🌀 Tic Tac Toe in the Quantum Multiverse

Unbeatable and Quantum

🧠 The Idea

Normal Tic-Tac-Toe is simple: players alternate ❌ and ⭕ on a 3×3 grid, and optimal play always leads to a draw. A classical AI (minimax) explores the game tree intelligently: pruning bad branches, reusing symmetric states, and always choosing the strongest line. Against minimax, you cannot win.

Minimax prunes many of the possible paths, so that we don't need to visit every node in the game tree in order to play perfectly. Visiting every node can be pretty slow on a classical computer.

Solving Tic Tac Toe the Hard Way: BRUTE FORCE

<tic-tac-toe solve> = Brute Force Tic-Tac-Toe Solver

🙀 This could be awhile…

QuantumMode™ is different.

The engine explores the entire game tree by brute force, in quantum parallel, using a lightning fast algorithm running logarithmic wall time.
But like Schrödinger’s cat, you become entangled with one branch of that multiverse.
When the move is revealed, the wavefunction appears to collapse — and you only see the outcome from a single random branch.

🤯 Yep—people have fully enumerated it.

Upper bound (naïve): (9! = 362,880) sequences (pretend the game always fills all 9 squares and never stops early).
Actual number of games (stop when someone wins): 255,168 distinct play-throughs (no symmetry reduction). Breakdown: X wins 131,184, O wins 77,904, draws 46,080. (jesperjuul.net)
Up to rotations/reflections (game-tree complexity): 26,830 games. State-space (boards): (3^9 = 19,683) boards; 765 essentially different boards up to symmetry. (Wikipedia)

If you want counts by when games end (ply/depth), the totals are known too (still without symmetry):

End on move 5: 1,440
End on move 6: 5,328
End on move 7: 47,952
End on move 8: 72,576
End on move 9: 127,872 (Sums to 255,168.) (Mathematics Stack Exchange)

So: 9! is an upper bound that ignores early terminations and alternation constraints.

The Universe is a quantum computer capable of mind-boggling parallelism.

That means:

The solver “knows” all possible outcomes, but you can’t use that knowledge — you can only observe one collapsed reality.
Sometimes you land in a dud universe where the AI blunders.
Sometimes you get a draw, or even lose, but only by chance.

It's important to note that these branches are not symbolic or hypothetical, they are physically instantiated realities which will continue to evolve on their own, according the wavefunction. After all, you're in one right now! Just look around!

It’s like the million monkeys on typewriters: given infinite universes, one might type Shakespeare, but the one you happen to observe might just smash the keys.

🔩 The Web Component

Examples

🕹️ Human vs. Machine

<tic-tac-toe> = Human vs Minimax

Try playing against the AI! Can you beat Grandmaster Minimax?

<tic-tac-toe quantum> = Human vs Quantum

It should seem a little easier with the Quantume Multiverse as your opponent, because of quantum parallalism.

🤖 Auto-Play Mode

<tic-tac-toe auto-play> = AI vs AI

Now let's let get rid of the MITM and let AI play against itself. First Minimax vs. Minimax:

<tic-tac-toe quantum auto-play> = Quantum vs Quantum

Now, let's take the quantum mutliverse for a test drive:

The quantum multiverse does not seem very strong at playing tic-tac-toe, but remember, in a parallel universe, it performed as well as classical AI.

✨ Key Features

🤖 Classic AI mode
- Minimax: Unbeatable classical solver.
🐈 QuantumMode™
- Calls the split-universe API, which returns a random branch index.
🗣️ Declarative attributes
- Say what you want, and the System will figure out the messy implementation details!
📜 Game history
- Every board state is recorded.
- The player can step back/forward (⏪ / ⏩) and restart (🔄).
- Perfect for kibitzers to dive deep on analysis! 💡
🔁 Auto-play loop
- In auto-play mode, both X and O are controlled by AI.
- To avoid deterministic repetition, the first move is chosen randomly with crypto.
- Subsequent moves follow the selected AI mode.
🎨 Rendering
- The component manages its own DOM and injects styles.
- Board cells are interactive only when it’s the human player’s turn.
- A .quantum-warning appears if the API fails.

