Key concepts
The Generic Approach
Translation & Transposition
Internal Balance
Strategical & Tactical Games
Essential Games

During my career as an inventor, I developed a number of 'mental tools' for direction and a number of definitions for clarity.

  • Objects
    The classification I employed is arbitrary. It served me well in exploring different alleys of thought, but it may at the same time have prevented me from 'thinking outside the box'. Fortunately it's a rather large box. Here it is:

    • Chess variants
    • Elimination games
    • Territory games
    • Connection games
    • Breakthrough & Race games
    • Configuration games
    • Mancala games

    The first two are 'existential' in that the object is to eliminate either the 'heart', as in chess, or 'body & limb' as in draughts. The other themes are 'co-existential' in that the object is to grab more, or connect sooner or to be faster, without necessarily cutting the other player's throat.

  • The Generic Approach
    Ask a Chess player about chess variants, and he'll probably envision a variant of Chess, not a game on the theme of 'checkmate'. That's because players tend to think specific rather than generic. That's all right for a player, but deadly for an inventor.
    I will distinguish accordingly: Shakti - one king and two pieces per side - is a chess variant, while Grand Chess is a variant of Chess.

    Being a musician or a builder of musical instruments requires different qualities. So does being a player or an inventor of games. There are some rather painful examples of players playing inventors. José Raúl Capablanca and Edward Lasker may be excused for getting it not quite right with Capablanca Chess, but Emanuel Lasker made a serious mistake that no inventor should make and thus Lasca ended up being a pretty bad game.
    And with his totally 'specific' mindset, Bobby Fischer messed up big time with Chess960.

    Consider this: there are two 'weird' moves in Chess, en passant capture and castling. The first follows the introduction of the initial double pawn move. That allows a pawn to conditionally avoid confrontation with an opponent's pawn on an adjacent file, a problem one felt should be addressed, whatever the solution. And the solution we now know and accept was implemented.
    The second arises from the difficulty of developing the rooks, because the king's safety can hardly be the argument in a chess game.
    Along comes Bobby Fischer, bored with his perceived supremacy in Chess. He invents a new variant that will do justice to his curtailed talents. A variant using a random set-up of the pieces behind a regular pawn placement. Rooks may still be tucked away, but most of the time they're not. No need for a castling rule, which is a means to an end, not of course an end in itself.
    That is: to an inventor. Not however to the totally specific mindset of Bobby Fischer, the only one qualified to improve on God's work: Chess as it is. Thus the first sentence of the wiki entry reads: "Chess960, or Fischer Random Chess, is a chess variant invented by the late former World Champion Bobby Fischer by modifying the rules of Shuffle Chess so that castling possibilities exist for all starting positions".
    Clearly Fischer perceives castling as an end in itself. As if having a weird move is somehow a prerequisite.
    Even more ridiculous is the game's large following of misguided players who obviously think brilliant players make brilliant inventors.

  • Grids
    Talking boardgames is talking grids. Since this is not a treatise on grids, I'll be short, personal and far from complete. For the larger part, I've employed only the square- and the hexgrid. Lotus is an exception, so if you happen to have an old Kensington board, you can put it to a better use. Rondo is another exception.
    Two other games, Medusa and MacBeth, employ a subset of the hexgrid, while the Glass Bead Game has no actual grid at all. All other games I invented use either the square grid, checkered or not as the case may be, or the plain hexgrid.

    If we consider a grid as consisting of cells and vertices, it is soon apparent that these are dual concepts. Put a point in the middle of each cell of a square grid, connect the points and you'll have another square grid. Put a point in the middle of each cell of a hexgrid, connect the points and you'll have a triangular grid, and vice versa.
    It means you can play Chess on the vertices instead of the cells, as in Xiangqi. You'll lose the checkering, but otherwise things are quite the same.
    It means you can play Hex or Havannah on the vertices of a triangular grid instead of the cells of a hexgrid, or Go on the cells instead of the vertices. It's only a matter of taste or tradition.

