Don Johnson's Blackjack Win
Quote: beachbumbabs
btw: if you haven't, you might want to read the forum rules here, and if you haven't noticed it, you're in good company on the ban list here. The point where you made the insult is hyperlinked next to your name, and the other comments would follow from that.
That 473 year ban for BlueEz makes me chuckle every time I look at that list lol. I'm sure there's a funny story behind that which I'm just not aware of.
I also like the wide variety of synonyms for the word "forever" that are used.
I am studying the question of optimal quitting points for Don Johnson's scenario and would
like to put the question up for discussion of whether or not a loss of five chips is an
optimal quitting point for losses. Don was offered a 20% rebate on losses of five chips or
more for the day. 1 chip = $100,000. I believe that -5 is an optimal quitting point for
losses and here is why.
If he is down 5 and takes the rebate he ends the day down 4. If he plays and wins he also is
down 4. If he plays and loses he is down 6 and if he takes the rebate he ends the day down
4.8, so he gains nothing if he wins and he loses 0.8 more if he loses. His best strategy is
not to play another hand if he is down 5 but to quit and take the rebate instead.
If he is down 4 he is not entitled to a rebate unless he plays again and loses, in which
case if he quits and takes the rebate he is still down 4. If he plays again and wins he is
down only 3. He has 1 chip to gain and nothing to lose by playing again. For that reason
down 4 is not an optimal quitting point.
Do you agree?
Quote: puzzlenutOptimal Quitting Points
I am studying the question of optimal quitting points for Don Johnson's scenario
I've never read what the time setting is on the discount/rebate. I don't think it's one "session". The problem with your logic is the period of time involved.
Quote: puzzlenutOptimal Quitting Points
I am studying the question of optimal quitting points for Don Johnson's scenario and would
like to put the question up for discussion of whether or not a loss of five chips is an
optimal quitting point for losses. Don was offered a 20% rebate on losses of five chips or
more for the day. 1 chip = $100,000. I believe that -5 is an optimal quitting point for
losses and here is why.
Do you agree?
No you are wrong.
First you need to figure out what he is maximizing. In my book he is maximizing win per day.
And assuming rebate kicks in every 24 hours (starts again) and that there is only one casino offering the rebate, then your conclucion is wrong.
If there were many casinos offerong the rebate and minimal time delay from travelling from one casino to the other, then your conclusion of optimal quitting point of -5 is correct.
After the -5 (minimum amount that rebate kicks in), the Marginal Increase in the Sessions win rate steadily decreases BUT the absolute Win amount continues to increase and as Teliot and others calculate the exit point is a lot bigger.
ie He is still increasing his win rate by continuing to play after -5 but at a lower and lower rate until an optimal point and then the reduction starts.
If there where 20 casinos one after the other offering this same rebate to him, Yes he just plays to -5 and moves to the next one.
Quote: IbeatyouracesIn my book, he manipulated casino managers to give him an edge on the game. That makes him a cheater, not an AP. Same goes for Ivey.
Just like convincing another poker player he has the better hand when you really do. Must be cheating there too!
......... mhm
If they say "Ok you can play by those rules." ... it's their own fault.
Quote: teliotHere are the results from the LRT for playing a number straight up on double zero roulette, $1000 wager, with a 20% loss rebate.
Loss rebate %: 20.0%
Loss exit point: $37000
Win exit point: $34000
Expected win: $1756
Expected rounds: 37.9
Win percent: 49.3%
What is the loss rebate threshold for these figures, that is, what amount does the loss have to exceed in order for the rebate to apply?
$0 -- but in practice it can be $1 less than the exit point.Quote: puzzlenutWhat is the loss rebate threshold for these figures, that is, what amount does the loss have to exceed in order for the rebate to apply?
Quote: TankoThis is a recent Bloomberg show on Don Johnson:
The whole show is worth a view, but the game rules are discussed at minute 36.
Secrets of a Vegas Whale
Here is the excerpt of link : http://www.blackjackapprenticeship.com/don-johnson-blackjack/
". . . .In addition to the loss rebate program above, Don Johnson piled on more positive expected value. First, Johnson negotiated $50,000 per day in “show-up” money. That is, by simply walking into the casino to play each day, he was given $50,000 cash. Taken together, the loss rebate program and show-up money yielded an expected cash profit of about $175,000 per day. In a recent Bloomberg documentary, Johnson admitted to intentionally creating havoc at the tables to induce frequent dealer errors in his favor. At $100,000 per error, even two errors per day was another $200,000 in his favor. In a later personal email, Johnson admitted to me that he did “other things” (that I can’t disclose) as well . . . . ."
