Here are some JoB statistics for a 9/6 FP machine. Values are per coin, unless otherwise
noted.
Results are non-linear in # of hands except for EV (mean).
100 500 1000 2000 4000 Hands
mode -45 -100 -165 -290 -550
median -30 -80 -135 -245 -445
EV -0.456 -2.280 -4.561 -9.121 -18.244
10% -130 -335 -510 -795 -1255
90% 110 230 315 460 1655
Prob<EV 61% 63% 66% 69% 73%
For comparison purposes, here are some values for Pick'em:
100 500 1000 2000 4000 Hands
mode -50 -170 -285 -305 -335
median -35 -95 -115 -125 -150
Pick'em has a return of 99.95%, higher than JoB's 99.54%. Nonetheless, for a (very) small
number of hands, the most likely outcome for JoB is better than for Pick'em. This has
nothing to do with the RF cycle or its contribution (the RF has almost no effect on the
mode; results for mode are the same with or without RF). BTW, if I had the choice, I'd
always play Pick'em over JoB. For some reason though, most VP players I've spoken to
prefer JoB. (Perhaps Pick'em is too boring? )
So which is the more risky game? I guess it depends upon what you mean by "risk" and
how many hands you play.
So what is the required bank roll to play either 5 coin JoB or Pick'em with 100% certainty of
not going broke for 4000 hands? Easy: 4000*5 = 20,000 coins, since that is the most you
could ever lose after 4000 hands. (This was a trick question and it has nothing to do with
log-optimal betting or RoR)
Homework: Compare JoB to a game with higher variance and better EV, like FPDW or NSUD
.