Dice Influence and Variance

Two dice are rolled. There are 36 possible combinations. Six of them make a 7, so normally a 7 appears 16.67% of the time (once every 6 rolls). We call this an SRR of 6 (Sevens to Rolls Ratio: on average, one 7 every 6 rolls).

One of the simplest bets is a Place 6: you bet $6 that a 6 will be rolled before a 7. If the 6 comes first, you win $7. If the 7 comes first, you lose your $6. All other numbers are irrelevant. The bet just waits.

Normally, there are 5 ways to roll a 6 and 6 ways to roll a 7, so you win 5 out of 11 times (45.5%). The casino pays 7-to-6 instead of the true odds of 6-to-5, which gives the house a 1.52% edge.

If a dice influencer can reduce the frequency of 7s even slightly, the 1.52% house edge on Place 6 erodes. The breakeven SRR is exactly 43/7 ≈ 6.14, which means one fewer 7 every ~263 rolls compared to a random shooter. That translates to P(7) dropping from 16.67% to 16.28%.

At that point, the probability of rolling a 6 before a 7 crosses above 6/13, and the expected value per bet flips from negative to positive. Go any higher, and the player has an outright mathematical advantage.

The dollar amounts remain tiny. Even at SRR 6.5 (a strong claimed skill level), the expected profit is about 22 cents per $6 bet.

Notice how the edge flips at SRR 6.14, but the dollar amounts remain small even at significant skill levels:

A dice influencer sets the dice in a specific orientation (e.g., the "hardway set" or "3V set") that keeps 7-producing faces off the primary rotation axis. They throw with a controlled arc, minimal wobble, and backspin. The goal is to land the dice just before the back wall so they make soft contact: enough to satisfy the dealer, not enough to fully randomize.

Everything depends on how much of the initial dice set survives contact with the back wall. If the wall fully randomizes the dice, the set is irrelevant and the SRR is 6.0 no matter what. The claim is that a skilled throw preserves some small residual correlation.

So what SRR can a human realistically achieve in casino conditions? The honest answer: nobody knows for certain. Rigorous data is scarce. Most tracking is self-reported, which is worthless for proof. But here's where the claims and physics live:

Why the Back Wall Makes This So Hard

The diamond pyramids aren't random bumps. They're precisely sized and angled so that no matter how the dice arrive, the contact point is unpredictable. A controlled throw with backspin and a soft landing can reduce the wall's randomizing effect, but it can't eliminate it.

If influence exists at all, its realistic range is probably SRR 6.1 to 6.5. Not because practitioners lack skill, but because the back wall is specifically engineered as a ceiling on how much influence is possible.

What Degrades the Throw

• Distance to back wall. Most tables are 12 feet long. The farther you throw, the more variables accumulate.
• Table surface. Bouncy felt amplifies randomness. Dead (worn) felt is better for control.
• Fatigue. After an hour, muscle fatigue erodes precision. After two hours, most practitioners admit their SRR drops toward random.
• Casino heat. If the pit boss suspects controlled shooting, they may ask you to throw differently or leave.
• Inconsistency. Even a good shooter doesn't hit their mark every throw. Maybe 60% of throws are "on axis" and the rest are random.

The Data Problem

Stanford Wong tracked some graduates of a dice-control course and observed results in the SRR 6.2 to 6.5 range, but the sample sizes were too small to be statistically conclusive. Most other "evidence" is self-reported by people who sell dice-control courses.

Dice influence lives in exactly the SRR range where it's nearly impossible to prove or disprove statistically. An SRR of 6.3 means one fewer 7 about every 126 rolls. To prove that's not random chance, you'd need to track tens of thousands of rolls under controlled conditions. Nobody has done this rigorously in a casino setting.

• The edge is tiny. Even a genuine dice influencer at SRR 6.3 is making about 10 cents per bet. At $6 Place bets, that's $10 per 100 bets. Compare that to the hundreds they might win or lose from pure variance in the same session.
• Variance makes it invisible. The casino can't distinguish a skilled shooter from a lucky one without tracking thousands of rolls.
• Time and table limits. Craps is slow. You might resolve 30 to 40 bets per hour. At SRR 6.3 and $6 bets, your expected hourly profit is about $3.
• Most players bet poorly. Even a player with genuine dice influence often makes high-house-edge bets (hardways, propositions) that wipe out any edge from controlled throwing.

The realistic range of SRR 6.2 to 6.5, even if achieved, sits in a no man's land where the edge may be real but the results look indistinguishable from a lucky player.

The "would bankrupt casinos" argument imagines dice influence as a switch: either you can control the dice or you can't. It's more useful to think of it as a dial. Even if you can turn it, the back wall means you can barely nudge it. And a barely-nudged dial, filtered through the noise of variance, looks exactly like randomness. That isn't a disproof. It's a statement about the limits of what we can observe.