How the dice roll tracker works
There are two ways to log. The quick way is one tap per total: the shooter rolls a 9, you tap 9. The detailed way is entering each die, 4 and 5, which keeps hard and easy numbers apart in the log and in the export. Either way, the chart and the counts update with every entry.
The quick mode suits a live casino table, where rolls come fast and you only glance at your phone. The detailed mode suits a practice session at home or an online game, where you can take a second per roll. Pick the mode that fits the table in front of you.
Everything you enter stays in your browser’s own storage on your device. There is no account and nothing goes to a server, so the log is there when you come back, until you clear your browser data. Before any fresh start, the export button saves the current log as a CSV file you can keep for good.
Reading the chart against the 36-way expectation
Two dice make 36 equally likely combinations, and the chart sets your results against that shape. The 7 should lead with six ways, the 6 and 8 follow with five each, and the 2 and 12 trail with one each. Your bars sit beside the expected ones, so any total running hot or cold stands out at a glance.
Short logs look lumpy. That is normal, and it fools people. With 100 rolls you expect under three 12s, and seeing none or six of them is entirely ordinary. The chart only starts to resemble the expected pyramid after several hundred rolls, which is the first lesson the log teaches. Our guide to craps odds explains the 36 combinations.
| Total | Ways out of 36 | Chance | Expected in 360 rolls |
|---|---|---|---|
| 2 or 12 | 1 | 2.78% | 10 each |
| 3 or 11 | 2 | 5.56% | 20 each |
| 4 or 10 | 3 | 8.33% | 30 each |
| 5 or 9 | 4 | 11.11% | 40 each |
| 6 or 8 | 5 | 13.89% | 50 each |
| 7 | 6 | 16.67% | 60 |
The sevens-to-rolls ratio and dice control
Players who practice dice control watch one number above all: how many rolls, on average, come between sevens. Fair dice give one seven in six rolls. A shooter who could really keep the 7 away would show a longer gap, one in seven or one in eight, and the roll log tracks that ratio live as you enter rolls.
The ratio is easy to fool yourself with. A hot night of twenty rolls without a seven feels like proof, but a run like that happens to fair dice regularly. What matters is the ratio over a long log, preferably one kept before you decide what it should show. The tracker keeps every roll, not only the memorable ones.
Dice setting and how you throw
Dice setting means arranging the dice before the throw, and dice-control guides pair it with a soft, even toss. You can record each set in your own notes and log the throws in separate sessions, one per set. Then the question of the best dice set for craps becomes a comparison of real counts instead of opinion.
How you roll dice in craps is fixed by the rules: one hand, down the length of the table, both dice off the back wall. That wall, with its raised diamond pattern, exists to scramble the dice. Any dice setting technique has to survive it, which is exactly why a long, honest log is the only real test.
The chi-square fairness test, in plain English
Once your log has enough rolls, the tracker runs a chi-square test. It compares the count of every total with its expected count, all eleven at once, and adds up how far they stray. Then it asks one question: would fair dice produce a spread this uneven often, or rarely? The verdict comes in a sentence, not a formula.
With eleven totals the test has ten degrees of freedom, and a score above about 18.3 happens with fair dice only one time in twenty. Below that, the verdict says the rolls look consistent with fair dice. Above it, it says the spread is unusual, which is a reason to keep logging, not a proof of anything.
That last point matters. Test enough logs and one in twenty fair sets will look unusual by chance. A single high score from a single session is weak evidence. A high score that stays high as the log grows from 500 to 2,000 rolls is much stronger, and the running test lets you watch the score as rolls are added.
How many rolls you need before the test means anything
The chi-square test needs every total to be expected at least five times, and the 2 and 12 arrive only once in 36 rolls. So the floor is 180 rolls. Below that, the tracker still charts your rolls, but treat any verdict with caution, because the math behind the test is not reliable on a smaller sample.
The sevens ratio needs far more. After 360 rolls, fair dice will usually show a seven somewhere between once every 4.9 and once every 7.9 rolls. After 3,600 the range narrows to between 5.6 and 6.5. A claimed one-in-seven shooter needs about 1,500 logged rolls before the difference stands out reliably from ordinary luck.
| Rolls logged | Usual range for fair dice (one seven per …) |
|---|---|
| 360 | 4.9 to 7.9 rolls |
| 1,000 | 5.3 to 7.0 rolls |
| 3,600 | 5.6 to 6.5 rolls |
Those figures explain why dice-control debates rarely end. Most claims rest on a few hundred remembered throws, where fair dice can land almost anywhere in the table above. Logging 1,500 throws is dull, but it is the only way to separate a real edge from a lucky weekend, and the tracker keeps count so you do not have to.
Is bubble craps rigged? Logging machines and online games
Bubble craps machines in licensed casinos use real dice shaken by air under a dome, and they must be approved by the state regulator and tested by independent labs before going on the floor. That makes a rigged machine at a licensed venue very unlikely. If you still want to check, log a few hundred rolls and run the test.
The same goes for online craps. Licensed games in New Jersey, Michigan, Pennsylvania and West Virginia are tested and approved by each state’s regulator, and live-dealer tables roll real dice on camera. Offshore sites answer to no US regulator, which is one reason to avoid them. Our page on bubble craps covers how the machines work.
Keeping a long log
A useful log is a long one, so plan for several sessions. Export to CSV at the end of each, and note the date, the table and whether you set the dice. The CSV opens in any spreadsheet, and you can add up sessions yourself or check the tracker’s numbers with the odds calculator next to them.
At a live table, log the come-out rolls too, since they decide whether the Pass Line wins at once. The come-out roll should produce a natural 22.22% of the time and craps 11.11%. If your log shows far more naturals over a thousand come-outs, what else changed about the way you throw?