Research
The working
There are real games inside Second Chance — a snooker table, a roulette wheel, a lad on a corner who plays you for money. Each of them had to be built properly rather than approximated, and each turned up something worth writing down.
These pages are the working: what was measured, where the numbers came from, what the evidence actually says, and what is still unproven. They are not marketing and they are not the devlog, which tells the story of a day. If a page here cannot say what it is uncertain about, it does not belong on this list.
-
Snooker · physics and geometry
Making the snooker real
A full-size table measured against the WPBSA rulebook to the millimetre, and a ball model with sliding, rolling, spin and ball-on-ball throw, verified in 61 checks against a reference implementation. Includes the three numbers everybody quotes that the rulebook does not actually specify, and a pocket that measured 9.382 cm at one end and 0.031 cm five centimetres later.
Cushion restitution · the slide-to-roll integration · positive controls -
Snooker · the rules
The rules of snooker, checked
The British game as the governing body actually states it. We gathered 298 sourced claims and put the 40 weakest to a reviewer whose job was to refute them — 33 came back corrected and one was refuted outright. Foul values are a maximum and never a sum; the last red does not end the reds; conceding turns on penalty points rather than snookers.
Free ball · re-spotting · the miss rule · what a referee actually says -
Roulette · fairness
Proving a roulette wheel is fair
A real ball in a real bowl, with nothing choosing the number — which means fairness has to be measured rather than asserted. Seven of our eight randomness tests missed both realistic failure modes; only one caught them. The honest position is that the design is proven fair and the wheel’s own data is still accumulating.
Every bet is −2.7027% · uniform is not random · the positive controls that found bugs in the tests -
Rock paper scissors · strategy
How to win at rock paper scissors
It is not a game of chance when humans play it. The biased opening throw, the win-stay lose-shift pattern and the population-level cycling that comes out of it — and the prediction engine behind Jack, who starts a match knowing nothing about you and gets harder as he learns.
The Nash trap · Markov prediction · why an opponent has to be made worse on purpose
A standing rule for this section. Every page here names its sources and marks what it could not verify. Where a widely repeated figure turned out not to be in the paper it is attributed to, we have published what the paper says instead and explained the difference. Where we have not measured something, we say so rather than reaching for a number that sounds right.