calculate baccarat win rate and lose rate by computation theory without distribution preset or any knowledge for it.
// by go run env
go build calculation
// run it by redirect standrad output
./calculation 1>result.log
...
11 11 13 13 3.333738451068494e-05 atimes 0 btimes 1
11 12 13 13 6.882556802205923e-05 atimes 0 btimes 1
11 13 13 13 6.452397002068053e-05 atimes 1 btimes 2
12 12 13 13 3.333738451068494e-05 atimes 0 btimes 1
12 13 13 13 6.452397002068053e-05 atimes 1 btimes 2
13 13 13 13 2.9237423915620866e-05 atimes 2 btimes 3
all types sum to 0.9999999999999297
the root types of base card range is 8281
the result win is 0.4585974226327612 chargerlose 0.4462466093436007 chargerdie 0.09515596802364121
sum is 1.000000000000003
For reason why sum is beyond 1 , you can personally remake data structure with more precise float, but it's fair enough for precision at 10^(-14) as mentioned at WIKIPEDIA