🧩 Algebraic Toe-Nation (ATN)

Tic-Tac-Toe commentary is an area of research and an evolving art form. Early attempts at move annotation were limited to gestures, grunts, and awkward silence, but with the advent of Algebraic Toe-Nation (ATN), analysts now possess a mathematically precise, intellectually rigorous, and grammatically well-defined language for drilling deep into arbitrary Tic-Tac-Toe positions.

ATN extends the brevity of chess notation into the absurd determinism of a 3 × 3 grid—encoding not just what was played, but how the cosmos felt about it. It acknowledges that in the quantum era of play, multiple outcomes can coexist: a move may be brilliant, disastrous, or both, depending on which branch of the multiverse one inhabits.

But naturally, ATN can record only a single path through the Quantum Maze; it, too, becomes entangled with its environment.

ATN is at once a notation, a philosophy, and a form of therapy for observers who have stared too long into the recursive abyss of optimal play. It has already produced extraordinary insights—among them, the once-unthinkable revelation of Double Checkmate.

📚 Learn by example

A textbook collapse of discipline. Both sides follow the Book for a single, blissful turn — symmetry, control, everything by the numbers. Then X breaks from orthodoxy with the audacious Xa2!!, launching an immediate flanking assault on the eastern front. O, perhaps overconfident, responds with the catastrophic Oc1??, abandoning the center entirely. The punishment is swift and absolute: Xc2#, a clean third-move checkmate. Commentators still describe this one as the shortest path from reason to ruin. A fan favorite!

[PlayerX "Deep Toe"]
[PlayerO "Dr. Toe-nator"]
[Opening "Center Gambit, Main Line"]
[Variation "Quantum Split-Universe Branch A"]
[Annotator "IM Schrödinger"]
1. Xb2  Ob3  
2. Xa2!!  
    X mounts an attack! A brilliant move!  
2... Oc1??  
    WOW! O veers dangerously!!  
3. Xc2# 1-0  
    X punishes O’s early mistake.

See below for an interactive replay!

Annotated Games

Game I — Deviation from the Book

Game II — Quantum Fork Collapse (Double Checkmate)

The celebrated Double Checkmate. At the fifth move, X completes both the middle row and the right column simultaneously — a dual-line victory unprecedented in classical play. Some theorists argue that this position represents a true collapse of the waveform, as both universes converge on X’s triumph. Others maintain that it is simply showing off. Either way, the crowd went silent, then applauded across adjacent realities.

ATN Formal Specification

Game        ::= Headers? Turns Result? EOF

Headers     ::= Header+
Header      ::= "[" HeaderKey WS? HeaderValue "]" NL?
HeaderKey   ::= /* any non-bracket, non-quote run (PGN-style) */
HeaderValue ::= Quoted

Turns       ::= Turn+
Turn        ::= Number "." WS* XMove (WS+ OMOVE)? NL?
OMOVE       ::= Ellipsis? Move

XMove       ::= Move          (* must start with 'X' *)
Move        ::= Player Square Suffix? Comment?
Player      ::= "X" | "O"
Square      ::= File Rank
File        ::= "a" | "b" | "c"
Rank        ::= "1" | "2" | "3"

Suffix      ::= "##" | "#" | "!!" | "!?" | "?!" | "!" | "??" | "?"
Comment     ::= WS+ FreeText  (* runs until NL or next Turn/Result token *)

Result      ::= "1-0" | "0-1" | "1/2-1/2" | "½-½"

Ellipsis    ::= "..."         (* optional before O’s ply: 2... Oc1 *)

Number      ::= Digit+
Digit       ::= "0"…"9"

Quoted      ::= "\"" QChar* "\""
QChar       ::= /* any char except " and newline; backslash escapes allowed */

WS          ::= (" " | "\t")+
NL          ::= "\r"? "\n"
EOF         ::= /* end of input */

FreeText    ::= FreeChar+     (* permissive; see lexer note *)
FreeChar    ::= /* any char except NL; parser should stop if a token
                  start for Number "." or Result is recognized */

Feeling depressed? Upgrade to Tic Tac Pro!