Losing x hands in a row, down a lot, question the legitness of the game
June 10th, 2011 at 5:11:31 PM
permalink
I know, ploppy all the way just from reading the thread title. I'm not, and play basic strategy and have even counted cards as little back when I played in real casinos, but I just can't help but think. I don't recall any ridiculously bad luck in real casinos (keyword: recall, I know), and I wasn't even playing a +EV blackjack game there. But I am currently playing a +EV blackjack game on 5Dimes, and only played a few times, but I just had an absolute horrible session:
Betting $1 flat, with no double downs or resplitting allowed, I was down $25 after under 100 hands. Then later in the session, I was down $15 or so, and lost 13 hands in a row, which set me close to $30 down. The chance of the 13 in a row are .0055% (or 1 in 18,800), based on losing 47% of all hands.
Basically, when do you stop saying a 1 in x chance is still a chance, and somebody has to hit it as some point, it's just how it goes, just got unlucky, etc., and start thinking OK this is "not possible". When would you think that maybe there are streaks in the RNGs that aren't supposed to be there?
BTW, it's funny I'm posting this, because I just found out all about Rob Singer and his nonsense. He sounds like a typical ploppy who can "feel the flow" or some crap. But now I'm sounding like him.
Betting $1 flat, with no double downs or resplitting allowed, I was down $25 after under 100 hands. Then later in the session, I was down $15 or so, and lost 13 hands in a row, which set me close to $30 down. The chance of the 13 in a row are .0055% (or 1 in 18,800), based on losing 47% of all hands.
Basically, when do you stop saying a 1 in x chance is still a chance, and somebody has to hit it as some point, it's just how it goes, just got unlucky, etc., and start thinking OK this is "not possible". When would you think that maybe there are streaks in the RNGs that aren't supposed to be there?
BTW, it's funny I'm posting this, because I just found out all about Rob Singer and his nonsense. He sounds like a typical ploppy who can "feel the flow" or some crap. But now I'm sounding like him.
June 10th, 2011 at 5:43:49 PM
permalink
(what's a ploppy?)
Someday, joor goin' to see the name of Googie Gomez in lights and joor goin' to say to joorself, "Was that her?" and then joor goin' to answer to joorself, "That was her!" But you know somethin' mister? I was always her yuss nobody knows it! - Googie Gomez
June 10th, 2011 at 5:46:19 PM
permalink
It's hard to believe in probability sometimes when it seems like you just can't win.. I was at a blackjack table where someone bought in for $200 and lost it all without winning a hand while flatbetting $10. So 20 straight losses and he was playing basic strategy..
June 10th, 2011 at 5:52:11 PM
permalink
iwannaiguana, If he lost 20 in a row, then that couldn't have bode well for you. I bet you were happy to see him and his black cloud of doom leave the table.
Someday, joor goin' to see the name of Googie Gomez in lights and joor goin' to say to joorself, "Was that her?" and then joor goin' to answer to joorself, "That was her!" But you know somethin' mister? I was always her yuss nobody knows it! - Googie Gomez
June 10th, 2011 at 5:53:41 PM
permalink
Quote: s2dbaker(what's a ploppy?)
A ploppy is that fat assed woman from WI, chain smoking, who, while you're in the middle of a winning streak with just you and the dealer, plops down next to you and says "Is this where I can play blackjack?" A ploppy is somebody who isn't serious about the game.
"It's not called gambling if the math is on your side."
June 10th, 2011 at 6:30:31 PM
permalink
Quote: EvenBobA ploppy is that fat assed woman from WI, chain smoking, who, while you're in the middle of a winning streak with just you and the dealer, plops down next to you and says "Is this where I can play blackjack?" A ploppy is somebody who isn't serious about the game.
Careful how you say that, because it kinda sounds like you're saying that that women will no doubt screw up your streak, which brings me to:
A ploppy is also somebody who thinks things like making a certain play, whether it's actually correct or not, will affect the entire table negatively and screw up the flow of the cards. Same with jumping in and out of play (Wonging was hard due to ploppies yelling at me). This is the #1 ploppy (see: dumbass) mistake in blackjack, is thinking that what other people do affect the chances of winning. Oh but that blackjack the ploppy gets the next hand, oh well that was just luck; the flow of the cards is still messed up. Riight.
June 10th, 2011 at 6:38:43 PM
permalink
My math was wrong. It's .0055%, which is 1 in 18,800! I was calculating the chance of winning or tying 13 hands in a row (.53), not losing (.47). Wow, so it's much worse than I thought. I went back and edited my previous posts.
