You just played a game you want to study later, or a position you want to send a friend. Typing "I dropped in the middle, then they went one to the left, then I stacked on top of that" works, barely, for three moves. It falls apart by move ten.
Connect Four has a notation for exactly this, and it is simpler than almost any other board game's. Here is how to read it, how to write it, and when to reach for which version.
Why Chess Notation Doesn't Apply Here
Chess notation names a piece and a destination square, because a piece can land on any of 64 squares from almost anywhere on the board. Connect Four doesn't have that problem. Gravity does the work for you.
Drop a disc in column 4 and it falls to the lowest open row in that column. There is no piece type to record, no ambiguity about which disc moved, no capture notation. The only decision in Connect Four is which column to play. Everything else, including which row the disc lands on, follows automatically from how many discs are already stacked there.
That one fact is why Connect Four notation collapses to something a chess player would find suspiciously easy.
The Column Notation
The standard way to record a Connect Four game is a plain sequence of column numbers, one digit per move, columns numbered 1 through 7 from left to right. Players alternate automatically, starting with player 1.
4 4 5 3 4 4 7
Read that as: player 1 drops in column 4, player 2 drops in column 4 and stacks on top, player 1 plays column 5, player 2 plays column 3, player 1 plays column 4 again, player 2 plays column 4, player 1 plays column 7.
Seven digits describe seven moves and the exact board after each one, because the row is never in question. This is the same convention used by Pascal Pons' widely cited Connect Four solver, and it is the format behind every example sequence on this site. Look at the solved-game explainer or the openings database and every replay on the page is built from nothing but a list of column digits.
Because the format is this compact, it travels well. Paste a move list into a chat message or a forum post, and anyone with a board, physical or digital, can replay it move for move. It is also exactly what the engine and game review need: feed in the column sequence and the position at every step reconstructs itself, no row coordinates required.
When Column Numbers Aren't Enough
Column notation tells you what move was played. It doesn't give you a quick way to name one specific square in isolation, the way "e4" does in chess, without first replaying the game up to that point and counting discs.
That's rarely a problem when you're just recording a game. It becomes a problem the moment you want to talk about a single square, like describing exactly where a threat sits. "The threat on column 3, the one three discs up" is a lot more awkward than just naming the square directly.
That is the gap Victor Allis's notation was built to fill.
The Allis Notation (Columns and Rows)
Victor Allis's 1988 master's thesis at Vrije Universiteit Amsterdam, the paper that proved Connect Four is a forced win for the first player, uses a different system: letters for columns, numbers for rows. James D. Allen, who solved the game independently that same year, reasoned about the same squares using the same column/row logic.
- Columns are lettered a through g, left to right. Column 1 is "a," column 4 (the center) is "d," column 7 is "g."
- Rows are numbered 1 through 6, counted from the bottom. Row 1 is the floor of the board, row 6 is the top.
A square is named column-then-row, no space: d1 is column 4, row 1, the very first square anyone can ever play. a3 is column 1, third row up.
Moves are numbered in pairs the way chess notation does it, with the second player's reply marked by an ellipsis: "1.d1" means move 1, first player plays column 4, row 1. "1...c1" means player two's reply lands on column 3, row 1 (the bottom, since nothing had been played there yet).
This letter-row system is what you will run into in the academic literature and in deep strategic writing, because it can name a threat square directly instead of describing its location. The double threat and odd/even parity articles on this site both lean on exact square references for the same reason: "the threat on e4" is precise in a way "somewhere in column 5" is not. It is also the notation behind the nine named rules from Allis's threat framework (claimeven, baseinverse, aftereven, and the rest), all of which are defined in terms of specific squares a threat sits on, not just a column.
Converting Between the Two
Both systems describe the same board, so converting between them is a straight lookup.
| Column # | Letter |
|---|---|
| 1 | a |
| 2 | b |
| 3 | c |
| 4 | d |
| 5 | e |
| 6 | f |
| 7 | g |
The row number in Allis notation is just the disc's position in that column, counted from the bottom. Column notation leaves that implicit: if a column digit is the third time that number has appeared in the sequence so far, it is landing on row 3 of that column.
A Worked Example
Game 1 of the 1995 Allen vs. Tromp match opened like this in Allis notation:
1.d1 c1 2.g1 d2 3.D3 d4 4.F1 e1
Converting each square to a column number with the table above: d to 4, c to 3, g to 7, d to 4, d to 4, d to 4, f to 6, e to 5. That gives the column sequence:
4 3 7 4 4 4 6 5
Open the replay widget on that game's writeup and you can watch the board build up exactly that way: a disc in the center, a reply one column over, a disc on the far right, then three more discs stacking straight up in column 4 before the action shifts to column 6 and column 5. Same game, two notations, same board, every time.
Recording a Single Position vs. a Full Game
Not everything worth sharing is a full game. Sometimes you only care about one position, like a puzzle or a specific tactical idea, and replaying an entire game to reach it is overkill.
For that, the practical move is a board diagram: a snapshot of the 7-by-6 grid showing exactly which squares hold which color, with no move history attached. Puzzle collections on this site are built this way, because a puzzle is defined by its position, not by how anyone got there. The column-digit sequence is still useful here too, just truncated: "replay the first 11 moves of this game" reconstructs the position without needing a separate diagram format.
Pick the tool that matches what you are actually communicating. A move sequence answers "how did we get here." A board diagram answers "what does the position look like." Most strategic writing, including the articles linked throughout this one, uses both: a short move list to set up the position, then a diagram to freeze it at the moment that matters.
Which One Should You Use
Use column digits for anything you are sharing or feeding to a tool: a game you want a friend to replay, a position you are pasting into the engine, a line you are saving in your match history. It is shorter, unambiguous, and every Connect Four tool reads it the same way.
Reach for letter-row notation when you are talking about a specific square rather than a move: writing up strategy, citing a threat, or reading primary sources like Allis's thesis or community breakdowns of famous games. It says "this exact square" when column notation can only say "somewhere in this column, eventually."
Most players never need to write Allis notation themselves. But recognizing it means you can read the research behind why column 4 is the only winning opening without translating every square reference in your head first. That is worth the five minutes it takes to learn.
Ready to put it to use? Play a game and check your move history afterward, or replay the Allen vs. Tromp match move by move and practice converting between both notations as you go.