    Diagonal grids
    A square board has two diagonal grids, the bishop domains. A hexgrid has three, that's why there are three bishops in Glinski's HexChess.
    A diagonal grid of a square grid is again a square grid, that of a hexgrid is again a hexgrid. To illustrate this, consider the following transformation:
    diagonal representation
    orthogonal representation

    The position is a Coup Raphaël that marks the beginning of my childhood fascination with 10x10 Draughts.
    But the point is: representation is only a matter of taste or tradition. Here tradition and my personal taste co-incide. So if a game is aligned with the diagonal grid, I've chosen that representation.
    Bushka and Emergo are examples, while Dameo is aligned with the ortogonal grid in the first place.

    Apart from their funcionality, I want boards to be as non-distractive as possible. It's the content of the game that matters.
    This Lasca board features circles, instead of squares. Very neat indeed, and we've used similar Emergo boards. Yet we've chosen for a traditional representation in MindSports: it's the content that matters.

    Where a square diagonal sub-grid constitutes half the board, a hexagonal diagonal sub-grid constitutes only one third. So here was this guy so blinded by 'specific' thinking that he made a hexagonal draughts game that was played on one of the three diagonal sub-grids of a hexboard, effectively using trice as many cells as needed. Laugh if you will, but these things happen. Anyone can be a 'games inventor' just as anyone who has thoughts can claim to actually 'think'.
    Of course Hexdame employs a plain hexgrid.

  • Translation & Transposition
    I use 'translation' when a game is to be adapted to another grid. Thus Glinski's HexChess is an as close as possible translation of Chess to the hexgrid and Hexdame is a literal translation of International Draughts.
    Go has been translated to the hexgrid in several ways. 'Hexagonal Go', a literal translation to the cells of a hexgrid, has never been a great success. Too many liberties. Mark Berger's 'Rosette', a translation to the vertices of a hexgrid, suffered the reverse problem, but Mark solved that brilliantly by introducing a 'rosette', a small hexagon completely occupied by one color, as an additional safety mechanism. Medusa is another hexagonal translation, less than literal this time.
    Actually Go can be translated to almost any grid, but all translations seem to fall short of the original - and not by a narrow margin either. Go is a very square game.

    macbethIt may sometimes be worthwhile to translate a game mechanism to different grids and I made it a standard routine. In the case of Hexdame it went smooth as honey, but the game has a totally different character than original square game. Translating Othello immediately reveals the need to abandon 'diagonal' capture: that would mean capture in twelve directions instead of eight. Allowing only 'orthogonal' capture reduces it to six directions, but then a surprise happens: put three black and three white stones on alternating cells around the centercell of a hexboard, and start playing. The stones will 'carve out' a hexboard with holes.

    So although capture in MacBeth is generally speaking in 'six directions', any particular infield cell allows capture in only four. These things happen. Translations may demand rule adaptions, and even if not, may render very different games.
    Also, don't expect miracles: if at all possible, translations are seldom smooth, and if they are, the resulting games may still lack appeal.
    To give an example of the latter, I personally feel that Chess too is a very square game, and there exist translations of almost all of my chess games, HexChad, HexShakti, HexCaïssa, HexDragonfly, HexLoonybird, you name it, but I discarded them all.

    The translation of Emergo went flawlessly, but Hexemergo turned out to be flawed. Shit happens.

    Some games may be impossible to translate. You may apply the idea of connecting opposite sides to a square, and come up with intersting games like Twixt or Gonnect, but you cannot translate 'Hex' to the square grid. Nor Havannah for that matter.
    A fellow inventor, Mark Steere, made Atoll, a Hex translation to a square board, but not to a square grid.
    I use 'transposition' where an idea or a mechanism is put into a different framework or context.

    • Lasker put the 'colomn capture' mechanism into a Checkers framework. The result was less than fortunate, but that's beside the point.
    • I narrowed down 'capture by appoach' as featured in Fanorona, and put it in an International Draughts framework. A literal transposition proved less than satisfactory, but after I also transposed the idea of 'linear movement', Bushka emerged as a great game.
    • Now that I've mentioned it: 'linear movement' was transposed from a game called Epaminondas and modified to become the backbone of Bushka and Dameo.
    • Apply the idea of 'different valued pieces' to a mancala, and you may end up with the Glass Bead Game (which of course would be unfortunate, since it already exists).
    • Introduce the Shogi-like dropping of pieces in western Chess, and you might end up with Neo Chess (aka Mad Mate), if you're 'specific' in your approach, or Dragonfly, if you're more 'generic'.
    • Use the mechanism of placement and capture in Go, for a different objective, and invent Gonnect, as João Pedro Neto did.