What "other things" Don Johnson did in his Loss Rebate game ?
Please don't get him started again. I like him and don't want to see him take a beating on the subject again. I think someone hacked his account and wrote that.Quote: RorryJust like convincing another poker player he has the better hand when you really do. Must be cheating there too!
......... mhm
If they say "Ok you can play by those rules." ... it's their own fault.
I saw a documentary or TV storey on this man...and he mentioned that he would force the dealer to make mistakes, and take advantage of getting a refund on a lost hand.....as one of the methods of gaining an advantage.
does anyone know what he was talking about. they recreated sexy party girls hanging around him that might break a dealers concentration.
Is it a simple as that?
Quote: LarrySIf this has been covered, just tell me and I will research it
I saw a documentary or TV storey on this man...and he mentioned that he would force the dealer to make mistakes, and take advantage of getting a refund on a lost hand.....as one of the methods of gaining an advantage.
does anyone know what he was talking about. they recreated sexy party girls hanging around him that might break a dealers concentration.
Is it a simple as that?
Hello. I saw the same documentary. The indication on said story was he would make signals for insurance in a vague manner and play it out. I assume taking a hit or not as well, not making a clear signal. I find it difficult to believe as the casino would have their best and brightest on the table but who knows? Money talks and he did win the money.
http://apheat.net/2014/03/25/fact-checking-an-article-on-donald-johnson-in-chance-magazine/
I don't know what DJ's actual strategy was because I wasn't there (and I don't blame him for not wanting to divulge all the details) but when he appeared on GWAE he did, in fact, claim that he would quit when down $500k. I have no idea if he was telling the truth or not (and, again, nor would I blame him if he chose not to tell the truth).
Johnson wants as much misinformation as possible about what he does to get out. He makes a point of it. He is a consummate AP.Quote: AxiomOfChoiceI have no idea if he was telling the truth or not (and, again, nor would I blame him if he chose not to tell the truth).
Quote: teliotJohnson wants as much misinformation as possible about what he does to get out. He makes a point of it. He is a consummate AP.
Yes, of course.
I have no doubt that some of the things that he says are false, but I have no idea which are true and which are not. I am not saying that he actually did quit when down $500k (again, I wasn't there), just that he claimed to.
Note that many people, including educated ones, actually believed that this was optimal, so it's not unbelievable that he did this. Also, it seems like a strange thing to lie about, since the casino certainly knows.
My only point here is, unless you have specific information that you are not sharing, I would not necessarily assume that he used the optimal quitting points that you calculated.
Yes, I do.Quote: AxiomOfChoiceMy only point here is, unless you have specific information that you are not sharing.
Quote: teliotYes, I do.
Fair enough.
Also, if I understand correctly, your optimal quitting points assume that he can play for arbitrarily long before quitting. However, this is not the case. There are only so many hands that someone can play in a day, after all.
Obviously, this makes the question much more difficult (the optimal points change as the number of hands remaining changes), but have you done any calculations to find out what the correct strategy is if we assume that he can't play more than, say, 2000 hands in a day?
(most loss rebate offers have time limits, so I think that this should be a good approach for your general "loss rebate theorem")
Somewhat related: I was reading random stuff on WoO and came across this: https://wizardofodds.com/games/video-poker/tables/guaranteed-play/ I've never seen this game offered anywhere, but it occurs to me that this is basically just a 100% loss rebate that you pay for up-front. So, this brings up two questions:
1. How much should you pay for a loss rebate, and
2. How does everything change if we can no longer assume that results are normally distributed? (the results of 500 hands of VP are NOT normally distributed)
There is also the question of when to quit (you clearly never quit when behind since all your losses are refunded, but how far ahead should you be / how many hands should be left in order to quit?) I recently played a loss rebate and made a nice little profit (relative to the sums involved) but I am sure that I quit too early (although if I put a value on my time, perhaps I didn't)
Regardless of what Johnson may have actually done, the thesis of the article is that Johnson lied about the amount he won, that the authors of the original articles did not double check Johnson's story, and that the magazines did not fact-check their authors. Yikes!Quote: AxiomOfChoiceFair enough.
Quote: teliotRegardless of what Johnson may have actually done, the thesis of the article is that Johnson lied about the amount he won, that the authors of the original articles did not double check Johnson's story, and that the magazines did not fact-check their authors. Yikes!
