The doubling cube is the most powerful and most misunderstood tool in backgammon. Handled correctly, it multiplies your wins and minimises your losses. Mismanaged, it turns winning positions into disasters. This guide covers the full framework for cube handling — when to double, when to take, when to drop, and the strategic concepts that tie it all together.
Key Takeaways
- Double when your winning probability is in the range of roughly 50–80% — this is the "doubling window"
- Take a double if your winning probability is at least ~25%; drop below that
- The player who owns the cube (last accepted) has exclusive right to re-double — a major strategic advantage
- "Too good to double" exists: above ~80%, play on without doubling to win a gammon instead
- Gammon risk shifts the take point downward — high gammon positions require better winning chances to take
- In match play, match equity modifies all cube thresholds — money game rules don't apply directly
The Doubling Cube: Quick Review
The doubling cube is a special die with faces numbered 2, 4, 8, 16, 32, and 64. It starts in the middle of the board at “1” (no cube in play). When you offer the cube, you propose to double the current stakes. Your opponent can accept (take) or decline (drop/resign).
This guide assumes you know the mechanics — if not, start with the doubling cube guide.
The Fundamental Cube Framework
The entire cube decision-making process is built on equity — the expected value of a position.
At any point in a game:
- Your equity = (probability of winning × value of win) − (probability of losing × value of loss)
The cube changes the value of wins and losses, which is why cube decisions are so consequential.
When to Double: The Doubling Window
The Basic Threshold
You should double when your winning probability is approximately 50–80%:
| Winning Probability | Cube Action |
|---|---|
| Below 50% | Do not double (you’re not favourite) |
| 50–80% | Double (correct doubling window) |
| Above 80% | Too good to double — play on for gammon |
Why Not Double Below 50%?
Doubling when you’re behind gives your opponent maximum leverage over the game at the highest stake. You should only double when you’re a genuine favourite.
Why Not Double Above 80%?
If your opponent would drop (decline) a double at this point, you’ve lost the value of the game above 1 point. Instead, play on without doubling — aim for a gammon (double the score) which is worth more than a doubled single game the opponent drops.
The “Too Good to Double” Concept
When you are extremely dominant (above ~80% win probability with significant gammon chances), the correct play is to not double. Your opponent would drop, giving you only 1 point (or current cube level). But by playing on, you might win a gammon for 2× the current stakes.
This only applies when your opponent would clearly pass the cube. If they’d take at 80%, doubling is still correct.
When to Take a Double
The Basic Take Point
Take a double if your winning probability is at least ~25%. Here’s why:
- If you drop: You lose the current stake (e.g., 1 point)
- If you take: Your equity = (25% × 2 points) − (75% × 2 points) = −1 point
At exactly 25%, taking and dropping have the same equity. Above 25%, taking is better. Below 25%, drop.
The 25% rule is approximate — adjust for:
- Gammon risk (discussed below)
- Match score (discussed below)
- Cube ownership value
Cube Ownership Adds Value to Takes
When you take a double, you own the cube. This means only you can re-double — your opponent cannot. Cube ownership is a real asset:
- You can re-double when the position swings back in your favour
- You have more flexibility in endgame cube decisions
- This extra value means you can take some positions even slightly below 25%
The approximate take point with cube ownership considered is closer to 22–23% in many positions.
Gammon Risk and the Take Point
Gammon risk significantly affects whether to take or drop. The key concept: if gammons are likely, the take point shifts upward (you need better winning chances to take).
Simple Example
Your opponent offers a cube at 2. If accepted and you lose a plain game, you lose 2 points. But if there’s a 30% chance they gammon you (win 4 points), your expected loss is:
- 40% × win 2 pts + 30% × lose 4 pts + 30% × lose 2 pts = much worse than a simple −2 drop
In high-gammon positions, even 30% winning chances may not be enough to take. You need significantly better chances to compensate for the gammon risk.
Gammon price: See backgammon gammon price for how to evaluate gammon risk precisely.