    The possibilities are infinite. It's almost a game in itself.

  • Simplicity, Occam's Razor & Minimalizing Choice
    Simplicity starts with bringing mechanisms down to basics and asking the right questions. Almost everyone is familiar with the capture mechanism in checkers type games. Basically it is a 'two men on three cells' scenario. The man on the first cell leaps his opponent on the second and lands on the third, removing the opponent.
    Notice that I'm talking 'cells', not 'squares': the hexagonal grid supports this mechanism of capture too, as Hexdame shows. Nor am I interested, for now, in whether it takes place in the 'diagonal plane', as in Checkers and Draughts, or in the 'orthogonal plane', as in Turkish Draughts and Dameo. Nor in the direction of movement, the direction of capture, whether or not multiple captures are allowed, and if, whether or not majority capture precedes. Nor in whether or not promotion is possible, and if, whether or not the king should be a long range piece.
    These are all 'specifics' that may lead to vastly different implementations. If you're interested, please have a look at Draughts Dissected, which is a comparative investigation into draughts variants, or at On the Evolution of Draughts Variants. But they're not the first things to consider: choices will present themselves eventually and should preferably not be made a priori.

    What I will note a priori, is that this basic mechanism of capture has given rise to a vast number of great games. For me that leads to a simple question: given this richness, are there similar basic mechanisms of capture? If so, are some great games hidden there? And the answer is yes. There are at least two more mechanisms of the 'two men on three cells' type, and both have rendered at least one great game.

    checkersIf you put the cursor on the pict on the left, you'll see the checkers type mechanism that has become the dominant form of capture in draughts type games around the world.

    emergoHere is the 'column checkers' mechanism, as employed in Bashne, Lasca, Stapeldammen and Emergo. The first two implementations were subject to a series of 'a priori' choices that were less than fortunate. The inventors failed to ask the right questions and obviously had too little faith in the mechanism to consider what it really wanted.

    bushkaThis mechanism is called 'capture by contact'. An extended form of it governs Fanorona, the national game of Madagascar. Its implementation in Bushka almost equals 10x10 Draughts as a 'mental sport weapon', with a smaller margin of draws as a bonus. Because that's where Draughts eventually falters, not as a game, but as a sport weapon.

    It's a well known fact that the greatest imaginable complexity may derive from the simplest mechanisms and rules. Games like Go, Hex and Symple bear testimony to that. It is therefore important, when considering a mechanism, to avoid preconceived ideas. Let the mechanism do the talking (beware, it whispers, listen very carefully) and don't introduce new ideas unless needed. This is called 'Occam's Razor', an instrument that should always be part of an inventor's toolkit. If applied to the checkers mechanism, it's not all that hard to 'reinvent' Checkers or Turkish Draughts.
    I will 'reinvent' Bushka in the section Organic mechanisms and show how it almost explained itself in the process of its invention. It's a two way street: you can't force a game into a jacket it doesn't want. A game like Lasca bears testimony to that.

  • Consistency
    Lasca is a good example of an inconsistent game. It's origins lie probably is some pub in 19th century Russia, where Shashki was the dominant proletarian game, an 8x8 draughts variant reminiscent of International Draughts, but without precedence of majority capture and with a 'flying king', meaning that a man reaching the back rank in a capture immediately transforms to king and, if further captures are possible, must proceed as such in the same move.
    This to sketch column checkers cradle.

    At some point someone got the basic idea of column checkers: what if, instead of removing a captured man, we take it along as a prisoner underneath the captor? That was a brilliant question, and I'm the first to understand that the idea was transposed to Shashki as it was. That's specific rather than generic thinking, but you'd have to see it 'work' to get some first impressions of the implications.
    So column checkers started out in a draughts jacket and it worked, sort of.