I totally get your point. But, I'm more interested in the math than the journalism :)
I'll give a look to the time-boundary problem. Hopefully I won't need to invoke the AoC.Quote: AxiomOfChoiceI totally get your point. But, I'm more interested in the math than the journalism :)
As Jacobson makes clear in his fine piece, changing parameters like the win goal and loss limit along with discarding unwanted results are almost guaranteed to produce useless conclusions. Although neither a "scientist" nor a "mathematician," I have had many projects that dealt with both fields and I must say that the distortions created by the folks from the Royal Military College to fit their wanted point sadly mirrors a growing trend in both of those formerly relatively rock-solid fields.Quote: teliotIn case you thought Monte Carlo analysis of loss rebates was easy, http://apheat.net/2014/03/25/fact-checking-an-article-on-donald-johnson-in-chance-magazine/
The more I re-read their report, the more confused I get. I just had to stop.Quote: SanchoPanzaAs Jacobson makes clear in his fine piece, changing parameters like the win goal and loss limit along with discarding unwanted results are almost guaranteed to produce useless conclusions. Although neither a "scientist" nor a "mathematician," I have had many projects that dealt with both fields and I must say that the distortions created by the folks from the Royal Military College to fit their wanted point sadly mirrors a growing trend in both of those formerly relatively rock-solid fields.
From what I can tell, their point is that a couple of magazines could have done some sort of "probabalistic fact-checking" to conclude that DJ didn't tell the whole truth to the authors of those articles. In demonstrating the kind of fact checking they had in mind, the RMC professors took a couple of anecdotal comments by Don Johnson, built a Monte Carlo simulation with those comments as quit-points and an incorrect blackjack model, ran a small simulation and concluded that DJ exaggerated his winnings. Yikes!
Their bj model got the probability of a blackjack wrong (they didn't account for a bj-bj tie), and only had outcomes of +1.5, +1, 0 and -1. They stated that they could ignore doubles and splits because those events were rare. They didn't consider the fact that DJ played a game that offered surrender.
It was reported in one of the articles that Don Johnson played 12 hours on *one* day (a day where he was playing against a specific dealer), so the authors assumed that DJ would always play until he either hit 12 hours (720 hands), or lost $500k, whichever came first.
Quote: teliotThe more I re-read their report, the more confused I get. I just had to stop.
From what I can tell, their point is that a couple of magazines could have done some sort of "probabalistic fact-checking" to conclude that DJ didn't tell the whole truth to the authors of those articles. In demonstrating the kind of fact checking they had in mind, the RMC professors took a couple of anecdotal comments by Don Johnson, built a Monte Carlo simulation with those comments as quit-points and an incorrect blackjack model, ran a small simulation and concluded that DJ exaggerated his winnings. Yikes!
Their bj model got the probability of a blackjack wrong (they didn't account for a bj-bj tie), and only had outcomes of +1.5, +1, 0 and -1. They stated that they could ignore doubles and splits because those events were rare. They didn't consider the fact that DJ played a game that offered surrender.
Don Johnson played 12 hours on *one* day (a day where he was playing against a specific dealer), so the authors assumed that DJ would always play until he either hit 12 hours, or lost $500k, whichever came first.
lol. Doubles and splits are rare so they can be ignored. That is a new one!!
Here is their exact quote:Quote: AxiomOfChoicelol. Doubles and splits are rare so they can be ignored. That is a new one!!
"We have not modeled some standard blackjack betting outcomes explicitly, outcomes such as splitting pairs, which can be favorable to the gambler. These situations occur so infrequently, they can be ignored without materially affecting the results described herein."
Here is their blackjack model:
Published in CHANCE magazine, a publication of the American Statistical Association.
Quote: teliotHere is their exact quote:
"We have not modeled some standard blackjack betting outcomes explicitly, outcomes such as splitting pairs, which can be favorable to the gambler. These situations occur so infrequently, they can be ignored without materially affecting the results described herein."
Here is their blackjack model:
Published in CHANCE magazine, a publication of the American Statistical Association.
Edit: Misread the column title..
Ok I am confused. What is the 1.5 in the bottom row, middle column?
Quote: AxiomOfChoiceSo, their blackjack model does not allow doubles or splits, but still has a house edge of only 0.25%? That does not sound right. Is this because the player wins all blackjacks?