June 10th, 2011 at 6:49:00 PM
permalink
Here is a table that shows you some parameters. Out of 9 plays for different house advantages what are the probabilities that you will lose at least k in a row at least one time where k is from 2 to 9. Different house advantages were shown.
The probability of losing 9 in a row out of 9 plays with no house advantage is 1/2^9=0.195% is the easiest one to calculate.
The same table can also be viewed as the primary relationship of the probability of at least one streak of at least k in length out of 9 plays, where there it is slightly more likely depending on the house advantage. You must add the following amounts to the top number.
The probability of losing 9 in a row out of 9 plays with no house advantage is 1/2^9=0.195% is the easiest one to calculate.
| House Advantage | 2 | 3 | 4 | 5 | 6 | 7 | 8 | 9 |
|---|---|---|---|---|---|---|---|---|
| none | 82.617% | 46.484% | 21.680% | 9.375% | 3.906% | 1.563% | 0.586% | 0.195% |
| 1% | 83.256% | 47.446% | 22.391% | 9.788% | 4.122% | 1.667% | 0.632% | 0.214% |
| Line bet in craps | 83.517% | 47.845% | 22.690% | 9.962% | 4.214% | 1.712% | 0.653% | 0.222% |
| 0 roulette | 84.311% | 49.089% | 23.632% | 10.519% | 4.510% | 1.858% | 0.719% | 0.248% |
| 00 roulette | 85.817% | 51.564% | 25.568% | 11.691% | 5.146% | 2.179% | 0.868% | 0.310% |
| 8% (i.e. bad slots) | 87.321% | 54.211% | 27.727% | 13.040% | 5.901% | 2.571% | 1.056% | 0.390% |
The same table can also be viewed as the primary relationship of the probability of at least one streak of at least k in length out of 9 plays, where there it is slightly more likely depending on the house advantage. You must add the following amounts to the top number.
| House Adv | 2 | 3 | 4 | 5 | 6 | 7 | 8 | 9 |
|---|---|---|---|---|---|---|---|---|
| none | 82.617% | 46.484% | 21.680% | 9.375% | 3.906% | 1.563% | 0.586% | 0.195% |
| 1% | 0.639% | 0.962% | 0.712% | 0.413% | 0.215% | 0.104% | 0.046% | 0.018% |
| Pl craps | 0.900% | 1.361% | 1.010% | 0.587% | 0.307% | 0.149% | 0.067% | 0.026% |
| 0 roulette | 1.694% | 2.604% | 1.953% | 1.144% | 0.603% | 0.295% | 0.133% | 0.053% |
| 00 roulette | 3.199% | 5.080% | 3.889% | 2.316% | 1.240% | 0.616% | 0.282% | 0.115% |
| 8% | 4.704% | 7.726% | 6.048% | 3.665% | 1.995% | 1.008% | 0.470% | 0.195% |
June 10th, 2011 at 6:53:16 PM
permalink
Quote: iwannaiguanaIt's hard to believe in probability sometimes when it seems like you just can't win.. I was at a blackjack table where someone bought in for $200 and lost it all without winning a hand while flatbetting $10. So 20 straight losses and he was playing basic strategy..
Chance of this is .0000276666%, or 1 in 36,150.
Somebody check my math, because how can 13 losses in a row be 1 in 18,800, but 20 in a row is only 1 in 36,150? I did .47^x, where x is the number of losses in a row. Then take that number, and move the decimal 2 to the right to get the %. Then to convert % to a number, you just do like if it's .5800, well that's out of 10,000, and you'd divide the latter by the former to reduce it down to 1 in whatever. All that's right, right?
June 10th, 2011 at 7:47:21 PM
permalink
"Basically, when do you stop saying a 1 in x chance is still a chance, and somebody has to hit it as some point, it's just how it goes, just got unlucky, etc., and start thinking OK this is "not possible". When would you think that maybe there are streaks in the RNGs that aren't supposed to be there?"
Well, fwiw, how I approach that is, especially when on the internet, is keep track of my results and, no matter how much I am up or down while flat-betting or not, is pretend later all my bets were for $1 and then calculate how many standard deviations I am up or down from expected value. It's so easy to record results exactly when playing on the internet, why not?
You say "hands" where I would say "rounds" or number of dealer upcards you played against.
Like splitting to 4 hands and doubling each for 8 units and losing -8 units is on one "round" not 4 "hands".
And do it for each session and also cumulatively for the same game over time.
