Investigations 4 min read

Chicken Road predictor: what the hacks and bots really sell

A Chicken Road predictor claims to know where the next oven is. It comes as an APK, a "Chicken Road hack bot" on Telegram, or a paid signals channel with screenshots of big wins. None of them can work, and the reason is built into the game.

We explain why in three steps: how a round is decided, what a "signal" would need to know, and what happens when you test the usual signals on 10,000 simulated rounds.

Chicken Road · on your phone

Plays in the browser,
nothing to install.

Chicken Road runs in a phone browser, in English, with all four difficulty levels and the Play, Go and Cash out buttons under your thumb. There is no official Chicken Road app, so there is nothing to download — and any "Chicken Road app" in a store is not from InOut Games.

The partner casino is not licensed in Australia. Gambling with money is 18+ and can be addictive: National Gambling Helpline 1800 858 858, free and confidential.
DEMO ON A PHONE Official Chicken Road demo on a phone in English, Australian dollars, with the Cash out and Go buttons

How a Chicken Road round is decided

The game's "Provably fair settings" panel says it plainly: the round result is determined from the server seed combined with the first three bets of the round. Before the round, the game shows you the SHA256 hash of the next server seed. After it, you can check from your bet history that the seed matches the hash.

That has two consequences for any predictor:

  1. The outcome is fixed before you press Go, and the server has committed to it publicly by hash. Nobody can change it mid-round.
  2. Nobody outside the server knows the seed until the round is over. A SHA256 hash cannot be reversed to reveal it.

A predictor would need the server seed in advance. If someone had it, they would not be selling it for A$30 on Telegram.

The usual "signals", tested

Signals channels sell rules based on past rounds: "go after three losses", "stop after a long road", "the oven repeats its lane". We simulated 10,000 rounds on Medium with the model the whole site uses (3 ovens among 25 slots, every placement equally likely, seed 20261116) and measured how often the chicken reached lane 5 after each "signal".

Bar chart of the share of rounds reaching lane 5 on Medium after each signal, all between 48.8% and 51.3%, around the model's 49.6%
Every "signal" lands within a couple of points of the model's 49.6%: that is noise, not information
Signal Rounds Reached lane 5
All rounds 10,000 49.9%
After a win 4,984 49.8%
After a loss 5,015 49.9%
After 3+ losses in a row 1,255 48.8%
After 3+ wins in a row 1,251 51.3%
After a long road (10+ lanes) 1,967 50.0%

The model says 49.6%. The signals scatter a point or two either way, which is what random samples of a thousand rounds do. Run the same test with another seed and the "best" signal changes.

"The oven repeats its lane"

Another favourite: the first oven tends to land where it landed last time. In the simulation the first oven repeated its lane in 7.3% of consecutive rounds; pure chance predicts 7.5%. The pattern people "see" is the one chance produces by itself.

Simulated rounds, not recorded ones

These are rounds simulated from the model, not rounds recorded from the demo; we say it plainly because the method matters more than the sample. The model fits every multiplier the official demo sends and every maximum InOut publishes. If the real rounds were predictable from the past, the provably fair hash would be pointless, and anyone could prove it by checking their own bet history.

What predictor apps actually do

  • Show random numbers that are right about half the time on Medium, by construction.
  • Collect your casino login, under the pretext of "connecting" to your account.
  • Ask for a subscription, then a "VIP upgrade", then disappear.
  • Install something else: an APK from outside the store can do anything Android lets it.

If you paid for one, the steps are on our page on is Chicken Road a scam. If what you want is to understand the odds, the Chicken Road odds page has every lane, for free.

FAQ about Chicken Road predictors

Does a Chicken Road predictor work?

No. Each round is fixed by a server seed you cannot see until it is over; its hash is shown in advance and cannot be reversed.

Is the Chicken Road hack bot on Telegram real?

It is real as a bot and fake as a predictor. Its "signals" are right about as often as a coin flip on Medium.

Is a Chicken Road predictor APK safe?

No. It is software from outside the store that asks for logins or payments; uninstall it and change any password you typed into it.

Can past rounds help predict the next one?

No. In our 10,000 simulated rounds, no rule based on past rounds moved the hit rate meaningfully away from the model's 49.6%.

Keep reading

← All blog posts