the rule of thumb or mantra recited by the gambling experts is that in a great number of trials the player will ALWAYS come very close to losing a % equal to the House Edge

okay - but let's think about a player who plays this way betting only on the Player at Baccarat

he parlays his winning bets until he loses

for example______ bet 10 - win - bet 20 - win - bet 40 lose__________he has bet a total of $70 and lost $10____________he has lost way, way more than the HE

another example:

bet 10 - win - bet 20 - win - bet 40 - win - bet 80 - win - bet 160 - win - bet 320 - lose_____________he has bet a total of $630 and lost $10 - again, way more than the HE

and that doesn't even begin to consider the great many times that he has no chance to parlay (about half the time) because he lost his initial bet

it looks like a player who plays this way will never ever lose an amount that approaches the HE no matter how many hands he plays

unless I'm missing something - which is quite possible

pretty interesting - to me anyway

.

You could start out with a session buy-in of 50 X $100 bets or $5,000 and flat bet Player for 1,000 hands and have a total bet of $100K and an HE of $1,500 (I forgot the HE, is it 1.5%?) and the 1 SD variance could be running +/- $3,100. So the HE will cut your winnings in half or increase your losses by 50% at this level. But you may just cash out at even when you can.

Quote:ChumpChangeThe HE goes up with your total bet amount, but your RTP is 0% because you busted out at all levels without winning.

link to original post

what you stated is true but it doesn't change the fact that -

again, no matter how many hands he plays it looks to me like he will never ever lose an amount that corresponds to what the HE is

I'm not trying to make the bettor's performance improve

I'm only trying to show that playing this way seems to break the rule of the bettor's performance in the long run corresponding to the HE

.

Quote:MDawgI've had this discussion with the Wizard recently.

Let's say someone is playing blackjack with a 0.19 house edge.

100 - 5000 table min/max. Varying bets - no flat betting, but let's not assume any card counting so that higher bets might come anytime. No specific strategy, no regular martingale, just, as they said about Dan Mahowny, "pretty big bets on impulse. No consistent pattern." Player will bet $5000. without any compunction.

Bankroll is $50,000. goal is to win $250,000. and player will not stop a session until he either wins $250,000. or loses $50,000.

Over time, will this person simply lose the house edge, or more?

We don't have to consider necessarily a very long term, let's just say even, a few dozen sessions.

link to original post

Quote:MDawgThe post I made does not concern MY play. It is a simulation based on the player who plays with no edge.

My point in making the post is that over the term of an average let's even say year or two, for the average player, playing that way he will be blown out far more times than will achieve his goal. Actually, many players play exactly like that as they get in the hole and try to win it all back in one session, unwilling to stop even with a partial recovery of their losses.

So this is an example of where the long term results and the shorter results will diverge.

Yes, over millions of sessions the house edge will be what the player should lose, but trying to win $250K every time with a max bet of $5000. and not necessarily betting that every hand, more often than not the player will lose everything.

You could extrapolate this even further, and make the goal whatever the player has lost so far. So after ten blowouts the goal could become $500K or nothing, for each session, and so on. And I see this sort of thing all the time, big players who have lost vast sums over the years or during that trip and as soon as they get on a roll think they'll get back to even in one session, and end up dumping hundreds of thousands of dollars ahead because they think they're going to get to a half million or million that session.

You could try this simulation with no house edge and just a coin toss and still over the shorter term the player would probably lose a lot more than expected via a 50-50 result, especially if the goal for each session increased but the bankroll remained constant or even dropped, and the max bet remained the same.

The risk of ruin is pretty high if the goal is unrealistic given the bankroll and max bet, especially if the player is unwilling to stop unless the goal is achieved or the bankroll is lost.

link to original post

Quote:lilredrooster.

okay - but let's think about a player who plays this way betting only on the Player at Baccarat

he parlays his winning bets until he loses

for example______ bet 10 - win - bet 20 - win - bet 40 lose__________he has bet a total of $70 and lost $10____________he has lost way, way more than the HE

another example:

bet 10 - win - bet 20 - win - bet 40 - win - bet 80 - win - bet 160 - win - bet 320 - lose_____________he has bet a total of $630 and lost $10 - again, way more than the HE

link to original post

(trimmed)

Unless I grossly misunderstand, the player in either of the instances has lost more than $10.

Quote:DieterQuote:lilredrooster.

okay - but let's think about a player who plays this way betting only on the Player at Baccarat

he parlays his winning bets until he loses

for example______ bet 10 - win - bet 20 - win - bet 40 lose__________he has bet a total of $70 and lost $10____________he has lost way, way more than the HE

another example:

bet 10 - win - bet 20 - win - bet 40 - win - bet 80 - win - bet 160 - win - bet 320 - lose_____________he has bet a total of $630 and lost $10 - again, way more than the HE

link to original post

(trimmed)

Unless I grossly misunderstand, the player in either of the instances has lost more than $10.

link to original post

in the first example he lost a bet of $40

in the 2nd example he lost a bet of $320

but he started playing in both instances with $10

I was referring to his net loss - $10 - in both instances

once in a great while he will have a long string on wins and in those instances lose less than the HE

but it wouldn't seem even remotely possible that that could make up for all the times he loses on his 1st, 2nd, 3rd, 4th, 5th or 6th bet and loses much more than the HE

.

