How to Play

While it is possible to play solo, you and a distant friend can decide on a ruleset, a RNG, and a seed and then decide who will be A and who will be B. Each player can record their selections and the outcome of their own selection. After making a selection, or to begin with a new set of selections, click the "Begin with new seed" button; this will clear all information in all counters and start a new random number generator at step 1. If the step count between the two players is off, the correlations are not true, so the step number should also be noted when testing with a friend.

After both players have made their selections, they can click the 'Get Results' button to proceed to the next result step.

Rulesets may be:

CHSH Test angles - these use the experimental test angles of 0, 22.5, 45 and 67.5 degrees. This is not the ideal score; because it loses points in the 4th part for excessive correlations.
Ideal Test Angles - uses 0, 30, 60, 90 degrees for the angles; this provides 100% non-correlation on the 4th setting, except by random chance; the other 3 values are only 80% correlated though.

RNG Choices: (implemented from)

SFC32 - Simple Fast Counter
MUL32 - Mulberry
XOR32 - xoshiro 128
JSF32 - Jenkins Small Fast

Seed: May be any string the two agree on; the string content is bit shuffled to generate a random start point in the random number generator.

Result Table meanings

The result table is somewhat abbreviated to fit in a smaller space.

same - when the two detectors are selected, the results were correlated.
diff - when the two detectors are selected, the results were different.
ratio - calculation of abs(same-diff)/(same+diff); where same+diff is the total results.

The proposed Bell samples then would have to be 3:1 3-1/(3+1) = 2/4 = 50% .

...

The scoring of this game is generous; and I'd pit myself against QM predictions with this ruleset applied to its data (if there was a similar experiment). The first place I heard about the game (the See Also page below), it was described as an ideal 4 point game. If the 4th term's points can only ever subtract, then the ideal is only 3, and a penalty from 3 for excess matches. This Game also counts excessive different results as a correlated negative result.

Other Information

This isn't much of a game itelf, so this keeps the running statistics for every set of choices both players might make, and reports on the number of correlated and non-correlated results for each of those settings.

See Also; very long discussion of the idea and math of this Local Hidden Variable (LHV).

also Spin Probabilities markdown that is fairly chaotic, and starts with some resources used to understand QM math.