And, then , if your analysis indicates results exceeding 4 Standard Deviations or so, I guess, I'd at least begin to fret a little.
It is amazing how painful being down 2 to 3 standard deviations feels despite being in the category of "*hit happens."
Like being down 25 units or more after 100 rounds might be a little over 2 standard deviations and nothing to raise concerns.
Well, fwiw, how I approach that is, especially when on the internet, is keep track of my results and, no matter how much I am up or down while flat-betting or not, is pretend later all my bets were for $1 and then calculate how many standard deviations I am up or down from expected value. It's so easy to record results exactly when playing on the internet, why not?
You say "hands" where I would say "rounds" or number of dealer upcards you played against.
Like splitting to 4 hands and doubling each for 8 units and losing -8 units is on one "round" not 4 "hands".
And do it for each session and also cumulatively for the same game over time.
And, then , if your analysis indicates results exceeding 4 Standard Deviations or so, I guess, I'd at least begin to fret a little.
It is amazing how painful being down 2 to 3 standard deviations feels despite being in the category of "*hit happens."
Like being down 25 units or more after 100 rounds might be a little over 2 standard deviations and nothing to raise concerns.
June 10th, 2011 at 8:10:35 PM
permalink
Quote: pacomartinHere is a table that shows you some parameters. Out of 9 plays for different house advantages what are the probabilities that you will lose at least k in a row at least one time where k is from 2 to 9. Different house advantages were shown.
The probability of losing 9 in a row out of 9 plays with no house advantage is 1/2^9=0.195% is the easiest one to calculate.
House Advantage 2 3 4 5 6 7 8 9 none 82.617% 46.484% 21.680% 9.375% 3.906% 1.563% 0.586% 0.195% 1% 83.256% 47.446% 22.391% 9.788% 4.122% 1.667% 0.632% 0.214% Line bet in craps 83.517% 47.845% 22.690% 9.962% 4.214% 1.712% 0.653% 0.222% 0 roulette 84.311% 49.089% 23.632% 10.519% 4.510% 1.858% 0.719% 0.248% 00 roulette 85.817% 51.564% 25.568% 11.691% 5.146% 2.179% 0.868% 0.310% 8% (i.e. bad slots) 87.321% 54.211% 27.727% 13.040% 5.901% 2.571% 1.056% 0.390%
The same table can also be viewed as the primary relationship of the probability of at least one streak of at least k in length out of 9 plays, where there it is slightly more likely depending on the house advantage. You must add the following amounts to the top number.
House Adv 2 3 4 5 6 7 8 9 none 82.617% 46.484% 21.680% 9.375% 3.906% 1.563% 0.586% 0.195% 1% 0.639% 0.962% 0.712% 0.413% 0.215% 0.104% 0.046% 0.018% Pl craps 0.900% 1.361% 1.010% 0.587% 0.307% 0.149% 0.067% 0.026% 0 roulette 1.694% 2.604% 1.953% 1.144% 0.603% 0.295% 0.133% 0.053% 00 roulette 3.199% 5.080% 3.889% 2.316% 1.240% 0.616% 0.282% 0.115% 8% 4.704% 7.726% 6.048% 3.665% 1.995% 1.008% 0.470% 0.195%
For blackjack, you want to take some figure between none and 1%, pretty much dead in between (.5% house edge), but I doubt you can just take the average of the 2.
June 10th, 2011 at 10:11:41 PM
permalink
Quote: SilentBob420BMFJFor blackjack, you want to take some figure between none and 1%, pretty much dead in between (.5% house edge), but I doubt you can just take the average of the 2.
I put in that case exactly. I also added a fourth decimal point for streaks of 8 or 9.
Someone mentioned losing 20 in a row right as they sit down. Since you are talking about a million to one, that is pretty unlikely.
