Take Points and the Doubling Window: When a Backgammon Double Is Actually Correct

How the doubling window works in backgammon — take points, market losers and the too-good zone — so you know exactly when to offer the cube, take, or drop.

Most players learn the doubling cube as three separate decisions — offer, take, or drop — and treat each roll as a fresh guess. But strong play flows from one idea that ties all three together: the doubling window. It's the band of winning chances inside which offering the cube is correct, bounded below by the point where your opponent must take and above by the point where you're winning so heavily you shouldn't double at all. Learn where those edges sit and cube decisions stop feeling like gambles and start feeling like arithmetic.

TL;DR

• Your opponent's take point in a simple money game is about 25% — below that they must drop, at or above it they must take.
• The doubling window is the range of your winning chances where a double is correct: roughly from ~50% up to ~80% in a plain race, wider or narrower once gammons are in play.
• Above the window you're too good to double — playing on for a gammon is worth more than cashing the game.
• Doubling windows open because of market losers: rolls that would push you past your opponent's take point, so you'd lose your chance to double for value.

What a take point is

The take point is the minimum share of games you need to win to justify accepting a double. In a straight money game with no gammons and no recube value, that number is 25%. The logic is simple: if you drop, you lose one point for certain. If you take, you're now playing for two points, so you only need to win one game in four to break even against the point you'd have conceded. Anything above 25% and taking is the cheaper long-run choice; anything below and you should drop and save the point.

Two things move the take point off its textbook 25%. Gammons raise it — if you might lose double, you need more winning chances to justify taking. And recube vantage, the value of owning the cube so you can redouble later, lowers it, often to around 20-22% in practice. That's why experienced players take cubes that look hopeless to a beginner: cube ownership is worth real equity.

The doubling window

If 25% is where your opponent is indifferent to taking, then 75% winning chances is where you, the doubler, are on the edge. Below that band a double gives away too much; inside it, doubling wins you value your opponent is forced to pay. The window has two edges:

1. The lower edge (~50%) — you generally want a real advantage before turning the cube. Doubling as a slight favourite hands your opponent an easy take and surrenders your cube ownership for little gain.
2. The upper edge (~80%) — past this, you risk becoming too good to double. If you're winning so decisively that gammons are likely, cashing a single point throws away the extra point a gammon would bring.

Between those edges lies the sweet spot: strong enough that the cube has bite, not so strong that you should play on. In a pure race the practical rule of thumb is that a lead of roughly 8-12% in the pip count opens the window, and a lead past that starts to close it.

Market losers: why the window matters

The reason you can't just wait for a huge lead is market losers — rolls that would blow past your opponent's take point before your next chance to double. Picture yourself a modest favourite. If a good sequence of dice would suddenly make you a heavy favourite, your opponent would then correctly drop, and you'd have lost your market: the game was worth doubling, but you missed the moment.

So the practical trigger is this: double when you have enough market losers that waiting risks giving up a valuable double. If a meaningful fraction of your rolls would push the position out of your opponent's take range, offer the cube now while they're still obliged to pay. This is why timing beats raw strength — a position with lots of market losers is a double even at 60-65%, while a stable position might wait.

Putting it together at the table

Ask three quick questions before you touch the cube. First, am I past 50%? If not, hold. Second, is my opponent above or below their ~25% take point? If they're below it, they'll drop and you cash a point — fine, but check the third question. Third, do I have enough market losers, or am I too good? If good dice would cost me a value double, offer now. If good dice would win a gammon and I'm already dominant, play on rather than cash. That single pass through take point, window, and market losers replaces guesswork with a repeatable read. For the mechanics of the cube itself, see our guide to the doubling cube, and for how these numbers shift in tournaments, match play and the Crawford rule.

Provably fair — verify

Cube decisions only mean something if the dice behind them are honest, and on 6proclub every roll is committed before you see it. Ahead of the roll the server draws a secret serverSeed and shows you only its SHA-256 hash — a fingerprint it cannot change later. Your own clientSeed is mixed in, and a nonce counts up one per roll, so the full input is serverSeed : clientSeed : nonce. When the game ends the server reveals the original serverSeed; you hash it yourself, confirm it matches the fingerprint shown up front, then replay every roll from serverSeed : clientSeed : nonce to prove the dice were fixed before you acted. See Provably-Fair Cash-Out Games for the full walkthrough, and the whole casino runs on the same commit-reveal chain.

FAQ

What is a take point in backgammon?

The take point is the minimum share of games you must win to justify accepting a double. In a simple money game it's about 25%, because taking doubles the stake, so winning one game in four breaks even against the point you'd lose by dropping.

What is the doubling window?

The doubling window is the range of your winning chances where offering the cube is correct — roughly from 50% up to about 80% in a plain race. Below it a double gives away too much; above it you may be too good to double and should play on for a gammon.

Why is 25% the take point?

Because dropping costs one point for certain, while taking plays for two. Winning 25% of the time earns back exactly that one point on average, so above 25% taking is cheaper in the long run and below it dropping is.

What does "too good to double" mean?

It means you're winning so heavily that gammons are likely, so cashing a single point by doubling would throw away the extra point a gammon brings. In that case you keep the cube and play on rather than offer it.

What are market losers?

Market losers are rolls that would push the position past your opponent's take point, costing you the chance to double for value. When you have enough of them, you should double now while your opponent is still obliged to take, rather than wait and lose your market.

18+. Crypto only. Backgammon involves risk — set a budget before you play, and treat strategy as a way to enjoy the game, not a guarantee of winning.