The Ocheckers Mastery Guide
You already know how to think in a system where small primitives compose into something expressive. Draughts is such a system. This guide teaches the intuition — what each rule really does, what shapes mean, and why a strong player's move looks obvious only afterwards. Every diagram here is a real position, and every line has been replayed through the engine's own rules to prove it is legal.
Part one
Before tactics, before openings: what kind of machine you are operating.
Sixty-four squares are drawn; thirty-two are real. Every piece begins on a dark square, moves diagonally, and therefore can never leave the dark squares. The light squares are scenery.
A city where every road runs diagonally and there is no way to cross onto the other grid. Two cars on different grids can drive past each other all day and never meet.
It is a narrower type than it looks. The board is not Square[64]; it is
Square[32] with a display that happens to render 64. Half the states you can
imagine are unrepresentable — and unrepresentable states are the ones you never have to
reason about. That is why players number the dark squares 1 to 32 and speak
in those numbers: 22-18, 22x15. Learn the numbering and the game
stops being spatial and starts being addressable.
If you have played the checkers most of the world knows, four things here are different, and all four matter more than they sound.
A man may only walk forward, but it may jump in any direction. Nothing behind you is safe just because it is behind you.
A king slides any distance along a diagonal, and captures at range — landing on any empty square beyond the piece it took. One king is worth about three men.
If a capture exists you must take it, and if several exist you must take the one that lifts the most men. You are not asked. You are told.
A man that passes through the last rank mid-jump keeps going as a man. It is crowned only if the move ends on that rank.
Mid-capture promotion is the one rule that genuinely varies from table to table. This engine uses the Brazilian rule above — pass through, stay a man. Some Nigerian tables use the Russian rule: crown instantly and fly on for the rest of the jump. It changes which combinations work. Ask first.
Because capture is compulsory, a capture is not something you hope for. It is something you can invoke.
This is the whole game in a sentence, and it is the thing beginners never internalise. In most games you influence your opponent. Here, when you place a man where it can be taken, you have not made a suggestion — you have made a call. They must answer. And because the rule also says take the maximum, you often know exactly which answer they must give.
Pulling a fire alarm. You do not persuade the building to empty — everyone must leave, and they must leave through the doors that exist. If you know the floor plan, you know where every person will be standing ten seconds from now.
An ordinary move is an event you emit and hope someone handles. A capture offer is a required handler: place the input and the framework must run it, synchronously, before anything else happens. "Take the maximum" is what makes it deterministic — not just "something will run", but "exactly this will run".
Once you see it that way, a combination stops being a lucky spot and becomes what it actually is: a small program you write in your opponent's pieces.
Dark squares are numbered 1 to 32, left to right, top to bottom. Black's men start on 1–12, White's on 21–32. Moves are written from-square then to-square:
22-18 — a quiet move, a slide from 22 to 18.22x15 — a capture, ending on 15.23x16x7x14x5 — a chain: four men lifted, one after another, finishing on 5.The board in the app prints these numbers on every empty square. Read a few games with them and you will start seeing squares rather than pixels — which is the point where you can begin to think a move ahead without moving anything.
Part two
Strong players are not calculating more than you. They are looking at fewer things.
Before you think about what you want, run this. In order. Every move, until it is automatic.
Checking your mirrors before you pull out. You are not admiring the road; you are looking at four specific things in a fixed order, and the whole thing takes a second because it is a habit rather than a decision.
These are the actual numbers this app's engine uses, in hundredths of a man. They are worth knowing because the move commentary in the app grades your moves against them.
| Piece | Worth |
|---|---|
| A man | 100 |
| A man, advanced towards crowning | +4 per rank |
| A man holding your own back rank | +6 |
| A piece in the four central files | +3 |
| A king | 320 |
Read the last row twice. A king is worth three and a bit men. That single fact reorganises the whole game: it means giving up two men to crown first is usually good, and it means that in the endgame material almost stops mattering next to who has a king.
Three delivery cyclists versus one van. On a short street the cyclists look fine. The moment the road opens up, it is not close — and the van can be anywhere on that road in one move, which means it is threatening everywhere at once.
The single most valuable mechanical skill in this game. Do it like this:
Because every step after the first is forced, this is not "calculating a variation" in the scary sense. It is following a chain of required handlers to the end. The board does the work; you just have to not stop early.
You count one exchange, see that it is even, and play it. But their recapture lands a man on a square that opens a second capture for them. Even trade, then a free man. Count to quiet, not to even.
Material is what you have. Tempo is whose turn it is when the position locks. In a blocked position, the side who must move is the side who must break something.
Practically, you count spare moves: quiet moves you can make that do not weaken anything — a man shuffling on a flank, a back-rank man that is not needed yet. When the position tightens, the player with more spare moves wins the standoff, because the other one runs out and has to push a man they wanted to keep.
Think of spare moves as slack in a queue. Two consumers are blocked on each other; the one with buffered work can wait, the one without must act and reveal its state. Tempo is who can afford to do nothing.
Part three
Individual men are almost worthless. What they form is everything.