Cube Timing: Early vs. Late Doubles
Early Doubles (Efficient)
A double offered at the start of your doubling window (~50–55% winning probability) is called an efficient double. The opponent is right to take (they have ~45–50% winning chances). The game is close and both players profit from the right cube decision.
Late Doubles (Inefficient)
A double offered when your winning chances are 70–80% is potentially inefficient — your opponent is a clear underdog but may still barely have a take. You’ve waited too long: you gave away some equity compared to doubling earlier.
Doubling for the Right Reason
Good cube timing means identifying the turning point of a game — when your advantage has just crossed the doubling threshold — and acting there, not earlier or later.
The Recube
Once your opponent takes your double, the cube sits at 2 on their side of the board. Only they can offer the next double (the recube or re-double to 4).
When to Recube
The same 50–80% window applies from the cube holder’s perspective. If the position swings and you (the cube holder) find yourself favoured, consider recubing when your winning probability crosses 50%.
Cube Pressure
Sometimes cube ownership pressures the other player into making suboptimal plays. They know you can recube if the position swings — this knowledge itself has equity value.
Match Play Cube Decisions
In tournament match play, the cube decisions differ from money games because each game’s value is non-linear — it depends on the match score.
Key Differences from Money Game
- Gammons are always worth double (no Jacoby Rule in match play)
- Crawford Rule prevents the cube in the game after one player reaches match point − 1
- Match equity tables give precise values for each score — use these for accurate decisions
Common Match Score Adjustments
| Match Score Situation | Cube Adjustment |
|---|---|
| Leading near match point | Double more aggressively (a win wins the match) |
| Trailing by 2+ points | Be more willing to take gamble cubes |
| Crawford game (no cube) | Focus entirely on position/gammon chances |
| Post-Crawford (trailer) | Always double immediately — nothing to lose |
| Both players at match point − 1 | Never double (DMP — Dead Man’s Point rules apply) |
At double match point (DMP) — both players need 1 point to win — the cube is irrelevant (all results score 1 point to the winner). Play purely for winning probability.
For the full match equity framework, see the match equity guide.
Practical Cube Handling Tips
1. Think Before Rolling, Not After
Cube decisions must be made before rolling your dice. If you pick up the dice and roll before offering the cube, you’ve given up your right to double that turn.
2. Don’t Be Rushed
Take your time on cube decisions — they are the highest-equity decisions in backgammon. Think through the position carefully before deciding to double or take.
3. Keep the Cube Accessible
The cube should be visible and accessible to both players. Don’t hide it or leave it in an ambiguous position.
4. Communicate Clearly
Announce cube actions clearly: “Double” (offering the cube), “Take” or “Drop” (accepting or declining). Avoid ambiguous gestures.
5. Use the Cube Aggressively (But Correctly)
Many intermediate players are too passive with the cube. If you are winning, double — don’t wait for a perfect position. The cube’s value comes from being used at the right moments, not from being saved.
Common Cube Mistakes
| Mistake | Correction |
|---|---|
| Doubling too early (<50% win probability) | Wait until you have a real advantage |
| Doubling too late (>80%, opponent will drop) | Double earlier to catch their take |
| Not doubling when well ahead | Being passive costs equity — double in the window |
| Dropping at 30% win probability | Calculate more carefully — 30% is usually a take |
| Ignoring gammon risk | Always factor gammon chances into take/drop decisions |
| Forgetting match score | Match play requires match equity analysis, not just raw probability |
Summary
Cube handling is the highest-leverage skill in backgammon. The core framework:
- Double in the window of ~50–80% winning probability
- Take when your winning chances exceed ~25% (adjusted for gammon risk and cube ownership)
- Drop when your winning chances are below ~22–25%
- Too good to double when above ~80% — play on for gammon value
- Match play requires matching raw cube decisions to the match equity implications of each score
For deeper study: see doubling cube mechanics, match equity, gammon price, and match play strategy.