It's that thing of referencing it to the original bet ... causes a lot of clucking out there as I found out tooQuote:lilredrooster

in the first example he lost a bet of $40

in the 2nd example he lost a bet of $320

but he started playing in both instances with $10

I was referring to his net loss - $10 - in both instances

once in a great while he will have a long string on wins and in those instances lose less than the HE

but it wouldn't seem even remotely possible that that could make up for all the times he loses on his 1st, 2nd, 3rd, 4th, 5th or 6th bet and loses much more than the HE

.

link to original post

Quote:odiousgambitIt's that thing of referencing it to the original bet ... causes a lot of clucking out there as I found out tooQuote:lilredrooster

in the first example he lost a bet of $40

in the 2nd example he lost a bet of $320

but he started playing in both instances with $10

I was referring to his net loss - $10 - in both instances

once in a great while he will have a long string on wins and in those instances lose less than the HE

but it wouldn't seem even remotely possible that that could make up for all the times he loses on his 1st, 2nd, 3rd, 4th, 5th or 6th bet and loses much more than the HE

.

link to original post

link to original post

I get it both ways.

Most on here look at the profit from each spin to now belong (even if only for a moment) to the player.

Take your same example spread out over a few days. Player wins $10, pockets money, second day wagers $20 and wins, leaves etc.

Sixth day returns, takes cash out loses $320.

Now did he lose $10 or $320?

The law of large numbers is not a mantra or rule of thumb, it is a mathematical theorem. https://en.wikipedia.org/wiki/Law_of_large_numbersQuote:lilredrooster.

the rule of thumb or mantra recited by the gambling experts is that in a great number of trials the player will ALWAYS come very close to losing a % equal to the House Edge

okay - but let's think about a player who plays this way betting only on the Player at Baccarat

he parlays his winning bets until he loses

for example______ bet 10 - win - bet 20 - win - bet 40 lose__________he has bet a total of $70 and lost $10____________he has lost way, way more than the HE

another example:

bet 10 - win - bet 20 - win - bet 40 - win - bet 80 - win - bet 160 - win - bet 320 - lose_____________he has bet a total of $630 and lost $10 - again, way more than the HE

and that doesn't even begin to consider the great many times that he has no chance to parlay (about half the time) because he lost his initial bet

it looks like a player who plays this way will never ever lose an amount that approaches the HE no matter how many hands he plays

unless I'm missing something - which is quite possible

pretty interesting - to me anyway

.

link to original post

The theorem doesn't care if you don't understand or believe it. In your example, the loss percentage will converge on the house edge for banker even if you don't believe it.

Because if the risk of ruin is not only great, but increasing with every blowout, as in the example I illustrated, then it could well be that the more likely outcome is blowout after blowout with a smaller and smaller chance of ever catching up. Why this is hard to accept, is what's unclear.

"Chasing losses" is what leads casino players to blowout after blowout. They walk into each session saying "I'm going to win back EVERY penny I lost, or bust," and more often than not...bust. And then even when they do win back, they stop as soon as they get even. So best case scenario, which happens less often, is break even. More likely scenario, which happens more often, BUST. Add up all those busts to the occasional break even...over all result is huge losses far greater than the HE. (Especially if the player raises his goal by adding to it the sum of each prior loss.)

Quote:MDawgIf the bankroll and max bet remain constant. The player plays the same way, sometimes betting max bet, not always. And the goal for each session increases, with a determined end result of either achieving that goal or losing it all, then most of what you just wrote above is wrong as far as real world results.

You're simulating 100,000 sessions. In five years I haven't had anywhere near 100,000 sessions and I have probably played more than anyone you've ever even heard of in Vegas.

No one is saying that the house edge changes.

Rather, that the real world results over the course of a year or two will almost assuredly be that the player will keep getting blown out.

The bankroll remains $50,000. The max bet 5000. The player does not max bet every hand. The initial goal is $250,000, all or nothing, but after each blowout the goal increases by $50,000. After ten blowouts now the goal is $750,000. with the same $50K bankroll and same max bet. You might think this sort of thing is just a simulation but it is not. I see this sort of thing all the time in high limit, with players who have dumped hundreds of thousands telling me they need to get back to even and won't stop until they are even. Then I see the same player a year later, and now he's saying he needs to win back two million or so, and hasn't even increased his bankroll by much. Keeps trying to win it all back in one session or trip. Keeps losing more and more, blowout after blowout.Quote:SOOPOOThe way you worded the question…. The great majority of time the player will lose his entire $50k. Without knowing how he varies his bets it’s just going to be a guess. My guess is 10% of the time he quintuples his money, 90% he loses it all.

link to original post

Yes I'd agree, at least 90% of the time blowout, more if you assume a rising goal where each prior blowout is added to the all or nothing goal.

link to original post

You tell me if it's accurate:

Risk of Ruin is about the likelihood of you dying in the short run so you never get to the long run. Imagine a coin that actually offered 1:1 odds, but it was skewed so it was 52% Heads / 48% Tails when flipped... you could make money on that game in the long run by betting on Heads.

But what would happen if you had a $1,000 bankroll and you bet $100 per hand? It only takes 10x Net Tails vs. Heads and you're finished. You go bust before the long run ever has a chance of getting there. You're susceptible to actual results deviating from the long run expectation of 52% Heads / 48% Tails.

I found this in the first week of lectures (in a ginormous hall) during first term freshman level Econ101, way far back in the day:Quote:MDawgI found this somewhere online.

...<SNIP>...