    • It was soon established that only the top man of a column should be captured, and that bi-colored columns always had one color on top and the other underneath, but never a color 'in between' the opponent's men.
    • Obviously a single man could make a multiple capture, thus capturing several prisoners. It was soon realized that this was dangerous: after such a man was captured, a strong column of the opponent would be released. Strong because it had to be jumped several times before being eliminated, especially strong because it would make prisoners in that very process, and prisoners can't assist in the process of their own liberation.
    • Nor did it take long to realize that it gave rise to a tactics called 'feeding': force a single opponent's piece to make several captures, while making sure the guard can be captured afterwards, releasing the 'fed' pieces as one strong column.
    • Also, it was noted that the capture of a single man reduced the number of pieces by one, but that no increase of the number of pieces was possible. Hence games kept 'spiraling up' in ever fewer but ever higher columns.
    • Finally it was noted that a piece had a 'life cycle', always starting out at its strongest, but transforming by involvement in captures into an increasing number of prisoners, held by a decreasing number of guards. This eventually makes it a liability, and in that case solid protection or timely sacrifice might prove necessary, lest the piece should be fed even more prisoners before its last guard faced capture.

    Fascinating stuff really. But problems emerged in practical play. Shashki features an initial position, a forward direction and promotion. So Bashne, as the newborn was christened, featured an initial position, a forward direction and promotion. And that's where the trouble started.

    It was established that only the top man of a column reaching the back rank would be promoted. Shashki features a long range flying king, so Bashne got a long range flying king. In consequence of its being long range and flying, the king not only captures easily, but it can also easily be fed. Being able to capture long range is all very fine in Shashki, but in its descendant it is a double edged sword to say the least. Promotion is meant to be a reward, but in Bashne it may become a liability: you end up capturing a lot of men under a vulnarable piece that, if captured, will release its prisoners as one strong piece for the opponent. This must have crossed Lasker's mind when he decided to transpose the mechanism to Checkers instead of Bashne. Checkers' king is neither long range nor flying.
    Transposition to a Checkers framework removed that particular suspicion of inconsistency, but a next one didn't take long to become apparent. Kings had to be marked as such, because by the nature of the game you can't use two stacked men. So you put a mark on one face of each man, while leaving the flipside unmarked, no problem. But kings can also be captured and disappear in a column, so if they're to be identified as such, they had to be marked on the side as well. Admittetly a small problem. But the real problem is structural, at least from an aesthetical point of view.

    • A king may be captured by a piece that already carries prisoners. By the nature of the mechanism such a king starts at the bottom, with like colored men above it. If and when released, the liberated column has one or more men on top, with the king in between or at the bottom.
      What a sight, soldiers imprisoning their own commander! The poor guy has to wait till the men above it are captured, before it can take command of the column.
      This problem persisted even after Lasker came along and transposed the mechanism to the checkers framework with its short leaping non flying king. To a degree it got rid of the liabilities concerning promotion, but it didn't solve the 'hampered king' problem at all. It's still there and players seem to accept it as something inherent in the mechanism. They even tend to defend the bug as a 'feature'.
      Doubtlessly one can have fun playing Bashne and Lasca - I know that from personal experience. But that doesn't make them good games. Good games don't hamper themselves and had Lasker been as good an inventor as he was a Chess player, he would have concluded that he hadn't been listening closely enough. And listening closely he would have realized that promotion was not an inherent part of the mechanism, but a result of the 'checkers framework' !
      So I will also 'reinvent' Emergo in the section Organic mechanisms and show how it completely explained itself in the process its 'invention', which was a joint venture with Ed van Zon.

  • Internal Balance
    The games I invented are all symmetrical in that both players have the same pieces and the same objective. Therefore the external balance is implicitly guaranteed. Though non-chess games can be excessively volatile, say Fanorona or Explocus, they seldom show internal imbalance. It's games with different kinds of pieces that may be susceptible to it, and that means primarily chess games.

    Imagine the queen in Chess would have an additional option to move as a knight. It then, if put in the center, would cover all squares of a 5x5 board. Most players would agree that this wouldn't constitute an improvement. The piece is simply off the scale in comparison to a mere rook or bishop. This I call internal imbalance. You can't blame the people who buy such a game. They can't see its flaw without playing it. But I do blame inventors because they should know better, and I do think manufacturers should care more about content and look further than the next shopping season.