A man cannot protect the man next to it. It protects the square behind that man — because if you jump into that square, the man behind takes you. So the useful shape is not a line across; it is a staircase back.
Bricklaying. A brick sitting alone on top of another brick falls off. A brick bridging two below it cannot. Advanced men need something under them or they are just exposed.
The four squares on your own last row are the only thing stopping an enemy man from becoming a king. Every man you move off that row is a door you have opened, permanently — men cannot walk backwards.
This does not mean never move them. It means: know that you are spending something that you can never get back, and get paid for it.
Sending your goalkeeper up for a corner. Sometimes it wins the game. It is never free, and you should know you have done it.
A hole is an empty square inside your own formation. It is an invitation: a man that lands there is often untouchable, because the men around it cannot reach behind themselves to walk it out — and in a game where jumps go backwards, an enemy sitting inside your camp is a permanent threat that also blocks your own moves.
The dog hole is the square in the corner of your own back rank. A man there has almost no moves and is often stuck for the entire game. Putting a man in it is usually a mistake; forcing your opponent to put one there is usually a small victory.
The long diagonal is the unbroken run of dark squares from one corner of the board to the other. It is the longest road on the map, and a king that owns it attacks along its entire length.
The double corner — the pair of squares at each end of that road — is the safest place on the board for a losing side, because a lone king there has two escape squares instead of one. Most drawn endgames in this game are somebody reaching a double corner in time.
A motorway and a cul-de-sac with two exits. Own the motorway and you can be anywhere. Reach the cul-de-sac when you are losing and they cannot pin you against a wall.
Part four
Where the game stops being about pieces and starts being about compulsion.
Every combination in this game has the same three-beat rhythm. Learn the rhythm and you stop memorising tricks, because you start recognising the shape of a trick.
A market trader handing you a free sample. The sample costs them something real. They are not being generous — they are moving you three steps closer to the crate, which is where they wanted you standing all along.
This is exactly what "give one to win two" means, and why it is the proverb of this game rather than a clever exception to it.
Here is the mechanism, and it is worth saying plainly because it is not obvious. When your opponent captures, two things happen at once: your man leaves the board, and their man moves. It lands on a square it was not on before. That landing square is the point. It is either now jumpable itself, or it has vacated a square that unblocks a road for you.
You are not sacrificing a man to "open lines" vaguely. You are using the compulsory-capture rule as a lever to relocate an enemy piece to a square of your choosing.
Six pieces. Black to move. Black's man on 15 walks forward to 18, where White's man on 23 is obliged to take it. Watch what that forced capture does to White's shape.
White's man had to leave 23. It landed on 14 — and the black king on 32, which had been looking down an empty road the whole time, now has two men lined up on its diagonal. One man given, two men taken.
The king on 32 did nothing for the entire combination. It did not need to. Its job was to already be there. Most combinations are won several moves before they happen, by a piece that was quietly put on the right road.
The same rhythm with a longer chain. White plays the quiet 26–23. Black must take. And White's man on 31 — which is nowhere near the action — walks through the wreckage lifting three.
Notice that no piece in this combination was ever "attacked" in the way you might expect. White simply arranged that after the forced recapture, three black men would be standing on alternate squares of one road, each with an empty square behind it. That is what a chain is: a line of men with gaps between them.
A capture chain is a fold over a list. Each hop consumes one element and moves you to the next. What you are looking for on the board is not "a capture" — it is an alternating pattern: enemy, gap, enemy, gap. Once you start scanning for that pattern rather than for individual threats, you see chains before they exist.
Fifteen pieces, a real middlegame. White gives a man with 19–16, Black is obliged to recapture with 12x19, and the man on 23 clears out four black men in a single move.
Step through it slowly. The thing to feel is that White's first move looks like a mistake, and there is no way to see that it is not unless you have counted the chain to its end. This is the whole reason the five-second scan insists on counting to quiet.
One rule governs how chains are built, and it catches everybody once. During a multi-jump, the men you have captured stay on the board until the whole move finishes. They are removed only at the end.
Two consequences, both important:
Clearing a table by carrying plates to the sink one at a time versus sweeping them all into a crate at the end. Here it is the crate: nothing actually leaves until you have finished walking. The plates you have already claimed are still in your way.
Watch this four-man sweep carefully. White's man travels 25 → 18 → 9 → 2 → 11. Square 2 is on White's crowning row. The man lands there mid-jump — and walks off again, still a man.
Under the Brazilian rule used here, that man finishes on 11 uncrowned. Under the Russian rule some tables use, it would have been crowned the instant it touched square 2 and could then have continued as a flying king — a completely different move, often taking more. Same position, same rules of capture, different game. Ask before you play.
Part five
Every combination is one of these, or two of them stacked.
Once you have the offer–forced–strike rhythm, the only question left is what the offer is for. There are five answers. Learn their names; naming a thing is what lets you notice it.
Pull a piece to a square. You want their man standing on 14; you give them something that can only be taken by landing on 14.
Like ringing a doorbell to get someone into the hallway.
Pull a piece away. A man is guarding a square. You give them a capture they must take with that man, and the square it was guarding falls open.
Like drawing the goalkeeper off their line.