"Risk of Ruin is about the likelihood of..."

link to original post

Amen,Quote:Nobel Laureate John Maynard KeynesIn the long run, we are all dead.

https://www.goodreads.com/quotes/6757924-in-the-long-run-we-are-all-dead-economists-set

D. Dead

That is very true but...Quote:MentalThe law of large numbers is not a mantra or rule of thumb, it is a mathematical theorem. https://en.wikipedia.org/wiki/Law_of_large_numbersQuote:lilredrooster.

okay - but let's think about a player who plays this way betting only on the Player at Baccarat

he parlays his winning bets until he loses

for example______ bet 10 - win - bet 20 - win - bet 40 lose__________he has bet a total of $70 and lost $10____________he has lost way, way more than the HE

another example:

bet 10 - win - bet 20 - win - bet 40 - win - bet 80 - win - bet 160 - win - bet 320 - lose_____________he has bet a total of $630 and lost $10 - again, way more than the HE

and that doesn't even begin to consider the great many times that he has no chance to parlay (about half the time) because he lost his initial bet

it looks like a player who plays this way will never ever lose an amount that approaches the HE no matter how many hands he plays

link to original post

The theorem doesn't care if you don't understand or believe it.

Sadly, I disagree. The Dollar amount he will exit with will tend to and DIVERGE towards two possible values: 'profit that is big enough to satisfy' OR a loss of 'As much as he can bear to lose'. No other option. His bankroll is not following a converging series. It will SNAP towards one of two outcomes.Quote:In your example, the loss percentage will converge on the house edge for banker even if you don't believe it.

link to original post

His Expected Loss only starts to approach HE if he proceeds to repeat the exercise with a new initial bet and restart of the series. We know that is likely, but then he's still converging his exit point towards 'profit that is big enough to satisfy' OR a loss of 'As much as he can bear', both in cash terms and percentage terms. He leaves no other option.

Quote:OnceDear

His Expected Loss only starts to approach HE if he proceeds to repeat the exercise with a new initial bet and restart of the series. We know that is likely, but then he's still converging his exit point towards 'profit that is big enough to satisfy' OR a loss of 'As much as he can bear', both in cash terms and percentage terms. He leaves no other option.

link to original post

I am sorry. I chose to address the hypothetical as posed:

Quote:it looks like a player who plays this way will never ever lose an amount that approaches the HE no matter how many hands he plays.

You are free to address any other hypothetical you choose.

Quote:lilredroosterbet 10 - win - bet 20 - win - bet 40 - win - bet 80 - win - bet 160 - win - bet 320 - lose_____________he has bet a total of $630 and lost $10 - again, way more than the HE

House edge is 1.2%. This shows a loss of 1.6%. I would disagree that a difference between expected results and observed results of 0.4% after six bets is "way more"? If anything, it is astonishingly close.

Quote:TomGQuote:lilredroosterbet 10 - win - bet 20 - win - bet 40 - win - bet 80 - win - bet 160 - win - bet 320 - lose_____________he has bet a total of $630 and lost $10 - again, way more than the HE

House edge is 1.2%. This shows a loss of 1.6%. I would disagree that a difference between expected results and observed results of 0.4% after six bets is "way more"? If anything, it is astonishingly close.

link to original post

it's an increase in % of loss of about 33%

in this instance to lose on the 6th bet he had to win 5 in a row

I believe the chance of flipping heads 5 times in a row is 3.125%

the chance of Player winning 5 times in a row is less than that

so many, many times a bet on Player will lose before winning 5 in a row

but anyway, I'm not really sure what I indicated is correct - which is why I used the language "looks like"

if I'm wrong it certainly wouldn't be the first time - I've been wrong lots of times before

it just seemed like a perplexing idea to me at the time - which is why I posted

I wanted to try and test it out on the Wiz's free bacc game with about 1,000 hands but in the history and stats that the game provides it doesn't record the amount bet - I would have to keep track of that separately - just too much work

yes, I know - 1,000 hands would be nowhere near being conclusive - but it would have been interesting to me

.

Quote:lilredroosterit's an increase in % of loss of about 33%

Compared to an increase of 8300% if the streak for player was 0, or 2700% if the streak was 1. It can sometimes take a while for expected results to converge to actual results, but it will happen.

The Wiz's free bacc game is just a computer program simulating the game with a RNG. I could probably write a Monte Carlo program that could simulate 100,000 games per second. So could the Wiz. I can tell you the result without even writing the program. Any betting system that involves varying the wager on otherwise identical bets will produce a cumulative house edge equal to the house edge of a single bet.Quote:lilredroosterQuote:TomGQuote:lilredroosterbet 10 - win - bet 20 - win - bet 40 - win - bet 80 - win - bet 160 - win - bet 320 - lose_____________he has bet a total of $630 and lost $10 - again, way more than the HE

House edge is 1.2%. This shows a loss of 1.6%. I would disagree that a difference between expected results and observed results of 0.4% after six bets is "way more"? If anything, it is astonishingly close.

link to original post

it's an increase in % of loss of about 33%

in this instance to lose on the 6th bet he had to win 5 in a row

I believe the chance of flipping heads 5 times in a row is 3.125%

the chance of Player winning 5 times in a row is less than that

so many, many times a bet on Player will lose before winning 5 in a row

but anyway, I'm not really sure what I indicated is correct - which is why I used the language "looks like"

if I'm wrong it certainly wouldn't be the first time - I've been wrong lots of times before

it just seemed like a perplexing idea to me at the time - which is why I posted

I wanted to try and test it out on the Wiz's free bacc game with about 1,000 hands but in the history and stats that the game provides it doesn't record the amount bet - I would have to keep track of that separately - just too much work

yes, I know - 1,000 hands would be nowhere near being conclusive - but it would have been interesting to me

.

link to original post

You intuitively feel this result is wrong for your scenario. This means your intuition is wrong because it disagrees with your the law of large numbers. You fought the law and the law won.

okay, I accept that the loss in my example will eventually correspond to the HE - that my intuition was wrong

but this is kinna interesting to me

the chance that a coin flip will end up heads if you flip it - times in a row

1 time - 50%

2 times - 25%

3 times - 12.5%

4 times - 6.25%

5 times - 3.125%

6 times - 1.56%

heads will not streak more often than 6 times 98.43% of the time on average

the Player will not streak more often than 6 times more than 98.43% of the time because the chance of Player coming is less than 50/50

if the bettor wins on the 6th time and loses on the 7th that will about equal the HE

all of the other times when Player loses on times 1,2, 3, 4,5 or 6 he will lose an amount greater than the HE

the bettor has to win 7 times or more in a row and then experience a loss to lose an amount greater than the HE

of course, when that does happen he will have bet much more which is why he will eventually lose an amount that corresponds to the HE

and it's easier to see the fallacy when the bettor does the opposite of this - a martingale

if my math is in any way incorrect I would appreciate a correction - thanks

.