| House Adv | 2 | 3 | 4 | 5 | 6 | 7 | 8 | 9 |
|---|---|---|---|---|---|---|---|---|
| none | 82.617% | 46.484% | 21.680% | 9.375% | 3.906% | 1.563% | 0.5859% | 0.1953% |
| 0.50% | 82.939% | 46.965% | 22.034% | 9.580% | 4.013% | 1.614% | 0.6088% | 0.2043% |
| 1.00% | 83.256% | 47.446% | 22.391% | 9.788% | 4.122% | 1.667% | 0.6324% | 0.2136% |
| Pass Line craps | 83.517% | 47.845% | 22.690% | 9.962% | 4.214% | 1.712% | 0.6525% | 0.2216% |
| 0 roulette | 84.311% | 49.089% | 23.632% | 10.519% | 4.510% | 1.858% | 0.7187% | 0.2483% |
| 00 roulette | 85.817% | 51.564% | 25.568% | 11.691% | 5.146% | 2.179% | 0.8677% | 0.3099% |
| 8.00% | 87.321% | 54.211% | 27.727% | 13.040% | 5.901% | 2.571% | 1.0556% | 0.3904% |
| House Adv | 2 | 3 | 4 | 5 | 6 | 7 | 8 | 9 |
|---|---|---|---|---|---|---|---|---|
| none | 82.617% | 46.484% | 21.680% | 9.375% | 3.906% | 1.563% | 0.5859% | 0.1953% |
| 0.50% | 0.321% | 0.481% | 0.354% | 0.205% | 0.107% | 0.051% | 0.0228% | 0.0090% |
| 1.00% | 0.639% | 0.962% | 0.712% | 0.413% | 0.215% | 0.104% | 0.0464% | 0.0183% |
| Pass Line craps | 0.900% | 1.361% | 1.010% | 0.587% | 0.307% | 0.149% | 0.0666% | 0.0263% |
| 0 roulette | 1.694% | 2.604% | 1.953% | 1.144% | 0.603% | 0.295% | 0.1328% | 0.0530% |
| 00 roulette | 3.199% | 5.080% | 3.889% | 2.316% | 1.240% | 0.616% | 0.2818% | 0.1146% |
| 8.00% | 4.704% | 7.726% | 6.048% | 3.665% | 1.995% | 1.008% | 0.4697% | 0.1951% |
At the other extreme, out of 1000 plays the odds of a losing streak of 8, or 9 in a row (given 0.5% house edge are) 87.07% for 8, and 63.91% for 9.
So if you have a roomful of blackjack tables and are playing 1000 hands per hour for the entire room, and everyone is playing basic strategy, you still have better than even odds that someone in the room will lose at least 9 in a row every hour.
Of course the probability of at least one person with a winning streak of at least 8 or 9 (given 0.5% house edge are) 85.1% for 8, and 60.9% for 9. The probability of a winning streak are slightly smaller than that of a losing streak because of the house advantage.
June 10th, 2011 at 10:40:03 PM
permalink
Quote: SilentBob420BMFJChance of this is .0000276666%, or 1 in 36,150.
Somebody check my math, because how can 13 losses in a row be 1 in 18,800, but 20 in a row is only 1 in 36,150? I did .47^x, where x is the number of losses in a row. Then take that number, and move the decimal 2 to the right to get the %. Then to convert % to a number, you just do like if it's .5800, well that's out of 10,000, and you'd divide the latter by the former to reduce it down to 1 in whatever. All that's right, right?
Yea, I think there might be a few errors in your math there. First off, I'm not sure what kind of game you're playing, but winning 53% of the time seems kind of high. In casino blackjack you can only expect to win maybe 45% of hands if playing perfect strategy. Assuming we use your 53%:
Losing 13 in a row: .47^13 = .0000546 or .00546% or 1 in 18,315
Losing 20 in a row: .47^20 = .000000277 or .0000277% or 1 in 3,614,466
June 10th, 2011 at 10:48:49 PM
permalink
I could be wrong but I'm not sure Pacomartin is using the right odds for blackjack. Even if the house advantage is .5% a player is still only expected to win maybe 40-45% because a player wins more for blackjacks, splitting doubling, etc. This means that there is a significantly higher chance of going on losing streaks than winning streaks, even though average earnings will average out in the long run.
As for the 20 hand losing streak I witnessed I guess I should be more specific. A player at my right lost a total of 20 bets without winning a hand, although there were a few pushes in between his 20 loses. I am generally not a very superstitious person, but damn I was pretty ready to leave that table.
As for the 20 hand losing streak I witnessed I guess I should be more specific. A player at my right lost a total of 20 bets without winning a hand, although there were a few pushes in between his 20 loses. I am generally not a very superstitious person, but damn I was pretty ready to leave that table.
June 10th, 2011 at 11:26:26 PM
permalink
Quote: iwannaiguanaI could be wrong but I'm not sure Pacomartin is using the right odds for blackjack. Even if the house advantage is .5% a player is still only expected to win maybe 40-45% because a player wins more for blackjacks, splitting doubling, etc.