They didn't model actual blackjack. They actually went backwards. They wrote,
"We make the following assumptions ... Blackjack occurs 4.833% of the time and pays $150,000 on a $100,000 bet. Tied hands—ones where the gambler’s stake is returned— occur 9.82% of the time."
They then split the wins and losses of 1 unit into the appropriate amounts to get a house edge of 0.25%.
They simply generated a random number between 0 and 1, and depending on its value, the result was 1.5, 1, 0 or -1.
I swear, your head would explode if you read it.
Quote: Mission146I could have modeled it 20 times better than that, and I suck! Do you guys think I have the chops to be an accredited Statistician? If they do, I can't be far behind.
There is no accreditation. Just call yourself a statistician.
Quote: teliotThey didn't model actual blackjack. They actually went backwards. They wrote,
"We make the following assumptions ... Blackjack occurs 4.833% of the time and pays $150,000 on a $100,000 bet. Tied hands—ones where the gambler’s stake is returned— occur 9.82% of the time."
They then split the wins and losses of 1 unit into the appropriate amounts to get a house edge of 0.25%.
They simply generated a random number between 0 and 1, and depending on its value, the result was 1.5, 1, 0 or -1.
I swear, your head would explode if you read it.
So, they got the EV more or less right, but are way off on the variance.
Which, you know, isn't that important for a loss rebate.... </sarcasm>
Yes, variance is the last thing you would want to waste your time with when modeling loss rebates.Quote: AxiomOfChoicesarcasm>
Curiously, we had this conversation with Puzzlenut a few weeks back.
None of these words mean anything. I call myself a mathematician, a computer scientist, or a computer engineer depending on what suits me at the time.
Yesterday I was talking to a (very) senior engineer about an interview question that he liked to ask, expecting people to give a dynamic programming algorithm to solve it. I said "that's just m+n choose n". My friend (sitting beside us) said "spoken like a true mathematician". So yesterday I was a mathematician. Or, at least, I had a basic understanding of first-year combinatorics.
Quote: teliotJohnson wants as much misinformation as possible about what he does to get out. He makes a point of it. He is a consummate AP.
What kind of consummate AP goes public with TV interviews, newspaper articles, high profile parties, etc? Or is that part of his dysinformation campaign?
After winning $15M, it was going to be hard to keep his win a secret. At the World Game Protection Conference, he didn't say much about what he actually did. He gave up a piece of his +EV in AC in his recent Bloomberg documentary spot.Quote: 21formeWhat kind of consummate AP goes public with TV interviews, newspaper articles, high profile parties, etc? Or is that part of his disinformation campaign?
Somehow, DJ does continue to get +EV blackjack playtime in casinos.
Quote: AxiomOfChoiceMost people who I consider to be statisticians tell me that they are not statisticians.
None of these words mean anything. I call myself a mathematician, a computer scientist, or a computer engineer depending on what suits me at the time.
Yesterday I was talking to a (very) senior engineer about an interview question that he liked to ask, expecting people to give a dynamic programming algorithm to solve it. I said "that's just m+n choose n". My friend (sitting beside us) said "spoken like a true mathematician". So yesterday I was a mathematician. Or, at least, I had a basic understanding of first-year combinatorics.
While everything you say is true, when you cross the line into electrical engineering (which is closely related to computer engineering), the titles actually start to matter as you can take the PE exam and become a Professional Engineer (same with Mechanical, Civil, etc.). Thankfully the software/CS fields don't have anything similar, or else I would have had to study for it :P
Marketed as a way for the state to insure the integrity of the industry, it's just another excuse for the state to employ an office full of lardbutted relatives and friends of the political elite, shuffling paper and collecting fees.
I actually have such a license, which no one has ever asked to see, just in case. I actually has to pass a test to get it, but it wasn't hard. There was just one question on the test.
In case you want such a license, here is the question, so that you can study for the exam.
Ready?
"Have you sent us $150?"
If the answer is yes, then congrats, you are a licensed telecommunication specialist. They change the question every year or so to insure that we all stay current with the latest state of art in the industry. Next years question will probably be "Have you sent us $200?"
Quote: AcesAndEightsWhile everything you say is true, when you cross the line into electrical engineering (which is closely related to computer engineering), the titles actually start to matter as you can take the PE exam and become a Professional Engineer (same with Mechanical, Civil, etc.). Thankfully the software/CS fields don't have anything similar, or else I would have had to study for it :P
Yes, that is true. None of the fields that I mentioned have professional accreditation.