Martingale and similar systems don't defeat the Law of Large Numbers. They just delay it.Quote:lilredrooster.

okay, I accept that the loss in my example will eventually correspond to the HE - that my intuition was wrong

but this is kinna interesting to me

the chance that a coin flip will end up heads if you flip it - times in a row

1 time - 50%

2 times - 25%

3 times - 12.5%

4 times - 6.25%

5 times - 3.125%

6 times - 1.56%

heads will not streak more often than 6 times 98.43% of the time on average

the Player will not streak more often than 6 times more than 98.43% of the time because the chance of Player coming is less than 50/50

if the bettor wins on the 6th time and loses on the 7th that will about equal the HE

all of the other times when Player loses on times 1,2, 3, 4,5 or 6 he will lose an amount greater than the HE

the bettor has to win 7 times or more in a row and then experience a loss to lose an amount greater than the HE

of course, when that does happen he will have bet much more which is why he will eventually lose an amount that corresponds to the HE

and it's easier to see the fallacy when the bettor does the opposite of this - a martingale

if my math is in any way incorrect I would appreciate a correction - thanks

.

link to original post

Most people have a poor intuition about low-probability events, and betting systems are often designed to mislead one's intuition by creating a system that is very sensitive to low-probability events.

Imagine that a 10 step Martingale is implemented based on a fair coin flip. Heads pays 1:1 and Tails loses. However, 10 different bettors divide up the betting. Bettor A makes all the single unit bets. Bettor B makes all of the double unit bets after A loses with Tails. Bettor C makes all of the four unit bets after B loses with Tails. And so on until Bettor J makes a 512 unit bet is bettor I lost with tails. After bettor J flips and wins or loses, bettor A always flips for one unit. After any other bettor flips and wins, bettor A always flips for one unit.

Bettor B bets only half as often as bettor A. Each subsequent bettor in the series bets half as often as his predecessor. In a series of 1023 bets, here is how often each bettor will place a wager (on average): 512,256,128,64,32,16,8,4,2,1.

After 10,230 flips, bettor A has already flipped about 5,120 times and is already subject to the law of large numbers. His result will be about 50% heads within a standard deviation of 1.4%. He might win or lose 35 units. Meanwhile, bettor J is definitely not into the large numbers yet. Bettor J will have flipped between 7 and 13 times. If J flips 13 times, then he might win or lose 4*512=2048 units. The variance of the group of bettors is dominated by the success or failure of bettor J (and to a lessor extent bettors I and H).

If you actually trust the math, then the group has a neutral expectation value. They are expected to break even over the long run. But to people who are just using their intuition, they might expect the group will make money based on not believing that tails could come up 10 times in a row. Or they might expect the group to lose money because they see bettor J flipping for a 512 unit bet and they just know that you always lose when the big bet is out.

If you believe that the martingale is not neutral EV, you must believe at least one of the bettors doesn't have neutral EV? Which bettor(s) don't have neutral EV? Explain your thinking.

Quote:MentalIf you believe that the martingale is not neutral EV, you must believe at least one of the bettors doesn't have neutral EV? Which bettor(s) don't have neutral EV? Explain your thinking.

I do believe that the martingale is neutral EV for a coin flip

when I stated that it was fallacious I was referring to if it was falsely believed to be a winning system by using it in betting on the Player at Baccarat

.

Quote:MentalI could probably write a Monte Carlo program that could simulate 100,000 games per second.

link to original post

Why simulate? Just keep pushing dose buttons.

Quote:MentalI played 17,000 hands in a single day when I was young and foolish.

link to original post

Meantime in the real world things may happen as described in this thread and mine. It takes a lot of hands to get to expected results and many undercapitalized players

with lofty goals I’ve observed bust so many times along the way that they aren’t around to experience the long term.Quote:MDawgsusceptible to actual results deviating from the long run expectation

link to original post

I don't see these as serious comments. I don't play systems. But OP wants to understand why a particular system seems to him like it won't converge to the HE quickly. OP wants to simulate bacc using the Wiz's program, but that program doesn't give him results in the form he would like. I also assume the Wiz's program is slow because it is simulating the visual aspects of the baccarat game.Quote:MDawgQuote:MentalI could probably write a Monte Carlo program that could simulate 100,000 games per second.

link to original post

Why simulate? Just keep pushing dose buttons.Quote:MentalI played 17,000 hands in a single day when I was young and foolish.

link to original post

Meantime in the real world things may happen as described in this thread and mine. It takes a lot of hands to get to expected results and many undercapitalized playerswith lofty goals I’ve observed bust so many times along the way that they aren’t around to experience the long term.Quote:MDawgsusceptible to actual results deviating from the long run expectation

link to original post

link to original post

I generally don't push buttons. I just hit autoplay and go do something else more useful and interesting than watching gambling. I am playing on three different site while I am writing this response. I am up $24K for the day even before I have had breakfast, thanks for asking. But even if I choose to make a living pushing buttons, what would be wrong with that? I know you think playing in high limit salons is the highest achievement of mankind. I think winning is more important than where you play.

