- Home
- How We Test Chicken Road Across 12,000 Demo Rounds
Our test
Inside Our Chicken Road Demo Recording Method
We captured the provider demos until a loss or the final crossing and published the retained rows and their limits.

Choose game and difficulty
Keep the eight groups separate.
Capture provider demos
Store round outcomes with dates.
Check each captured attempt
Exclude attempts disturbed by shared activity.
Publish rows and calculations
Share counts, uncertainty and reusable files.
The people and demo behind this capture
We are Bet Chicken Road. We captured the InOut Games browser demos of Chicken Road and Chicken Road 2, using virtual credits throughout the work. The game windows came from the provider, while the collection and presentation of observations are ours. We did not place these sessions through a casino cashier or make deposits to obtain the results.
Our purpose was to describe how far captured attempts progressed under a clear rule. The files keep the game and difficulty attached to the observations, so readers can trace a chart back to its matching group. Dates shown beside the data identify the recording period. Those dates matter when comparing a saved copy with a later visit.
The site name identifies who assembled these observations. It does not imply employment by InOut, permission to speak for a regulator or independent certification of gambling software. We publish the method so you can assess what the rows support before using them in your own work.
A fixed stopping rule across all eight groups
We recorded 1,500 retained rounds for each combination of game and mode. Four modes across two games account for all 12,000 recorded rounds. We keep these groups separate instead of making a pooled curve from roads with different lengths and multipliers. Select the group matching the game and setting you want to examine.
For the survival recordings, each attempt continued until a loss or completion of the road. No early cash-out was chosen. A successful move contributes to the reached-step count; the unsuccessful move ending a round does not count as another reached position. An opening collision therefore contributes no successful crossing. Progress already achieved remains in the capture when a later move fails.
This stopping rule lets us observe later positions when the game permits them. A player who collects early makes a different choice, so the survival cohort cannot describe the outcome of that player’s strategy. Keep the rule beside any calculation you reproduce, especially when a file label alone would leave the stopping behaviour unclear.
Shared-demo interference is left outside the calculation
The provider demo can be shared with another visitor. If that activity disturbs a recording and prevents us from assigning its actions or outcome confidently, we drop the affected attempt. We do not decide whether it would probably have won or lost. An uncertain association is insufficient support for either result.
The page widget displays how many attempts were excluded from this site’s collection. Read that live count alongside the retained totals rather than borrowing a number from another website. Attempted captures and usable rounds answer different questions. The public retained-round files leave the discarded attempts outside the statistics.
This rule concerns whether a recording belongs to the attempt being captured. It is not a way to discard ordinary losing rounds or improve the appearance of a curve. Clear losses remain observations. Removing interference also cannot eliminate every limitation of a browser recording; it makes one specific source of uncertainty visible.
Reading a survival share and its interval
At any position, survival is the share of retained starts that reached that many successful crossings or more. The denominator stays the full starting group. It does not become only the rounds reaching the preceding position. A curve therefore answers how far starts progressed, rather than the conditional chance of the next crossing after arriving there.
The shaded range is a pointwise Wilson 95% confidence interval. Under its sampling assumptions, repeating the procedure would give intervals covering the underlying proportion at the stated long-run rate. That interpretation concerns the procedure, not a promise about your next round. It also does not assign a personal winning chance to an individual player.
Few attempts may reach the later positions. An empty count is not proof that the position is impossible, while a captured success is not proof it is easy. Read the interval with the share and group size. Each interval concerns its own position; the separate ranges do not form a simultaneous guarantee for the entire curve.
What you can reproduce and what remains outside the test
These recordings do not predict future crossings, certify the game or audit a casino’s random-number generator. Demo observations cannot verify real-money payment behaviour. A large reached multiplier also need not be a paid prize: the same attempt may have continued and lost. For theoretical returns, consult the provider RTP explanation separately.
Use the download area to obtain the CSV rows and accompanying notes. Choose the same game and mode as the chart, read the field definitions, then count reached positions from retained starts. Keep the recording dates and stopping rule with a table you create. That makes the calculation understandable to a reader who did not see this page.
Our round data is reusable under CC BY 4.0. Credit Bet Chicken Road, link to the source and licence, and say what you changed. You can publish a recalculation or a chart with that attribution. Data reuse does not transfer ownership of InOut’s artwork.
Adults 18+ only: set time and money limits, take breaks and never chase losses.
Take the data with you
Download rows and read field definitions. Reuse the data with attribution.
Captured by: Bet Chicken Road · CC BY 4.0. Credit Bet Chicken Road. Link the licence. Explain any changes. InOut retains its game artwork.