Somewhere in 100 spins
84.58%At least one qualifying run, counting overlapping possibilities correctly.
STREAK LAB · 18+
How surprising is a losing streak?
A rare start can become a common event across a longer session. Explore the difference.
Set up an independent-spin experiment
Fair European roulette: 18 red, 18 black, one zero. We follow a fixed choice of red: black and zero both count as losses. There are no stakes, bankrolls or prizes here.
The default probabilities below are ready to read. Explore to generate an example sequence.
At least 5 losses in a row
In the first 5 spins
3.57%A single specified starting position answers a different question.
Next spin: red still has an 18/37 = 48.65% chance. A previous streak does not make red due, and it does not predict another loss.
These are exact model probabilities evaluated numerically and displayed rounded. Very high probabilities appear as >99.99%, not a guarantee. They are not estimated from the example sequence and do not change when you choose another seed.
Your reproducible sequence
Inspect the numbered spins
W = red, L = black or zero. Colour is also shown in each spin's description. The outlined blocks mark the first longest losing run in this example.
Run the same settings and seed again to reproduce this example, or choose another seed to see a different illustration. A single selected example cannot validate a method or a real wheel.
Method, limits and sources
The calculator tracks each possible trailing-loss count and absorbs a sequence only when it first reaches the chosen run length. This prevents double-counting overlapping runs. The first-position probability is (19/37)k; multiplying it by the number of possible starting positions is not an exact probability of at least one run.
The illustrative sequence uses the Lab's deterministic seeded generator. It is for reproducible education, not a certified real-money random generator. Both the calculator and example assume independent uniform pockets; they do not test whether a particular casino or wheel is fair. A streak alone does not establish unfairness.
No Martingale funding, ruin, table limit or early-stop model is calculated. Previous outcomes and the seed cannot reveal the next real spin. With consent, Analytics counts experiment runs and exploration of different session/run settings; seeds and results are not sent.
Primary research: outcome runs and roulette predictions · UKGC: random outcomes