Of course most of the dreamers that you see bust out. They are playing negative EV games. Sometimes they bust out quickly, but other times they rally and stay in the game a bit longer. Lofty goals don't mean diddly without a coherent plan to gain an edge and a bankroll sufficient to support the plan.

You auto play 17000 hands of video poker too? ☺️

Quote:MDawgJust keep pushing dose buttons.

Oh so you’re an online player on top of machine player? I did not know that. At that point it becomes one’s entire existence so no wonder you think in terms of hundreds of thousands of “hands” played.

until a couple of weeks ago I played around with a system - I got tired of it and stopped playing

it was a reverse Oscar's Grind on the Wiz's bacc game - instead of betting up when you win - betting up when you lose

the base bet was $100 and I would bet up alternating Player or Banker when one or the other won - if one or the other didn't win after I alternated I repeated the bet until either Player or Banker got a win -

The goal was to win about $100 and then I would start over trying to win another $100

so, for example it was bet $100 - lose - bet $200 - lose - bet $300 - lose - bet $400 win - now down $200 - so bet $300 and win - since a winning bet on the Banker took 5% commission I would sometimes have to make more bets than that to win about $100 - if the win was too much less than $100 (subjective judgement) I would bet $100 again

the goal of the entire session was to win about $1,000 (such as $970 or $1030) before busting out - the Bank the game gives you is $10,000

when I won about $1,000 the session was over - I played about 5 sessions per day for several weeks - a session took only about 3 to 10 minutes so I didn't spend a great deal of time on it per day

I only busted out of my Bank of $10,000 one time - I didn't actually keep track but I know I won about $1,000 at least 100 times - for a net win of at least $90,000

YES, I KNOW THAT THIS IS NOT A WINNING SYSTEM AND I'M NOT CLAIMING THAT IT IS

obviously, the system will eventually lose in the long run

but I was amazed at how long I was able to play it and be profitable by that large of an amount

.

Quote:lilredrooster.

until a couple of weeks ago I played around with a system - I got tired of it and stopped playing

...

The goal was to win about $100 and then I would start over trying to win another $100

...

the goal of the entire session was to win about $1,000 (such as $970 or $1030) before busting out - the Bank the game gives you is $10,000

when I won about $1,000 the session was over - I played about 5 sessions per day for several weeks - a session took only about 3 to 10 minutes so I didn't spend a great deal of time on it per day

I only busted out of my Bank of $10,000 one time - I didn't actually keep track but I know I won about $1,000 at least 100 times - for a net win of at least $90,000

...

but I was amazed at how long I was able to play it and be profitable by that large of an amount

.

link to original post

Interesting result.

Let's ignore house edge and do some estimating.

'won about $1000 at least 100 times' means 100 sessions? at 5 sessions a day about 20 days. Is that right?

Using OnceDear's rule of thumb, for each session, with starting bankroll of 10,000 and session target bankroll of 11,000, each session had probability of success of about 10,000/11,000 = 90.9%

You should have expected to bust out about 10 times in those 100 sessions. But you only bust out once?!? That's pretty cool.

rough numbers...

You took 10,000 at risk and won 99,000 and at some point lost 10,000 giving you a net profit of 99,000?

Which is the same as turning 10,000 into 99,000

Which is analogous to winning a single wager with probability of 10.1%

I.e. a 9:1 shot paid off for you.

Not fantastically 'amazing', but still remarkable.

Shall we try it again with carefully logged results?

Quote:OnceDear

I.e. a 9:1 shot paid off for you.

Not fantastically 'amazing', but still remarkable.

Shall we try it again with carefully logged results?

I'm really bored with it now - I don't wanna do it again_________maybe you or somebody else

here is something else interesting to me that I can't explain

as I mentioned - I was playing a reverse Oscar's Grind - bet up only after a loss

but I started with the actual Oscar's Grind - which is only bet up after a win

the actual Oscar's grind was getting creamed - losing a lot - that's why I switched to the reverse

and immediately when I switched to the reverse Oscar - betting up after a loss - the results completely flipped

just the Gods of Chance laughing at me_____?________making me think I actually found out something when I really didn't_________?

I guess

.

This player walked up after I'd already won about six Players, and probably figured the next one would be a Bank. Put down $1000. on the Bank - the ones I've seen play like that in high limit with their Martingales don't mess around they aren't trying to win just a hundred dollars.

I got six more Players and the guy had $35,000 (the table limit at that time) on Bank, this time he got a natural 9 and won. I could just feel the relief.

Saw the same player pulling the same thing again another time, and had to go 6 deep again before winning. But he did win. I suppose he figured that once I had 6 Players it was more than time for the Bank based on his "must win sixth hand" strategy, but

1000

2000

4000

8000

16000

35000

I don't Martingale. After losing a hand I am just as likely to lower my bet.

Why 35K at the end? Well, why not I mean what's the difference between losing 32 and losing 35 I suppose, at that point. But if the player had lost on that sixth hand, almost seventy grand! gone. And now the need to win almost seventy 1000 winning hands/sequences just to get back to even.

Quote:lilredroosterQuote:OnceDear

I.e. a 9:1 shot paid off for you.

Not fantastically 'amazing', but still remarkable.

Shall we try it again with carefully logged results?

I'm really bored with it now - I don't wanna do it again_________maybe you or somebody else

link to original post

I'm doing it. Just for 5h175 and giggles

Advantage Players often use Kelly to help avoid going bust. They might even use half Kelly. If you were to use that on a 2% Advantage while in a coin-flip bet situation, it isn't going to take that long to see a profit.Quote:MDawgI found this somewhere online.

