventions (the 4 of mine all had these stringent criteria in mind in their design, though): 10x10 Shako (the only one of the 5 not mine), 10x10 Sac Chess, 10x8 Hannibal Chess, 10x8 Frog Chess and 12x10 Wide Nightrider Chess.]