## Poll

 1 in 2.18 4 votes (40%) 1 in 8 1 vote (10%) 1 in 18 5 votes (50%) 1 in 58 No votes (0%) 1 in 108 No votes (0%) 1 in 1 million No votes (0%) 1 in extremely rarely No votes (0%)

10 members have voted

mustangsally
• Posts: 2463
Joined: Mar 29, 2011
March 31st, 2016 at 8:06:49 AM permalink
this is part 2 from
https://wizardofvegas.com/forum/gambling/craps/25403-double-1-000-into-2-000/

the poll is for fun like the other was
some will be surprised of course (like me!)
<<<>>>
poll question
is
what is the chance to
turn \$1000 into \$2000 betting the pass line using the Marty betting system.
<<<>>>
the original turn \$1000 into \$2000 flat betting the pass line
turned out to be very boring

the Marty!
we all love Marty
hehe

1st bet = \$10 on the pass line
when lose bet \$20 or double the last losing bet
no odds... huh
comps will be better than flat betting
when win a bet, go back to basic \$10 bet

if can not bet double the last losing bet
bet it ALL!
Yahoo Yahoo!!

maybe it takes less time too (1+1)
and more can win more money
with Marty!

Mo Mo Mo
Go Go Go
Sally
I Heart Vi Hart
TwoFeathersATL
• Posts: 3616
Joined: May 22, 2013
March 31st, 2016 at 9:18:46 AM permalink
What if you jump from \$10 loss to \$100 bet? Then from \$100 bet to \$300 bet, assuming loss again. Then from \$300 bet to \$800 bet? You lost 100+300+800, jhit happens. Start over, try that a gazillion times or so. I would do it but I don't know how ;-( cheers, just 2F...
Youuuuuu MIGHT be a 'rascal' if.......(nevermind ;-)...2F
ThatDonGuy

• Posts: 6405
Joined: Jun 22, 2011
March 31st, 2016 at 10:27:07 AM permalink

If my equation solver is working properly, I get 45.620221%, or about 1 in 2.192.

mustangsally
• Posts: 2463
Joined: Mar 29, 2011
March 31st, 2016 at 11:28:18 AM permalink
the info I do have (due have) on the Marty for this poll is small but I like the average number of rolls to complete the experiment-challenge. looks to be abouts 546 rolls

from wincraps classic (goes good with skittles)
I Heart Vi Hart
Steen
• Posts: 126
Joined: Apr 7, 2014
March 31st, 2016 at 1:54:12 PM permalink

`If Beginning new session Then   bankroll = 1000 :   autotake full odds = falseEndIfIf passline loses Then   Bet last passline * 2 on passlineElseIf comeout roll Then   Bet \$10 on passlineEndIfIf bankroll < 10 Or bankroll >= 2000 Then   start new sessionEndIf`

Last edited by: Steen on Mar 31, 2016
OnceDear
• Posts: 7481
Joined: Jun 1, 2014
May 5th, 2016 at 11:46:08 PM permalink
Quote: mustangsally

this is part 2 from
what is the chance to
turn \$1000 into \$2000 betting the pass line using the Marty betting system.

Oncedear rule of thumb applies very well indeed to low house edge, even money bets using Marty with a small base unit.

P=Initial Bankroll/(Target Profit+Initial Bankroll) . Approx.
P=1000/(1000+1000)
P=50%
1 in 2 (and a bit for house edge)

FAR FAR BETTER than 1 in 18 !!!!!!
Psalm 25:16 Turn to me and be gracious to me, for I am lonely and afflicted. Proverbs 18:2 A fool finds no satisfaction in trying to understand, for he would rather express his own opinion.
mustangsally
• Posts: 2463
Joined: Mar 29, 2011
June 27th, 2018 at 10:55:00 AM permalink
Quote: ThatDonGuy

If my equation solver is working properly, I get 45.620221%, or about 1 in 2.192.

I was going to add my calculation from Excel (I think I lost it in a different folder, in other words I lost it!)
so as I have yet to convert Excel to online R use
*****
my sim data (in Excel) shows
Went Bust: 54.16200%
Hit Target: 45.83800% <<< 1 in 2.18
avg bet = \$36.17

take that average bet and make it \$40
flat betting \$40 now shows this (remember 0 odds) (turning 25 units into 50)
> gambler.ruin(25,25, 244/495, "Craps pass line 0x odds")
[1] "Craps pass line 0x odds. Stake:25, Target:50"
[1] "p(goal):0.330236, p(ruin):0.669764"
[1] "mean Trials:600.237, mean given Goal:600.237, mean given Ruin:600.237"
section 2r.

p(goal):0.330236 is about 1 in 3 on average
much better (but expected)
*****
how does this compare to 100 units into 200 units
with the free odds bet?