You tell me if it's accurate:

Risk of Ruin is about the likelihood of you dying in the short run so you never get to the long run. Imagine a coin that actually offered 1:1 odds, but it was skewed so it was 52% Heads / 48% Tails when flipped... you could make money on that game in the long run by betting on Heads.

But what would happen if you had a $1,000 bankroll and you bet $100 per hand? It only takes 10x Net Tails vs. Heads and you're finished. You go bust before the long run ever has a chance of getting there. You're susceptible to actual results deviating from the long run expectation of 52% Heads / 48% Tails.

link to original post

Oftentimes, when people are talking about the law of large numbers and the long run they are addressing what it takes to achieve the exact calculated percentage of that game. They tend to forget that you don't need to be in the extreme long run to achieve 1%- 99% of the expectation. Someone with a 2.7% advantage might be running badly and only achieve 1% of their expectation. Someone with -2.7% expectation might only be losing 1.7 %.

No, not really. You are already part of the long run, therefore you are in the long run.Quote:MDawgMore or less what you are saying is exactly the point - that the long run isn't always achieved if you aren't around to see it.

link to original post

My theory is... You can't trick math and EVERYTHING averages out to what it should be.

So I could buy-in with $2,750 at a $50 table and try this.

Lilredrooster mentioned quitting when he lost $10K on his sessions. That would be $100 + $200 + $300 + $400 + $500 + $600 + $700 + $800 + $900 + $1,000 + $1,100 + $1,200 + $1,300 + $1,400 = $10,500. So 14 levels of loss. He quit by the time he got above $11K, so there's no 15th level of loss.

A 15th level of loss would make the total $12,000. Maybe just add an extra level of loss as I go up instead of raising my unit bet? Ahh, it will be easier to calculate 15X unit max bets.

I continued my game with $60 units and I ran up to 15X my bet right off the bat ($900) with the 15 losses adding up but "last bet if I lose" I didn't lose and eventually won back and reduced my bet to $60 and totaled a win of near $600. I had another run up to $720 and got it back to $60 and now I'm up $2,300. The whole 15X $60 unit progression is a $7,200 if I lose. I'm starting to wonder if the 5% commission is getting to be too much for this kind of play and should I play Player instead since Banker only pays one extra hand per shoe; but Banker is winning this shoe 9 to 2 so far. Yeah, well, I busted out before the shoe ended with a $900 bet and now I'm down to a $9K balance, so I would have been up to over $16K if I made a come-back, so I'm at -$1K now.

Maybe I should switch sides if one side is up by 10 or more hands in the shoe.

I started a new game at $10K and bet Player instead so there's no commission. 14 hands in Player has won 10 hands, Banker has won 4 hands, and I'm up 10X $60. Win 40 hands per shoe as Player with $60 bet units pays approx. $2.4K in a shoe. So need 3 shoes of wins to recoup loss. I busted out with a $2,740 loss and I'm at $7,260 so I'll try again. I over bet by $400 on one bet (made a $600 a bet a $1,000 bet and lost), so my last bet was $840 and I had $620 left, and I busted out again. That was a one way train to ruin. I like the strategy, it's just the luck I'm having is just cursed. Maybe it'd work on the Don't Pass Line.

I played the Don't Pass Line on the Wiz's site for awhile and right off I was hit by right siders and my bet jumped to the 12th level of $720 on $60 units. It's a long climb down and by the time I got to a $60 bet again I was at +$4,360. I may have misbet slightly a couple times so the profit doesn't divide by $60 evenly, but that's 72 units ahead. Maybe if I'm coming off that high a disadvantage, I could forfeit when I get down to the 5th level bet and take a 15 unit loss and start over betting the minimum.

Ahh, I busted out at $8,600, or -$1,400. I'm not getting the 120+ hits that are needed to keep this thing going. Didn't quite make 100 hits this time. But, I'm ready if I can get 500+ hits without a loss. But really, I keep busting out with way under 120 hits repeatedly.

The expectation value can be calculated based on the rules of a game. You don't even have to play a single game. In fact, you can calculate the expectation value of a game, a wager, a prop, or an investment that has never happened. The LLN says that this EV number is special. In the long run, the average of the results will always converge on this one special number. That is what makes EV such a useful tool for decision making in the presence of risk and uncertainty caused by randomness. The EV does not predict the outcome of the next event.Quote:MDawgMore or less what you are saying is exactly the point - that the long run isn't always achieved if you aren't around to see it.

link to original post

Risk of ruin is another useful tool for some APs. But the RoR is not a precise number in the same way as the EV is a precise mathematical concept. To calculate the RoR, I have to know the bankroll, whether the bankroll can or will be replenished, whether the bet sized can be varied, whether this variation is continuous or lumpy, whether the gambler is going to stick to the plan and not tilt or chicken out, whether the casino will limit the player or bar them, whether the gambler will have a heart attack before ruin, etc. Even then, I cannot calculate the RoR in closed mathematical form like I can usually do for EV.

The risk of tapping out does not effect the EV of the game in any way. Consider your biased coin example. For every gambler that taps out, there are multiple gamblers who make unlimited amounts of money. If you average over all gamblers, the EV is still the same.

The point I am trying to make as strongly as possible is that having a finite bankroll does not effect the EV. You are saying that the RoR is high in a coinflip game if you bet 10% of the bankroll on each flip. I agree. But if you cap losses at 10 units and cap wins at 10 units, more gamblers will hit the win cap than will hit the loss cap. If you calculate the net amount won by the group and divide by the total action, the resulting average will converge to the EV in the long run. This is the LLN. The losers don't get to the long run, but neither do the winners. Yet EV is a very important number in deciding whether one should play your biased coin flip game. RoR is another consideration.