    In the past, the market has provided some good examples. One such culprit is called Ploy and was invented by Frank Thibault. Eventually it disappeared of course. You can fool all people sometimes, and some people all the time, but you can't fool all people all the time.

    Ploy is a chess variant with 'rotational' pieces and no pawns. Each player has one 'commander, six 'lances', five 'probes' and three 'shields'. These pieces have arrows on top indicating the directions in which they may move. Instead of moving a piece may rotate to a different direction. Shields may move and rotate (in that order) in the same turn.
    Pieces have a limited scope: they may move as many spaces as they have directions to move in. That's the first indication something may go wrong. Coupling 'strength' with 'scope' is positive feedback. Positive feedback creates imbalance.

    It's not necessarily wrong. In Focus 'scope' is coupled with a column's height. But in Focus the column must move precisely as far as it is high, not 'up to and including'. This restricts positive feedback. Focus is quite volatile, but not imbalanced.

    Frank Thibault however was so blatantly unaware that he was going down the wrong track, that he actually increased the dubious effects of his coupling by allowing six lances, able to move up to three squares in three directions, but only three shields, able to move one square in one direction, albeit with the right to rotate. What the hell are these guys doing there? It's like having six pitbulls to fight with, and, to make sure, three pekingese.
    So I did a better job with Rotary. And it wasn't too difficult either. I needed a couple of minutes. A complete and consistent set presented itself so emphatically that it would have been hard to miss. The mechanism by nature is very capricious, so I transposed regular Chess pawns to provide a strategical framework. Rotary may not be a great game, but it's certainly a good game.

  • Strategical & Tactical Games
    On our homepage one can read:
    "We're more committed to strategy games than to tactical ones. Here's the difference:
    Strategy games have strategies varied enough to allow different styles of play, tactics varied enough to induce their own terminology, and a structure that allows advantageous sub-goals to be achieved as calculable signposts along the way. Tactical games have strategies that are either fairly obvious (however deep), like Pente, or fairly obscure, like Othello".

    It's not a clear-cut division. Hex and Symple are pure strategy games, with less than obvious tactics. Havannah is a strategy game, but features a number of obvious tactics that beginners may appreciate even without a deeper understanding of strategy. Hexade has a similar structure, but it is a tactical game with a basically simple strategy. Emergo can be argued from both sides.
    There's no value judgement involved either, more a matter of taste. Tactical games typically have a lower treshold, and are more fun for beginners. Strategical games demand more of an investment but give more of a reward.

    For me this has always been a post-invention issue, never an a priori one. You can't decide to invent one or the other. My ambition was simply to find the best possible implementation of any given idea. If it turned out to be a strategical game, great. If not, then the result would at least be a good tactical game.

  • Essential Games
    I call a game essential if the basic idea translates into the rules without much of a choice on the inventor's part. Noughts & Crosses is fairly essential. So is Hex. Checkers is also essential, but International Draughts is not: it features backward capture and a long range king, obvious deviations from the basic idea of Checkers. Go is quite arguably the essential game par excellence.

    Chess games are never essential, and many great games aren't - in that sense there's no value judgement implied. But you create a chess variant. You can never find a one like you can find an essential game. Reversely, you cannot create an essential game because in a sense it is already there, and just waiting to be discovered.

    There's a value judgement involved nevertheless in that essential games usually are simple in terms of rules and structure, but consistentently resourceful in their consequences. Noughts a & Crosses may be excused and Checkers isn't a 'bottomless pit' anymore, since Schaeffer et al wrapped it up, but that's merely a matter of size. Go is extremely rich and resourceful, and Hex, with increasing boardsize, is not only a 'bottomless pit', but unlike Checkers resists programming quite tenaciously.

    Emergo, co-invented with Ed van Zon, is the essential 'column checkers' game.
    Symple is also essential with regard to its most unusual theme. Its invention was a notion drifting up while I was drifting off to sleep, and my last thought was "so simple ? ... what's wrong?".
    Both games were preceded by a number of other games based on their respective themes. How they explained themselves can be read in the Organic mechanisms section.