A pattern needs a fresh test
A wheel-bias hypothesis is a claim about a physical wheel, not a claim that a number is “due”. Record consistent observations, choose a prediction from earlier data and check it against later spins. Do not keep changing the prediction to match results already seen.
The five-pocket baseline
On a fair European wheel, a fixed selection of five pockets hits with probability 5/37, or about 13.51%. Across 100 spins, that is about 13.5 hits on average—not a promise of that result in every hundred.
If you stake one unit straight-up on each of those five pockets, the total stake is five units. A hit returns 36 units including the winning stake, leaving 31 units profit after the other four lose. A miss loses five units. Expected profit is therefore 36p − 5, where p is the combined hit probability. Break-even is 5/36, about 13.89%, before any other costs.
Why 16 hits is not enough on its own
Sixteen hits in 100 spins gives an observed hit rate of 16%. That is above the break-even point, but it is only a sample estimate. Its uncertainty matters; choosing the best-looking five pockets after seeing all 100 spins also biases the comparison.
Keep a separate later test. Fix the selection method before measuring its performance, record misses as well as hits and keep different wheels separate. EdgeLab21 shows validation results and a conservative confidence bound rather than treating the observed percentage as certain.
What the tool records
The roulette tool records start pocket, landing pocket, direction and wheel type. Its offset calculation uses physical wheel order. Try the public sample to see why moving from 0 to 21 on a European wheel is five clockwise pockets, not twenty-one.
These observations apply to physical wheels, including live-dealer wheels. An RNG game does not supply the same physical-wheel measurement.
Frequently asked questions
Can a red or black streak prove bias?
No. A streak alone does not establish a persistent, exploitable physical-wheel pattern.
Why test with later spins?
It reduces the risk of reporting a pattern that was selected only because it fitted the data already observed.
Research foundation
These primary references provide historical or mathematical context. EdgeLab21’s product outputs remain estimates and should be independently validated for the exact rules and conditions being analysed.