I have not even considered a wager in the last 20 years that posed any risk of ruin to me. I know these gambling opportunities exist, for example, in the lottery, pick-6, or ultra-high limit table games. My gambling revolves around games where the bet is usually capped at $200 or less and the variance is usually well less than 100 and I have a decent edge. I don't even need to consider ruin.

https://www.scientificamerican.com/article/the-gambling-strategy-thats-guaranteed-to-make-money-and-why-you-should-never-use-it/

Quote:A disconcerting 28 percent of participants went broke despite having an advantage, and a shocking two thirds bet on tails at some point in the game, which is never rational. On average, the participants walked away with $91 (winnings were capped at $250). This might seem like an ample take for someone starting with $25, but the researchers calculated that over the 300 coin tosses time allowed for, the average winnings of players using the optimal strategy (described below) would be more than $3 million!

I think some systems can be fun for a player who doesn't desire to be a pro and can accept the fact that he will eventually lose in the long run

I would guess that quite a few could enjoy playing a system that has a lot of smallish wins and only a few large losses

in baccarat after betting $50,000. the bettor's expected loss is only about $575

that's a very small number to me and probably to many others here

of course the pros on the site will say it only makes sense to gamble when you have an advantage - I used to feel that way too when was I counting cards in the 90s

I say each to his own - if a non pro can enjoy playing this way - hey - whatever

.

Makes sense that all Mental does is push buttons and give us theory because the reality is at the tables, not in some number churning that doesn't take into account risk of ruin. Even the minority who win a lot end up giving back more.

As well, I have an advantage that is calculable and yet I have won much more than expected, but that's not what I'm talking about here. I'm just talking about the hard math reality of that these undercapitalized players with lofty goals keep losing it all time and again. Among that subset - which it's a pretty big subset - they're not making it to the long term of just a HE loss.

Just the fact that he has to write volumes to try to Explain His Position is evidence of that he needs to get back to pushing dose buttons. 😆

Quote:MDawgMost of these players I see at the big tables are getting blown out trip after trip losing much more than expected, for reasons I have articulated above and elsewhere. There are so many players who don't stop until they have lost it all and have no other outcomes.

Makes sense that all Mental does is push buttons and give us theory because the reality is at the tables, not in some number churning that doesn't take into account risk of ruin. Even the minority who win a lot end up giving back more.

As well, I have an advantage that is calculable and yet I have won much more than expected, but that's not what I'm talking about here. I'm just talking about the hard math reality of that these undercapitalized players with lofty goals keep losing it all time and again. Among that subset - which it's a pretty big subset - they're not making it to the long term of just a HE loss.

Just the fact that he has to write volumes to try to Explain His Position is evidence of that he needs to get back to pushing dose buttons. 😆

link to original post

MDawg, I can and do disagree with some aspects of Mental's posts, but can't we do so without implicitly insulting him?

"he needs to get back to pushing dose buttons." looks like an insult to me.

I understand that he plays significantly high stakes online and that he knows how to exploit certain advantage plays. "pushing dose buttons" doesn't quite cover it.

The point on which I disagree with him is his assertion that a player's 'hold' will tend to converge on the game's house edge. That is undeniable for small flat betting, but where bets are stupidly large or growing by progressives, the hold will be 100% or negative, as experienced by the sort of over betters that you describe.

I understand what he is saying. For example, say my per session bankroll was $100K and I bet nothing but $100K. Assume even, a 50-50 game. Well, if I walked away after just one bet then half the time I would win $100K half the time I would lose $100K. But what if I was determined to either win $5,000,000. or lose everything, every session. The bet remains $100K each time. What are the odds I will win that $5,000,000.? Sure, over an infinite number of attempts I would win that $5,000,000. an infinite number of times such that the net win would even out with the infinite number of $100K losses.

But in reality, I'd probably run out of money and give up before I ever won $5,000,000. even one time, unless I was extremely lucky.

Now, let's say that I added whatever I had lost in all prior bets to the goal. Every $100K lost is added to the "all or bust" $5,000,000. goal, sending the goal higher and higher. Now the chances of ever stopping with other than a total loss increase even over time.

While this may seem theoretical, I see this basic determined to win it all or lose everything theme all the time in Vegas. Even just the simple greed of trying to empty the rack every time a big player is on a run, contributes to this kind of thing.

Casinos clean out high rollers not from the little HE, but from this all or nothing chasing sort of mentality. Even if Baccarat commission were lowered to where the Bank became an even fifty fifty proposition, you'd still see high rollers betting on Player sometimes.

If the equalityQuote:ChumpChangeThe HE goes up with your total bet amount, but your RTP is 0% because you busted out at all levels without winning.

[...]

link to original post

HE = 1 - RTP

was correct and RTP in this case is 0%, shouldn't then

HE = 1 - 0% = 1 = 100%

and moreover be constant and not varying with the total amount bet?

Quote:ThomasKIf the equalityQuote:ChumpChangeThe HE goes up with your total bet amount, but your RTP is 0% because you busted out at all levels without winning.

[...]

link to original post

HE = 1 - RTP

was correct and RTP in this case is 0%, shouldn't then

HE = 1 - 0% = 1 = 100%

and moreover be constant and not varying with the total amount bet?

link to original post

I believe that would be inaccurate.

HE and RTP in that sense are related to EV; your bust-out is AV. If the number of trials is insufficient, AV and EV may not converge.

Could you please explain the term "AV"?Quote:Dieter[...]

I believe that would be inaccurate.

