Quite interesting:
TheGamblist ran a simulation over the last 500 to test the profitability of different cash out points, these were his results:
1.01 = 99% 1.05 = 97% 1.11 = 97% 1.21 = 98% 1.31 = 97% 2.50 = 83% 4.00 = 91% 4.80 = 104% 5.00 = 106% 5.10 = 109% 5.25 = 112% 5.35 = 111% 5.50 = 104% 6.00 = 102% 10 = 82% 13 = 97% 14 = 112% 14.5 = 105% 15 = 108% 16 = 100% 20 = 95% 50 = 87%
Then taking the theory into practice, he started wagering on the profitable cash out amount:
and then with another account going for the other one:
In somewhat of a panic, I wrote up my own little simulation:
function genGameCrash() {
if (Math.random() < 0.01)
return 0;
var r = Math.random();
var perfect = (1 / (1 - r));
var houseEdge = (perfect-1) * 0.01;
var multiplier = perfect - houseEdge;
return Math.floor(multiplier * 100);
}
for (var p = 101; p < 2000; p += 1) { // which cashout points to check (101 = 1.01x )
var balance = 0;
var runs = 1000000; // how many wins at each point (amount of wins)
for (var i = 0; i < runs; ++i) {
balance -= 100;
var cp = genGameCrash();
if (cp >= p) {
balance += p;
} else {
++runs; // we didn't win, so lets increase
}
if (cp != 0)
balance++; // add the bonus
}
console.log('[Over ', runs , ' games] Cashing out at ', p/100, ' gives ', balance/i, ' per game');
}
but I'm unable to reproduce the results TheGamblist is seeing, yet i'm oddly nervous (hence posting this here)
