The Patricia DeMauro Craps Roll: 154 Rolls Without a Seven-Out

The odds of avoiding a seven on any single roll of two dice depend on the point established. But a simplified model: across a full craps game, the probability of not sevening out on a given roll is approximately 5/6, since five faces of six possible two-dice totals are non-seven outcomes... no. That's too loose. The actual calculation requires tracking the sequence of come-out rolls and point phases.

A cleaner estimate: gaming mathematicians who analyzed DeMauro's run calculated the odds at approximately 1 in 1.56 trillion. That number has been cited in various gaming publications and stands up under scrutiny. Even if you disagree on the exact figure by an order of magnitude, you're still looking at an event that should occur roughly once in hundreds of billions of craps sessions.

For context, there are an estimated 100 billion craps decisions played globally each year across all regulated casinos. DeMauro's run should occur, statistically, once every 15,600 years of global craps play at current volumes.

The DeMauro record is frequently cited as evidence that craps streaks are real, that certain shooters are "hot," or that strategy adjustments during a cold table or a hot shooter make mathematical sense. None of those conclusions follow from the data.

Craps uses standard casino dice, physically tested to tolerances of 0.0002 inches (within 1/5000th of an inch of perfect cube geometry) by manufacturers like GPI and Midwest Game Supply. Each roll is, within measurement precision, independent. The dice have no memory.

The DeMauro roll is not evidence of controlled shooting. It is a tail event. A sample size of one spectacular outlier does not validate the general claim that extreme outliers are reproducible.

Some advantage-play communities maintain that skilled dice control can shift the outcome distribution. This claim has been studied. Stanford Wong, whose work on card counting is well-documented, analyzed controlled throwing in a 2006 book. His conclusion was that modest influence is theoretically possible but requires extraordinary practice and degrades in casino conditions due to felt type, backwall texture, and chip clutter.

The evidence standard here is not casual. A shooter would need to demonstrate statistically significant deviation from expected outcomes across thousands of rolls, under casino conditions, in front of independent observers. No one has produced this to academic standard.

DeMauro, by her own account, was a relative beginner at craps when the record happened. She was playing at the Borgata with a companion. Her technique was not the source of the outcome.

There are a few concrete things the DeMauro record usefully illustrates.

• Sample size illusions are real. In a short craps session, almost anything can happen. A player who experiences 40 rolls without a seven and starts believing in their "hot hand" has observed roughly 1/3.85 of DeMauro's run. Their streak is not exceptional.
• Casino variance is meaningful. With a house edge of approximately 1.36% on pass line bets (taking 3-4-5x odds), a skilled bettor can survive extended winning streaks and extended losing streaks. The DeMauro event would have made a patient pass-line player significant money. A field bet player less so.
• The Borgata's surveillance infrastructure verified the count. This is the least discussed part. Casino surveillance systems generate shot-by-shot records. That record was preserved and authenticated. If you want to argue with the 154 number, you need to argue with the Borgata's tape.

The record is what it is: a documented, authenticated, physically improbable event that tells us nothing about what the next shooter will do.