HE and RTP in that sense are related to EV; your bust-out is AV. If the number of trials is insufficient, AV and EV may not converge.

link to original post

Sounds very much like parlaying to me. I'd like to quote the Wizard:Quote:darkozQuote:odiousgambitIt's that thing of referencing it to the original bet ... causes a lot of clucking out there as I found out tooQuote:lilredrooster

in the first example he lost a bet of $40

in the 2nd example he lost a bet of $320

but he started playing in both instances with $10

I was referring to his net loss - $10 - in both instances

once in a great while he will have a long string on wins and in those instances lose less than the HE

but it wouldn't seem even remotely possible that that could make up for all the times he loses on his 1st, 2nd, 3rd, 4th, 5th or 6th bet and loses much more than the HE

.

link to original post

link to original post

I get it both ways.

Most on here look at the profit from each spin to now belong (even if only for a moment) to the player.

Take your same example spread out over a few days. Player wins $10, pockets money, second day wagers $20 and wins, leaves etc.

Sixth day returns, takes cash out loses $320.

Now did he lose $10 or $320?

link to original post

Quote:Wizard[...]

It depends how you define the house edge involving parlays. I generally only count the original wager as money bet. The reason the house edge is so high is the same money is often bet multiple times.

link to original post

Quote:ThomasKCould you please explain the term "AV"?Quote:Dieter[...]

I believe that would be inaccurate.

HE and RTP in that sense are related to EV; your bust-out is AV. If the number of trials is insufficient, AV and EV may not converge.

link to original post

link to original post

Actual Value - what was actually won or lost.

EV is Expected Value, the theoretical win or loss that actual results should approach.

I definitely like your example and I think that your question is absolutely legitimate. In my opinion it deserves some more detailed analysis. Here is my take on it.Quote:lilredrooster.

the rule of thumb or mantra recited by the gambling experts is that in a great number of trials the player will ALWAYS come very close to losing a % equal to the House Edge

okay - but let's think about a player who plays this way betting only on the Player at Baccarat

he parlays his winning bets until he loses

for example______ bet 10 - win - bet 20 - win - bet 40 lose__________he has bet a total of $70 and lost $10____________he has lost way, way more than the HE

another example:

bet 10 - win - bet 20 - win - bet 40 - win - bet 80 - win - bet 160 - win - bet 320 - lose_____________he has bet a total of $630 and lost $10 - again, way more than the HE

and that doesn't even begin to consider the great many times that he has no chance to parlay (about half the time) because he lost his initial bet

it looks like a player who plays this way will never ever lose an amount that approaches the HE no matter how many hands he plays

unless I'm missing something - which is quite possible

pretty interesting - to me anyway

.

link to original post

The house edge of a game is a factor within the expected value of that game. I'd like to show how the house edge can be made visible inside the expected value. I'll use the PLAYER bet in Baccarat as an example for this method. Then I will apply that method to your scenario.

In order to calculate the expected value of a game only two ingredients an two constraints are needed.

The ingredients are

1) the outcomes of the game and

2) the corresponding probabilities of these outcomes.

As constraints,

1) the sum of all probabilities have to add up to 1 (i.e. 100%) to make sure that all possible outcomes of the game are taken into account and

2) the game has to be repeated infinitely often in order to approach the given probablities. In real life the term "in the long run" approximates this constraint quite well already after some hundred thousands of repetitions.

The ingredients for the PLAYER bet in Baccarat based on a wager of $10.

1) Outcomes as losses and gains

1a) Player loses: -$10

1b) Player ties: $0

1c) Player wins: $10

2) Probabilities (taken from the Wizard's Baccarat calculator for 8 decks)

2a) Player loses: 0.458597

2b) Player ties: 0.095156

2c) Player wins: 0.446247

Constraints.

The sum of all probabilities equals 1 and therefore fullfills the constraint.

In order to come close to these given probabilities the player needs to play many hundred thousands of PLAYER bets in Baccarat (wagering $10 each hand).

Expected value.

First we define the random variable for which the expected value will be determined.

X_{1}= Player bets $10 on PLAYER in Baccarat.

The common form of the expected value is the term containing losses and gains.

E(X_{1}) = 0.458597 * (-$10) + 0.095156 * ($0) + 0.446247 * ($10)

An equvalent expression can be given based on the payouts and the wager.

a) Player loses: payout is $0, wager is -$10, i.e. player doesn't receive any money back

b) Player ties: payout is $10, wager is -$10, i.e. player receives wager back

c) Player wins: payout is $20, wager is -$10, i.e. player receives wager plus the same amount back

= 0.458597 * ($0 - $10) + 0.095156 * ($10 - $10) + 0.446247 * ($20 - $10)

In a first intermediate step we bring together the wager parts in the expression and thus separate the payouts.

= (0.458597 + 0.095156 + 0.446247) * (-$10) + 0.458597 * $0 + 0.095156 * $10 + 0.446247 * $20

In a second intermediate step we factor out the wager completely.

/ $10 $20 \

= (-$10) * | (0.458597+0.095156+0.446247) * 1 - 0.095156 * --- - 0.446247 * --- |

\ $10 $10 /

We know that the sum of all probabilities equals 1 and multiplied by 1 remains 1.

Also, factoring out the minus of the payout components, reveals the RTP (return to player).

= (-$10) * (1 - (0.095156 * 1 + 0.446247 * 2))

= (-$10) * (1 - (0.095156 + 0.892494))

= (-$10) * (1 - 0.987650)

In a last step the amount is factored out and the RTP subtracted from 1 to produce an expression of the form

"amount wagered" times "per dollar bet" times "house edge".

= 10 * (-$1) * 0.01235

So the the game represented by the random variable

X

_{1}= Player bets $10 on PLAYER in Baccarat

has a house edge of 1.24%.

In other words:

The house edge defines the percentage the player loses on average from every dollar bet in the long run.

In the case of the PLAYER bet in Baccarat the player on average loses 1.24 cents from every dollar bet in the long run.