Fill a square so their chain dies. Because captured men stay until the move ends, you can put a man where their jump would have continued and cut it short.
Like parking across a driveway.
Get your own man out of the way. Sometimes the piece blocking your king's road is yours. Give it away with tempo and the road opens.
Like moving your own car before the ambulance arrives.
Two jobs, one man. A man is holding two squares. Attack both; whichever way the forced capture goes, the other falls.
Like one person covering two exits.
The two-for-one diagram above is a decoy: White wanted a black man standing on 27. The big one is a deflection followed by a chain: the man on 12 was doing a job, and after 12x19 it was somewhere else and the road was open.
When you feel there is "something here" but cannot find it, stop looking for captures and ask the five questions instead. Which square do I wish an enemy man were standing on? Which of their men do I wish were somewhere else? Which of my own men is in my way? The combination usually answers itself.
Part six
A king is not a better man. It is a different kind of object.
Three men or one king, at 100 against 320, is close to an even trade on the arithmetic. It is not close on the board, because the three men can only be in three places and the king can be anywhere on its roads. Material counting understates kings; every strong player overpays for the first one and is right to.
Most middlegames are secretly a counting problem: how many moves until I crown, how many until they do, and who moves first? If you are two tempi ahead in that race, a lot of otherwise reckless play is justified, including giving up a man to clear the road.
Two people running for the last train. It does not matter who is fitter. It matters who is closer, and whether anything is standing in the aisle.
Two situations where pushing for the crown is wrong:
Part seven
Five studies. Each one was played out by the engine at full strength from a quiet position, and the result below is what actually happened — not folklore.
Endgame knowledge is the highest-leverage thing you can learn, because it tells you which advantages are worth having. There is no point winning a man in the middlegame if the resulting endgame is a dead draw, and there is every point giving a man to reach one that is not.
Three against one is winning. The method is not "chase" — a flying king cannot be chased, it simply flies past you. The method is to take away roads: occupy diagonals so the lone king's options shrink one at a time, then drive it to an edge where its two escape squares become one.
This is the single most useful fact in this chapter, and it surprises people. Two kings do not beat one. The lone king reaches a double corner, sits on the long road, and cannot be squeezed — there are always two ways out and only two attackers.
If you are heading for an endgame of two kings against one, you are not winning. Which means: a middlegame plan that trades down to that is a plan to draw. Keep a third piece, or keep men on the board, or do not make the trade.
Same reason. The extra man is not enough to close the second exit, and the defending king is fast enough to harass it. Knowing this stops you from playing on for forty moves in a position that was never going anywhere.
The other side of the same coin. Two men, with no king of their own to answer with, cannot defend a road. The king picks them off from range and they have no reply, because a man can only threaten the square directly diagonal to it.
Together with the previous two studies, this gives you the rule of thumb that decides most endgames:
Kings answer kings. An extra king wins against pieces that cannot fly. An extra king against a side that has a king and a corner is usually only a draw. So the question is never just "am I a piece up" — it is "will they still have a king?"
Included because it shows what balanced king play looks like: both sides holding roads, neither able to commit, nothing achievable without help. Watch how little actually happens. That is not the engine being timid; that is the position.
Part eight
You do not need a repertoire. You need four habits.
That is genuinely it. In a game where captures are compulsory and maximal, opening theory matters far less than the ability to count a forced sequence, because a single miscount ends the opening immediately. Spend your study time on chains, not on move orders.
Openings here are conventions, not APIs. There is no framework contract to learn — there is a small set of invariants to preserve (centre, back rank, support, no loose men), and everything else is derived at runtime by calculation.
Part nine
Using this app deliberately, in four weeks.
The app has three tools that most people never turn into a practice loop: the move commentary that grades every move against the engine, the replay theater that plays your games back, and the fact that you can drop the engine's strength to make it punish you slowly instead of instantly.
Play at Casual with commentary on. Ignore whether you win. After each game, open the replay and step through every move marked ? or ??. For each one, do not look for the better move — look for the chain you did not count. Goal: stop hanging pieces to two-move combinations.
Same setting, one new rule: before every capture, say the whole sequence out loud before you touch anything. "I take, they must take with that one, then I take two." If you cannot say it, do not play it. Goal: your Mistake count per game halves.
Move to Club. Now play for combinations: each game, try to create one position where a quiet move forces a capture you have prepared. You will lose games doing this. That is the price. Goal: earn one Brilliant.
Sharp strength, and a new question every time you consider a trade: what endgame does this lead to, and is it one of the drawn ones? Re-read part seven. Goal: stop trading into two-kings-against-one when you are the one who is a piece up.
Tape this to the inside of your head. It is four questions and it costs five seconds.
After a loss, watch the replay before you play again. Your grade marks show exactly where the game turned — usually one move, usually four or five moves before you noticed. The gap between "I lost" and "I lost here, because I did not count that chain" is the entire difference between playing a lot and getting better.
Reference
22-18 a slide · 22x15 a capture ·
23x16x7x14x5 a chain, one x per man taken, finishing on the last square.
Recent
Everything here
Nothing here by that name.
Keyboard shortcuts