That is correct, I don't consider a push to be either a win or a loss. And yes, the house advantage is not really the correct term to use, it is the advantage of winning a hand minus that of losing a hand. That may be over 2%, so that it is closer to the single zero roulette bet.
But more importantly, the chart was supposed to show that house advantage is not a terrifically strong factor.
Plus hypothetically if you lose 9 in a row, is that out 9 out of the last 12 hands you played, or 9 out of the 1000 hands that were played that hour in the casino? This statistic has always been dangerous, and has been misused since the dawn of time. The man that broke the bank at Monte Carlo a century ago, cleverly marketed a winning streak, and published the statistics based on a short list of plays bounded more or less by his winning streak beginning and ending. He was able to get worldwide fame and notoriety by convincing people he had a system. In the modern age when gambling is more widespread and people pay better attention to streaks, the "Man that broke the Bank" would be a minor footnote.
The Wizard has always said in gambling streaks don't really tell you anything. Now in other areas of life they can be very important. If you are a casino owner, streaks can be very profitable since it encourages people to start gambling by betting either on or against the streak. Likewise in the stock market, investors buy and sell stocks based on patterns they think they see. Streaks are the easiest pattern to spot (i.e. if the stock goes up 20 days in a row, they automatically sell). Winning streaks sell tickets to sporting events, and boost tv audiences.
June 11th, 2011 at 5:04:14 AM
permalink
Without taking a mathematical approach, I think gamblers should realize streaks of 12, 16, even 20 in a row are not that unusual in roughly equal chance events.
It's one of the problems with the Martingale. The player thinks he has protection in his plan as it survives what seem to be rare long streaks, then they turn out to be not so rare he finds out.
It's one of the problems with the Martingale. The player thinks he has protection in his plan as it survives what seem to be rare long streaks, then they turn out to be not so rare he finds out.
the next time Dame Fortune toys with your heart, your soul and your wallet, raise your glass and praise her thus: “Thanks for nothing, you cold-hearted, evil, damnable, nefarious, low-life, malicious monster from Hell!” She is, after all, stone deaf. ... Arnold Snyder
June 11th, 2011 at 6:44:56 PM
permalink
Quote: iwannaiguanaYea, I think there might be a few errors in your math there. First off, I'm not sure what kind of game you're playing, but winning 53% of the time seems kind of high. In casino blackjack you can only expect to win maybe 45% of hands if playing perfect strategy. Assuming we use your 53%:
Losing 13 in a row: .47^13 = .0000546 or .00546% or 1 in 18,315
Losing 20 in a row: .47^20 = .000000277 or .0000277% or 1 in 3,614,466
I didn't say winning 53%, I said losing 47%, which would leave winning + tying = 53%. I did the math correctly, with exactly as many zeros as you have in your numbers, but I guess I just misread the 20 in a row decimal to the 10 millionth decimal place, instead of 1 billionth. Must have converted to a percent, then did it, which would explain being off by 2 places.
BTW, after doing some more reading, and it's actually 48.5% loss rate for blackjack.
June 11th, 2011 at 6:53:39 PM
permalink
Nevermind, already addressed.
June 11th, 2011 at 10:04:04 PM
permalink
Quote: odiousgambitIt's one of the problems with the Martingale. The player thinks he has protection in his plan as it survives what seem to be rare long streaks, then they turn out to be not so rare he finds out.
I would say that it is entirely the problem with Martingale. The illusion is that the streaks are much rarer than they actually are. Psychologists have proven this fact countless times, that people have a tendency to underestimate streaks. My personal belief is that it goes against our learned behavior. If we are walking down the street in a storm, you don't expect five or six tree branches to continually fall on you nearly breaking your head. The repetitive behavior of dice or slot machines has no equivalent in nature. Nor does the simple interpretation that some event is good and some are bad.
It is one of the easiest ways to tell if data (like voting data) has been faked. People who fake data, even simple data like a vote for candidate A or candidate B, will not put enough streaks in the data then would be predictable by normal random processes.
June 12th, 2011 at 12:14:42 AM
permalink
I think, maybe, the problem here might be whether "wins or losses in a row" ignores pushes or not.
Say, one loses 48.5% of all rounds dealt, pushes 9%, and wins 42,5%.
Take half of the pushes and add to both win/loss figures if you want to ignore pushes.
So losing x in a row ignoring pushes would be 53%
Winning x in a row ignoring pushes would be 47%
But, getting back what I tried to point out and what Iwannaiguana confirmed the 20 unit loss included some pushes.