well, I used some R code (and a Markov chain solution)
(will post that code online as soon as it is cleaned up)
It was my 1st attempt long ago.
I do way better now
`##### suggest bankroll_target no lower than 20# Gambler's Ruin - mean time in transient states # and probabilities to target goal and ruintitle <- 'pass 345x odds. over-shoot target possible'bankroll_target <- 200 # Enter bankroll target (in units)bankroll_start <- (bankroll_target/2)   # Enter a different starting bankroll, default is 50% of target##### do not change any below herebankGainUnit <- (bankroll_target - bankroll_start)startBr <- (bankroll_start)    # for avg # of trials sum (from Q returns Index #)states <- (bankroll_target + 7)   # (target = 1st term))#example target==20;(20+7)(over-shoot target states=7)pass <- 244/495    # prob of win 1 unit flat bet (not enough for odds bet)miss <- 251/495    # prob of loss 1 unit flat bet (not enough for odds bet)SevenOut <- 196/495Point <- 134/495 ptFourLoss <- 1/9ptFiveLoss <- 2/15ptSixLoss <- 5/33nat <- 2/9craps <- 1/9target <- bankroll_target # absorbing state target goal EXACTstate2state <- (target - 1) #last state indexabsorb_states <- target # begin absorb statesrow.col_Names <- c(1:(states - 1),0)# enter transition probability matrixP <- matrix(rep(0,states^2), nrow=states, ncol=states, dimnames = list(row.col_Names, row.col_Names))# absorbing goal/ruin states = 1	for (i in target:states) {P[i, i] <- 1 	} # to ruin from 1:6 bank trying to double-up	for (i in 1:5) {P[i, states] <- miss } 	P[6,states] <- ptSixLoss # to pass - 1 to 5 bank double-up success	for (i in 1:5) {P[i, i * 2] <- pass } # Point diag starts at P[6,13]	for (i in 6:state2state) {P[i,i + 7] <- Point }# nat diag starts at P[6,7] 	for (i in 6:state2state) {P[i,i + 1] <- nat } # craps diag starts at P[6,5] 	for (i in 6:state2state) {P[i,i - 1] <- craps } # ptFourLoss diag starts at P[6,2] 	for (i in 6:state2state) {P[i,i - 4] <- ptFourLoss } 	# ptFiveLoss diag starts at P[6,1] 	for (i in 6:state2state) {P[i,i - 5] <- ptFiveLoss  } # ptSixLoss diag starts at P[7,1] 	for (i in 7:state2state) {P[i,i - 6] <- ptSixLoss } 					 	#P # uncomment to show matrix#print(formatC(P),quote=FALSE)P_sums <- as.matrix(rowSums(P))colnames(P_sums) <- list("total")#print(P_sums)# take sub-matrix for transient states to statesP.s <- P[1:state2state, 1:state2state]#P.s# take sub-matrix for transient states to absorbing statesP.t <- P[1:state2state, absorb_states:states]#P.t# compute mean number of revisits matrix SQ <- solve(diag(state2state)-P.s)#Q# sum entries in Xth row of Qsum(Q[startBr,])#bankroll_start#startBr #Index numberQ_sums <- as.matrix(rowSums(Q))colnames(Q_sums) <- list("avg trials")#print(Q_sums)QT <- Q %*% P.t#QTgoalG <- QT[1:state2state, 1:7]#goalGsuccess_target <- as.matrix(rowSums(goalG))colnames(success_target) <- list("p(target)")#print(success_target)ruinR <- QT[1:state2state, 8]#ruinRruinR <- as.matrix(ruinR)colnames(ruinR) <- list("p(ruin) ")#print(ruinR)S2Srow_names <- as.matrix(seq(1, (bankroll_target - 1), by=1 ))colnames(S2Srow_names) <- list("bank")#print(S2Srow_names)final <- cbind(S2Srow_names, success_target, ruinR, Q_sums)rownames(final) <- c()#print(final)data <- as.matrix(final[startBr,  ])colnames(data) <- list("data  ")print(title)print(formatC(data,digits=10),quote=FALSE)bankroll_targetbankGainUnit`

Full 2x odds(\$10flat wit \$25 odds on 6&8): p(target) 0.4229117 <<<lower than a Marty
345x odds: p(target) 0.4704061 <<< a bit higher than a Marty
*** note this is hitting target exactly *** most do and will not play this way
345x odds: p(target) 0.4647375755 where one can go over the target (from 20 to 26 units)
INTERESTING, I do say so
`[1] "pass 345x odds. over-shoot target possible"> print(sprintf("Bankroll target:%g, unit gain:%g",bankroll_target,bankGainUnit))[1] "Bankroll target:200, unit gain:100"> print(formatC(data,digits=10),quote=FALSE)           data        bank               100 p(target)  0.4647375755p(ruin)    0.5352624245avg trials   422.46213 [1] "pass 345x odds. Exact target"> print(data)                data  bank       100.0000000p(target)    0.4704061p(ruin)      0.5295939avg trials 413.7700891> bankroll_target[1] 200> bankGainUnit[1] 100[1] "pass Full 2x odds. Exact target"> print(data)                 data  bank        100.0000000p(target)     0.4229117p(ruin)       0.5770883avg trials 1077.0895754> bankroll_target[1] 200> bankGainUnit[1] 100[1] "don't pass 345x odds. Exact target"> print(data)                data  bank       100.0000000p(target)    0.4713885p(ruin)      0.5286115avg trials 415.9557935> bankroll_target[1] 200> bankGainUnit[1] 100`

this shows why Marty has been around for a long time
and keeps on being discovered
even to this day!

I am guessing that in truth
the Marty does have a place in the sun!
Sally
Last edited by: mustangsally on Jun 27, 2018
I Heart Vi Hart
billryan