The Doubling Cube for Beginners: When to Offer, Take, Drop

The doubling cube is the engine of backgammon's strategic depth. Most beginners never learn to use it properly. Here are the four practical rules that cover 90% of real cube decisions — without any equity tables required.

The Doubling Cube for Beginners: When to Offer, Take, Drop

If you've played a few backgammon matches and the cube is still mysterious to you, you're not alone. The doubling cube is the single biggest reason new players think backgammon is "harder than chess." It isn't — the cube just has rules that nobody bothered to teach you. This post is the small handful of practical decisions you actually need, written for someone who knows how the pieces move but freezes when the cube comes out.

For the deeper mathematical treatment, see the full cube strategy guide. This post is the cheat-sheet version.

What the cube actually does

The doubling cube is a six-sided die showing 2, 4, 8, 16, 32, 64. At the start of a match, it sits in the middle of the bar at "1" (centered, unowned). Either player can, on their turn before rolling, offer to double the stakes. The opponent must either:

• Take. Accept the double. The cube becomes theirs, value goes to 2. From now on, only the cube's owner can offer the next double. If they win, they win double; if they lose, they lose double.
• Drop. Refuse the double and concede the game at the current value (which was 1, so they lose 1 point).

The cube can keep getting redoubled — 2 → 4 → 8 → up to 64 in theory, though anything above 8 is rare in human play.

That's the mechanic. The strategy is knowing when to offer and when to accept.

The four practical rules

Rule 1 — The "Match Equity" rule for taking

The simplest, most useful rule for whether to TAKE a double is the 25% threshold. If you think you have at least a 25% chance to win the current game from the current position, take the cube. If less, drop.

Why 25%? Because the math of a doubled cube means you need a 25% win rate to break even on the take. Below that, dropping (and losing 1 point) is cheaper than taking and losing 2 points 76%+ of the time.

This is approximate — there's a small correction for gammons (more on those below), and the threshold creeps up slightly when you can be re-doubled later. But for the first few hundred matches you play, "25%" is the right answer.

How do you estimate your win rate at 25%? Look at the position:

• If you have a clear positional advantage (lead in the race, opponent on the bar, your home board half-built) → take, easily
• If positions look roughly equal → take
• If you're clearly behind but not blown out → take
• If you're WAY behind (opponent has closed-out half-board on you, you're on the bar facing a 5-point board) → drop

You will take too much in your first 50 matches. That's fine. The cost of an over-take is much less than the cost of an over-drop.

Rule 2 — The "good but not great" rule for offering

You should OFFER a double when:

1. You have a clear positional or racing advantage
2. Your opponent can still reasonably take (you don't want them to drop)
3. Things are about to change — your good position might evaporate next roll

That last bullet is the bit beginners miss. If your position is so good that your opponent will obviously drop, you've doubled at the wrong moment — you would have won the game at the cube value of 1 anyway. The right time to double is when the position is just barely good enough for them to take.

A useful mental check: "If I were the opponent, would I take?" If your honest answer is "probably yes, with some hesitation" — that's the moment to offer.

Rule 3 — The "race" simplification

In a pure race (no contact, both sides just rolling home), the cube decision is almost purely about pip count.

If your pip count is lower by 8% or less of your opponent's, you have a good take. If your lead is 10–14% lower, you have a good double-and-pass-back-to-them. If your lead is bigger than 14%, your opponent should drop.

These percentages assume both rolls are still ahead. Adjust slightly for who's on roll (the player whose turn it isn't has effectively a half-pip handicap).

This is a useful shortcut. You don't need exact equity tables — pip count comparisons are quick mental math and good enough for ~80% of race decisions.

Rule 4 — The "gammon danger" override

A gammon (winning before your opponent has borne off any checkers) doubles the cube value. A backgammon (winning while opponent still has a checker on the bar or in your home board) triples it. When gammon risk is high, the take/drop math shifts dramatically.

If your opponent has gammon chances:

• Threshold to TAKE drops from 25% to 22% if they have ~15% gammon chances on top of game wins
• Threshold to TAKE drops to 18% if they have ~30% gammon chances

In short: never take a double from a position where the opponent might close you out. Even if you'd technically still "win the game" 25% of the time, the gammon premium turns those losses into double-losses. Drop the cube and minimize damage.

The "opponent has a strong home board and I'm on the bar" position is the textbook drop scenario. Don't be brave; just drop.

Common beginner mistakes

Doubling too early. The first time you have a small lead, the instinct is "lock it in now!" — but the opponent just takes, and now you've raised the stakes on a marginal lead. Wait until your lead is real.

Dropping too often. The 25% rule looks scary, but 25% is not high in a game with as much variance as backgammon. If you're dropping more than you're taking, you're being too cautious.

Forgetting you can redouble. If you took the cube and the position has improved for you, you can offer back at 4. The opponent now has to take/drop the same way you did. Many beginners never redouble; this is leaving money on the table.

Beavering when you shouldn't. A "beaver" is taking and immediately redoubling on your own turn. It only makes sense when you think the opponent doubled prematurely (their position isn't actually good enough). Beavering when you're genuinely behind is just doubling your losses.

A simple decision tree

When opponent offers a double:

1. Am I winning the race comfortably and not at gammon risk? → Take, consider beaver
2. Am I winning the position but slightly behind in pips? → Take
3. Am I in a contact position, behind but with realistic chances? → Take
4. Am I on the bar against a 5-point or 6-point home board? → Drop
5. Am I so far behind that I see no path to winning even 25% of the time? → Drop
6. Did I just get hit and now face gammon risk? → Drop

When considering offering a double:

1. Do I have a clear positional or racing lead? → Maybe — go to 2
2. Would the opponent reasonably take this? → If no, don't offer (you'd lose value)
3. Is the position about to change next roll? → If yes, offer now

Practice it on 6proclub

Match buy-ins from $5, ELO ladder, fully-implemented cube (no, the cube isn't optional — every real backgammon match uses it). The provably fair dice mean you can verify each roll yourself, so any cube decision you make is grounded in honest randomness. Browse the lobby.

In one paragraph

The doubling cube isn't optional in real backgammon. Four practical rules cover almost everything: take if you have 25%+ win chances, offer when your lead is real but the opponent will still take, use pip count in pure races, and drop hard when gammon risk shows up. Skip the equity tables for now — these four rules will take you from "cube freezes me" to "I'm reading the position" within 20 matches. The depth comes later. The basics get you playing.