mustangsally
Joined: Mar 29, 2011
• Posts: 2463
February 19th, 2014 at 1:40:03 PM permalink
Quote: endermike

1) Using flat betting on the banker and quitting when 1 full unit (or more) up or 100 hands have been played
2) Based on 10,000,000 independent attempts (about 1200 seconds run time)

Prob of winning (quitting with 101): .8372

1 full unit?
a win on a \$20 bet nets \$19 after the \$1 comm
I ran a 101 unit bankroll (2020) sim and stopped when the bankroll hit 2039 or 100 trials
that seems more than fair to quit after the first win.

here is the value calculated using a transition matrix for success = 88.483079%
from a quick 1 million sim I get 884,658 attempts that were successful = 88.4658%
`                       grouped dataitems:                  1000000   minimum value:             1073.00first quartile:            2039.00median:                    2039.00third quartile:            2051.00maximum value:             2057.00mean value:                2015.44midrange:                  1565.00range:                      984.00interquartile range:         12.00mean abs deviation:          51.44sample variance (n):       8496.01sample variance (n-1):     8496.02sample std dev (n):          92.17sample std dev (n-1):        92.17`
the OP may still be asking
"how do you calculate this?"

either simulate and/or use a spreadsheet or a program to raise a transition matrix up to any power (number of trials)

I think the OP may too be bummed out that the probability is not closer to 99.95% or even higher.

just remember that the longer one plays against a house edge game, the less chance that player has of coming out even or ahead
in other words
the chances increase with every trial that the player will not be even or ahead no matter how many trials are completed each day

Sally
I Heart Vi Hart
endermike
Joined: Dec 10, 2013
• Posts: 584
February 19th, 2014 at 1:44:12 PM permalink
Quote: mustangsally

1 full unit?
a win on a \$20 bet nets \$19 after the \$1 comm

Yup, it means a win on the first hand isn't enough. I just had the sim pay .95 of a unit for a win and lose 1 full unit for a loss.
mustangsally
Joined: Mar 29, 2011
• Posts: 2463
February 19th, 2014 at 1:52:22 PM permalink
Quote: endermike

Yup, it means a win on the first hand isn't enough. I just had the sim pay .95 of a unit for a win and lose 1 full unit for a loss.

OK,
OP has now two sets of sim data for Banker bet

and I think most Bac players would disagree that
the Player bet has the higher probability of winning just 1 unit over 100 hands played than the Banker bet
even with paying the comm

I did at first
Sally

my matrix shows with a 2020 bankroll betting \$20 and trying to hit a target of 2040
success = 84.471064%

latest:
how about something I think would not work well at hitting a 1 unit win

Single 0 roulette
betting 2 columns (or sections) so bet 2 units to win 1
not much variance here, my first guess was yuck, but a nice win probability

It does even better than Craps Don't Pass or Pass or Player bet in Bac
I had to check it twice and sim it too.
interesting
91.549408%
100 spins max, bankroll of 201 units and target goal of 202 units
I Heart Vi Hart
GWAE
Joined: Sep 20, 2013
• Posts: 9854
February 19th, 2014 at 5:41:58 PM permalink
Quote: zenith27

i had thought of that idea..i would just wait for a 5-streak player hands and bet on banker in baccarat. I would invest 10k as my bankroll and win limit is 1k with loss limit 5k..after win, i could go home and comeback tommorow. thats 30k a month.means tripple your investment in
1 month

varmenti?
Expect the worst and you will never be disappointed. I AM NOT PART OF GWAE RADIO SHOW
Walkinshaw30t
Joined: Apr 11, 2013
• Posts: 91
February 19th, 2014 at 8:06:40 PM permalink
Quote: mustangsally

Single 0 roulette
betting 2 columns (or sections) so bet 2 units to win 1
not much variance here, my first guess was yuck, but a nice win probability

It does even better than Craps Don't Pass or Pass or Player bet in Bac
I had to check it twice and sim it too.
interesting
91.549408%
100 spins max, bankroll of 201 units and target goal of 202 units

That is interesting- what was the average loss?
Time will tell
Walkinshaw30t
Joined: Apr 11, 2013
• Posts: 91
February 19th, 2014 at 8:08:52 PM permalink
Quote: endermike

Yup, it means a win on the first hand isn't enough. I just had the sim pay .95 of a unit for a win and lose 1 full unit for a loss.

Yes that is true, I was just thinking that if you were 1 banker win infront late in the spins the commission would have eaten up any profit
Time will tell
mustangsally
Joined: Mar 29, 2011
• Posts: 2463
February 20th, 2014 at 7:58:35 AM permalink
Quote: Walkinshaw30t

That is interesting- what was the average loss?

first something to compare against
recall the Player bet in Bac 101 unit bankroll and 1 unit win goal and 100 max bets
my calculated values - faster and the most accurate using a transition matrix
(verified with a simulation - slower and the least accurate)

Prob of winning (quitting with 102): 0.907286313 <<< this value is much higher when removing the time factor as I pointed out in an earlier post
Prob of pushing (quitting with 101 after 100 hands): 0.001546102 (1 in 647)
Prob of losing (quitting with less than 101 after 100 hands): 0.091167586
Avg loss, given you quit with less than 101 after 100 hands: -12.37590892 units
expected value on this strategy: -0.220995423 units

Now the Single 0 Roulette making 2, 2to1 type bets (columns or sections with prob of winning at 24/37)
Bankroll 201 and target goal at 202
Prob of winning (quitting with 202): 0.91549408
Prob of pushing (quitting with 201 after 100 hands): 0.00 <<< either lose 2 units or win 1 unit
Prob of losing (quitting with less than 201 after 100 hands): 0.08450592
Avg loss, given you quit with less than 201 after 100 hands: -13.97135554 units
expected value on this strategy: -0.18066225 units

use (y)our wings and fly!
Sally
I Heart Vi Hart
Walkinshaw30t
Joined: Apr 11, 2013
• Posts: 91
February 20th, 2014 at 8:57:14 AM permalink
I wonder what the B E H combo on roulette would fare under the same rules?
Time will tell
mustangsally
Joined: Mar 29, 2011
• Posts: 2463
February 20th, 2014 at 9:22:02 AM permalink
with those type of bets (combining even money bets for 18/37)
now 3 units are required to be wagered and one can lose 3 or 1 or win 1 or 3 and actually overshoot the 1 unit win goal (301 bankroll)
that should lower the probability of at least a 1 unit win and it does to just 89.183220%
and raises the avg loss to -18.56314162 units

Sally
I Heart Vi Hart
Walkinshaw30t
Joined: Apr 11, 2013
• Posts: 91
February 20th, 2014 at 8:04:11 PM permalink
Quote: mustangsally

with those type of bets (combining even money bets for 18/37)
now 3 units are required to be wagered and one can lose 3 or 1 or win 1 or 3 and actually overshoot the 1 unit win goal (301 bankroll)
that should lower the probability of at least a 1 unit win and it does to just 89.183220%
and raises the avg loss to -18.56314162 units

Sally

Sorry I should have explained myself, I mean betting 1/3rd of a unit on each.
Time will tell