Quote: rhodyBobActually, RI has a license for what they call "telecommunication specialist". You are supposed to have such a license if you want to take the covers off a computer (or whatever passes for a "cover" nowadays). Extends to stringing cable, fiber, or anything related to computers.
Somewhere, in a datacenter far, far away, are computers which run the code that I write. I don't really know any more than that.
He was not the mastermind of this operation.
He was a low-mid BJ Civilian player playing recreationally with a small disadvantage and being smart enough not to lose too much and get good comps.
He was also professional into racing gamling, betting high with advantage. Being in this professionally he came accross other professional APs who were also into BJ. The documentary mentions Phd Mathematicians who helped Don develop mathematically the strategies but declined to appear on the program.
It appeares to me that they are the masterminds of the operation. Not only they developed the strategies on Don's request but they probably approached with with the operation.
The operation was that he would increase his betting for a period (say play for 1-2 months at higher stakes) with a negative EV, so that he will be recorded in the casinos datrabase as a Negative Ev player.
And then go to them and negotiate very good terms (including the 20% rebate) to play at very high stakes.
Based on his past performance (which was at a relatively high stake) the casinos accomodated.
The real operators (refererred to as the Phd mathematicians) probably partly bankrolled the operation and had a big % share on the profits.
+1000 your exactly right. This is why all the information is mixed up and changes. He is not doing this to confuse anyone as an AP play he is confused himself on the math of everything and the details including the exact profit numbers that are probably inflated in the first place. He wants the notoriety. The masterminds probably have the next front man in place to pull this or something else off. I have a feeling he thought he was burnt out and is using this as a way turn this into other opportunities like books, TV, investors etc etc.Quote: AceTwoFrom what I have read about the Don Johnson story, it appears to me that the following has happened.
He was not the mastermind of this operation.
He was a low-mid BJ Civilian player playing recreationally with a small disadvantage and being smart enough not to lose too much and get good comps.
He was also professional into racing gamling, betting high with advantage. Being in this professionally he came accross other professional APs who were also into BJ. The documentary mentions Phd Mathematicians who helped Don develop mathematically the strategies but declined to appear on the program.
It appeares to me that they are the masterminds of the operation. Not only they developed the strategies on Don's request but they probably approached with with the operation.
The operation was that he would increase his betting for a period (say play for 1-2 months at higher stakes) with a negative EV, so that he will be recorded in the casinos datrabase as a Negative Ev player.
And then go to them and negotiate very good terms (including the 20% rebate) to play at very high stakes.
Based on his past performance (which was at a relatively high stake) the casinos accomodated.
The real operators (refererred to as the Phd mathematicians) probably partly bankrolled the operation and had a big % share on the profits.
LOL..... You said this teliot " The more I re-read their report, the more confused I get. I just had to stop.Quote: teliotAfter winning $15M, it was going to be hard to keep his win a secret. At the World Game Protection Conference, he didn't say much about what he actually did. He gave up a piece of his +EV in AC in his recent Bloomberg documentary spot.
Somehow, DJ does continue to get +EV blackjack playtime in casinos.
From what I can tell, their point is that a couple of magazines could have done some sort of "probabalistic fact-checking" to conclude that DJ didn't tell the whole truth to the authors of those articles. In demonstrating the kind of fact checking they had in mind, the RMC professors took a couple of anecdotal comments by Don Johnson, built a Monte Carlo simulation with those comments as quit-points and an incorrect blackjack model, ran a small simulation and concluded that DJ exaggerated his winnings.
Then you quote the same numbers he is boasting about.
He didnt say much about what he did probably because, he didn't know exactly what he did.He knew if he said to much, they would realize he didn't really know much of anything. He was told what and how to do it, he followed directions and handed over the cash to the people who set this up. He was just a front man following the lead of people who knew exactly what was needed to be done and how to do it.
Quote: AceTwoHe was also professional into racing gamling, betting high with advantage. Being in this professionally he came accross other professional APs who were also into BJ. The documentary mentions Phd Mathematicians who helped Don develop mathematically the strategies but declined to appear on the program
If this Phd theory is right, I wonder how much of it was wanting the money [I assume they benefited] and how much of it was getting some pleasure out of gaming the system.
Part of Don Johnson's strategy was to take advantage of any dealer errors that occurred. He said repeatedly that whenever a dealer made a mistake, he got a free bet. What did that entail? If the dealer made a mistake in the middle of a hand, that hand became a freeroll? The next hand became a freeroll? He was instantly paid?