I mean it could be such a 20 unit loss, in say 25 rounds, is less than 4 standard deviations and maybe only in the 1 in 8000 range or so?
Yeah, bad, but not crazy bad.
After all, he could have lost 20 units in 3 rounds. That's off the charts, obviously.
Anyway, don't focus so much on streaks, focus more on losing x units over so many dealer upcards.
Getting back to your original post of losing 30 flat-units in, say, 125 rounds, is, while "bad", a total -non-event.
So that is definitely not the point at which to question the legitness of the game.
Say, one loses 48.5% of all rounds dealt, pushes 9%, and wins 42,5%.
Take half of the pushes and add to both win/loss figures if you want to ignore pushes.
So losing x in a row ignoring pushes would be 53%
Winning x in a row ignoring pushes would be 47%
But, getting back what I tried to point out and what Iwannaiguana confirmed the 20 unit loss included some pushes.
I mean it could be such a 20 unit loss, in say 25 rounds, is less than 4 standard deviations and maybe only in the 1 in 8000 range or so?
Yeah, bad, but not crazy bad.
After all, he could have lost 20 units in 3 rounds. That's off the charts, obviously.
Anyway, don't focus so much on streaks, focus more on losing x units over so many dealer upcards.
Getting back to your original post of losing 30 flat-units in, say, 125 rounds, is, while "bad", a total -non-event.
So that is definitely not the point at which to question the legitness of the game.
June 12th, 2011 at 6:02:21 PM
permalink
I like to include pushes, but now that I think about it, it totally changes the math, because the loss % changes, without any other factor being known. So I guess what you need to say is that the chance of losing 20 in a row is x% if you're including pushes, and x% if you're ignoring them. Very important.
Chance of losing 10 in a row:
With pushes: .485^10 = .072% = 1 in 1389
Without pushes: .53^10 = .175% = 1 in 571
HUGE difference. More than double the chance!
OK so from now on this needs to be clarified. I think ties should not be ignored, and one reason is that most people are looking to get entertainment value for their dollar, meaning if they win 100x, lose 100x, and tie 100x, that will take a lot longer, thus more fun. Another reason is that saying that losing twice and then tying 10x is the same as losing twice in a row, come on, you just played 12 hands, how is that losing twice in a row as far as bad luck goes? Total money lost, yes, the same, but other than that, no.
Chance of losing 10 in a row:
With pushes: .485^10 = .072% = 1 in 1389
Without pushes: .53^10 = .175% = 1 in 571
HUGE difference. More than double the chance!
OK so from now on this needs to be clarified. I think ties should not be ignored, and one reason is that most people are looking to get entertainment value for their dollar, meaning if they win 100x, lose 100x, and tie 100x, that will take a lot longer, thus more fun. Another reason is that saying that losing twice and then tying 10x is the same as losing twice in a row, come on, you just played 12 hands, how is that losing twice in a row as far as bad luck goes? Total money lost, yes, the same, but other than that, no.
June 24th, 2011 at 5:51:27 PM
permalink
OK, I just had another HORRIBLE session at this same game at 5Dimes. I started playing really late in the promo time, so I was only able to play for about 90 minutes. In this 90 minutes, I lost 50 units. When I was down 20 units, I figured it would only get worse, and I was curious how much worse; this was like a test almost. And then an hour into it, I was down around 60 units.
Could somebody do the math on this, of being doing 60 units in 1 hour? I don't know how many hands I played, but let's say I played 500 hands in that first hour. So -60 units in 500 hands. And it's a little worse than it appears, because some of the volatility is taken away by the fact that you can't double (at all) or resplit in this game, and also the game has a +EV of about .05%, so really this shouldn't happen IMO. Don't need an exact calculation, but I think this is pretty far out there (at least .001%, or 1 in 100,000), so I'm curious. Definitely more rare than losing 10 hands in a row.
And this is just random paranoia for the record, but if they were to fix their games, what better to fix than the promo games which aren't available often?
Could somebody do the math on this, of being doing 60 units in 1 hour? I don't know how many hands I played, but let's say I played 500 hands in that first hour. So -60 units in 500 hands. And it's a little worse than it appears, because some of the volatility is taken away by the fact that you can't double (at all) or resplit in this game, and also the game has a +EV of about .05%, so really this shouldn't happen IMO. Don't need an exact calculation, but I think this is pretty far out there (at least .001%, or 1 in 100,000), so I'm curious. Definitely more rare than losing 10 hands in a row.
And this is just random paranoia for the record, but if they were to fix their games, what better to fix than the promo games which aren't available often?
