Jean-Baptiste Queru wrote:
Cindy Liu's calculator indicates that the long-term bankroll for TDB
for a CB of 0.6% is approximately 5 times as large as that of JoB in
the same conditions.
That sits much better with me than your original "more than twice the
bankroll" statement -- way too much wiggle room on that one; sort of
like saying they should have topped off those NO levees at least a tad
before the storm. 
I'd be ready to not hit any of those hands, and therefore I'd want
to be bankrolled enough to lose 15% or more of my coin-in.
I don't have hard numbers, but my guess is that would be an equivalent
ROR stake to playing JB with funds to cover 5% of your targeted
coin-in. TDB is wicked and a short-term loss of 20% is almost a
no-brainer. You're looking at an expected loss of 27% in any hour
without something better than a FH.
As for how long it takes to get into the "long term", for my rule of
thumb of "long term" (30 cycles of the top jackpot), I was wrong.
The 800-coin top jackpot occurs a lot more often in TDB than in JoB,
with a cycle of less than 11000. My rule of thumb of "long term" is
an empirical value for which I feel confident that modeling the
results of VP with a normal distribution is accurate enough for
practical purposes.
When you consider the additional concentration of return in the other
quads as well, this well could be an isolated case where "in the (vp)
long run we're all dead" might be an apropos adage.
Sorry for harping on this one, but this is a game that no disciplined
vp player with realistic goals can take seriously, except under the
most extraordinary of circumstances.
- H.