July 24th, 2011 at 12:32:25 AM
permalink
Quote: SilentBob420BMFJIn this 90 minutes, I lost 50 units. When I was down 20 units, I figured it would only get worse, and I was curious how much worse; this was like a test almost. And then an hour into it, I was down around 60 units.
Could somebody do the math on this, of being doing 60 units in 1 hour? I don't know how many hands I played, ...
Well, assuming you flat-bet every round, chances are this is a complete non-event. Assuming 500 rounds, 2+ SD. Of course actually knowing how many rounds you played would be useful...
Can you download logs from the casino?
And if, you played, implied 750 rounds in 90 minutes, and finished 50 flat-units down, that's, likely, ridiculously to be expected to happen frequently. Could be as low as 1 in 30 or so.
And, if you weren't flat-betting every round, learn to re-state your results as if you had been flat-betting. Like take a piece of paper and write down W's, L's, T's, winning BJ's and be able to know if you split or doubled. You'll also know how many rounds were played.
Anyway, rather than go thru life feeling persecuted, or that the games cheat, because you think results might be 1 in 100,000 rather than 1 in 200 or 300, spend a little time before you play until you feel you can analyze your results after you play.
Somewhere I picked this formula up
I played t hands of blackjack one at a time, flat betting and religiously following basic strategy. My
final outcome was a gain of r (a negative r represents a loss). The house edge under the rules I
played was h. What is the probability of losing this much or more in a fair game?
Using Excel type this into any cell, substituting the correct values for h, t, and r:
=normsdist((r+t*h+0.5)/(t^0.5*1.16)).
The 1.16 is the standard deviation per hand. This will vary slightly from one set of rule to another
but I feel 1.16 is a good benchmark. Let's take an example. Suppose the number of hands is
1000, the house edge is 0.41%, and the player lost 100 units. The formula for the probability of
losing this many units or more is =normsdist((-100+1000*0.0041+0.5)/(1000^0.5*1.16)) =
normsdist(-2.600700765) = 0.004651712 .
Anyway, it's the formula I tried to use for your question.
Probably, it came from the WIZ but I don't know for sure.
I'm not even saying I really fully understand why it works but I can type it into Excel. Especially that "+0.5" stuff. But, you know, what I lack in brains, I make up in faith and just apply it as best I can. Note, importantly, it assumes flat-betting.
The fine points - over time I think this formula assumes you finish the X hands with y units, not that you achieve losing those y units at some point while you play.
Finishing down 60 units in 500 hands is not the same as being down 60 units at some point during those 500 hands. It therefore also, I came to think, assumes one has adequate bankroll to weather all low points achieved during play.
In other words, I don't think you should use it for Risk of Ruin calculations. I think such calculations would be at least double or so the risk of the above equation since half the time one may be below one's finishing point during play but recover.
Also, I don't see the point of focusing on a limited amount of rounds that may represent "the worst case". At least ask, at the same time, how lucky you were to win 10 units in the last half-hour of play!
July 25th, 2011 at 1:30:18 AM
permalink
Ya I overreact, and I know the game isn't rigged. I just wanted to know how rare it was, with the rarer it was making me feel better. I guess though that I'd have to lose 499 units in 500 hands (exaggeration) for it to be rare.
That formula seems easy. Not sure why you'd need a spreadsheet though, unless you're documenting all your hands. Then again, I don't understand what "=normsdist" is about. Probably a command for a spreadsheet, but just wondering why you need it at all. And your answer for your example came up with like -2.6=.004? I don't get it. Is that saying there's a .4% chance of that result happening? But exact result, not "or worse.
That formula seems easy. Not sure why you'd need a spreadsheet though, unless you're documenting all your hands. Then again, I don't understand what "=normsdist" is about. Probably a command for a spreadsheet, but just wondering why you need it at all. And your answer for your example came up with like -2.6=.004? I don't get it. Is that saying there's a .4% chance of that result happening? But exact result, not "or worse.
July 26th, 2011 at 2:45:32 PM
permalink
Quote: SilentBob420BMFJYa I overreact, and I know the game isn't rigged. I just wanted to know how rare it was, with the rarer it was making me feel better. I guess though that I'd have to lose 499 units in 500 hands (exaggeration) for it to be rare.
That formula seems easy. Not sure why you'd need a spreadsheet though, unless you're documenting all your hands. Then again, I don't understand what "=normsdist" is about. Probably a command for a spreadsheet, but just wondering why you need it at all. And your answer for your example came up with like -2.6=.004? I don't get it. Is that saying there's a .4% chance of that result happening? But exact result, not "or worse.
That's OK - I understand how painful it can feel when mathematically, unfortunately, it may not really be that "rare".
Well, you're right the formula isn't that hard to type into a spreadsheet. I make one row # of hands, the row below house edge, the next row standard deviation, and the next row the flat-bet loss. Then below that I use that formula referencing the cells above to get my answer. I even copy that over 4 or 5 columns. It just makes it easier, and more interesting, to only have to type in number of hands and loss to compare what happens under different assumptions.
Well I used a spreadsheet to try to answer your question. It took me 2 seconds. It's just a tool to at least estimate what it takes to be how far out on that "Lady Luck" bell-shaped curve. Like if you knew the formula and had a spreadsheet, you never would have had to suffer agonizing nights and days about your unbelievable "bad luck". Basically, I was hoping to put you out of your misery and erase any doubt about the fairness of the game.
Yes, in that example it means you finished with a result 2.6 SD below expected value. And, like it said, it also assumed all your play was flat-betting. And, also, like it said, it also means of finishing that bad or worse. In other words, all the area to the left of the curve below -2.6 SD. It also, I believe, assumes you have enough bankroll to never go broke during the 500 hands. So it also means if you played 1000 hand sessions forever 1 in 215 of those would finish with a loss of 60 or more flat-units.
It would get harder to estimate if one is not-flat-betting though. Like maybe in the 501st hand above you bet 1000 units and win and now you're like 500 units ahead but the reality would still be it was an impossible occurrence although the formula would not show that. Although, if an internet game it's really not hard to write down w's,l's t,s and BJ's. And when splits and doubles occur. And later assume a $1 bet/rd.
At least, maybe, it's one tool to start with. Like maybe you could plug in 300 hands in that one hour and see how "rare" that is. etc.
Anyway, good luck to you.
July 27th, 2011 at 12:22:34 PM
permalink
Although it may seem rigged, I doubt it . Why would they risk a very profitable game already ?
I sometimes wonder when the dealers/pit bosses go through and count the cards, or a few were left in the shuffler. I would assume there could be short decks on occassion .
I have been at tables where I could not win a hand, everytime I gotta stiff, hit and busted, everytime I gotta pat hand, the dealer turned and burned. DD's/splits, just cost me more $$$. :l ( still stayed and played, stubborn or dumb, probably both-lol)
I have also had a few( very few) tables where I could not lose . :)
I sometimes wonder when the dealers/pit bosses go through and count the cards, or a few were left in the shuffler. I would assume there could be short decks on occassion .
I have been at tables where I could not win a hand, everytime I gotta stiff, hit and busted, everytime I gotta pat hand, the dealer turned and burned. DD's/splits, just cost me more $$$. :l ( still stayed and played, stubborn or dumb, probably both-lol)
I have also had a few( very few) tables where I could not lose . :)
" If I had the money and the drinking capacity, I'd probably live at a blackjak table and let my life go to hell." Don Pedro
July 27th, 2011 at 12:33:36 PM
permalink
Didn't that baseball team, the Anaheim just lose 17 in a row? It happens.
Someday, joor goin' to see the name of Googie Gomez in lights and joor goin' to say to joorself, "Was that her?" and then joor goin' to answer to joorself, "That was her!" But you know somethin' mister? I was always her yuss nobody knows it! - Googie Gomez
July 28th, 2011 at 11:48:15 AM
permalink
Quote: DonPedroAlthough it may seem rigged, I doubt it . Why would they risk a very profitable game already ?
I sometimes wonder when the dealers/pit bosses go through and count the cards, or a few were left in the shuffler. I would assume there could be short decks on occassion .
I have been at tables where I could not win a hand, everytime I gotta stiff, hit and busted, everytime I gotta pat hand, the dealer turned and burned. DD's/splits, just cost me more $$$. :l ( still stayed and played, stubborn or dumb, probably both-lol)
I have also had a few( very few) tables where I could not lose . :)
That BJ game for them isn't profitable actually. Well, it is, but not nearly as much as your usual games. Remember, this game has about a .07% player advantage. Also no double downs or splitting, so the volatility is low (assuming the 2:1 payout on BJs doesn't boost the volatility a lot).
Yes I too wonder why casinos do things that they think is helping them, when it does nothing. Stuff like stopping a card counter who is flat betting the table minimum and Wonging (me). Who cares as long as people aren't waiting to get on the table.
