Overhauling the Betting System Challenge

What you are doing here is effectively only making ~1000 bets and hoping variance helps you out.

It's not just the stop limit that hurts this... it's the stop win/sit out clause as well.

If you forced the initial big winners to then make min bets, I couldn't find a way to make the odds of winning to go high enough to be worthwhile to beat a 10:1 bet (you need 38 initial winners to give you enough cushion... 2.5% chance of that).

If you knew your lifetime totals so far, you could probably do some magic to increase your chances.

Quote: thecesspit

I make your chance of this schema working 16.3%

45% chance of hitting 28 or more winners on the first and 36% chance of hitting one or more winners on the second stage. = 16.2%

I'll agree with the first number (roughly, my estimate is somewhat lower), but 1 or more out of 28 is, I think 1-(37/38)^28 = 52.61% for the total of about 23% (I said 20% initially, because my number for 28 or more out of 1000 was about 40.3% ... I could not compute it analytically, and had to resort to simulation, so I am not totally sure about the accuracy).

I am still thinking about the "no sit out" variant, but it looks like I am going to have to agree with you, that having to continue making bets after a win, kills the scheme even if you keep the limit on the bankroll (I think, you need even more than 38, because, if you win 175000 on the first bet, it gets added to your bankroll, and you can't stop after losing just $5000, and have to continue for the expected lifetime gain of only about 85,000)

I think that leads to a good next tweak that still avoids the "no sitting out" problem -- as stated in the rules, you'll play additional lifetimes *unless* you bust out early, but the goal is to minimize the number of trials and let variance trump mean. Version 2:

On a single-zero roulette wheel, place a $5000 bet on any inside number. If it wins, go to the craps table and bet $5 on don't pass for the rest of the lifetime. Or if the single-wheel table offers partage, stay there and bet $5 on any outside bet.

1 in 37 times you'll end a lifetime up roughly 150k, the other 36 times you'll lose 5k. The average number of inside number hits in 1000 plays is 27. If you get 33, 6 more than average, you'll be ahead for all 1000 lifetimes. That's getting lucky but not unbelievably so.

MathExtremist's original idea, modified. I tweaked MathExtremist's original system (bet $5000 each hand to try to hit $50,000, then ride it out with small bets). My tweaks were (1) having it take full odds, (2) betting only as much as would be necessary to get to the $50,000 (assuming that we'd be placing an Odds bet and that it would win), and (3) betting more again to try to get back to that $50,000 goal if we ever drop below it, rather than just always betting $5 until the end of the lifetime after hitting the $50k goal even once. The results weren't much different. EV ranged from -$1350 to -$2400, and winning lifetimes were 4.2% to 6.2% of the lifetimes (winning ~$54,000 during winning sessions).

Ayecarumba's system. Ayecarumba didn't give instructions for when to start betting again after a "stop". If the idea was to give up for that lifetime, that would clearly fall under the category of "trying to exploit the challenge" instead of "trying to prove a winning betting system". After all, what system seller says that if you start losing you then have to stop playing *for your entire lifetime*?! But it would indeed be allowed under the current rules. All the challenger has to say is, "If bankroll drops to $2500, then the table is cold and we have to wait for 50 winning Pass Line bets in a row before we start betting again," which would effectively end the lifetime without making any more bets. However, I don't think it matters if we allow this exploit, because it's not really an exploit. Remember that we bet additional lifetimes for hands that were sat out, so ending lifetimes early doesn't ultimately reduce exposure to the house edge, because the full number of hands are going to eventually be played.

I did run a sim and results showed only a ~3% win rate and an EV of -$2600, and that's *without* adding in the additional lifetimes for sitting out. Then again, I'm not entirely confident that I programmed it correctly, because there's lots of room for error. If anyone wants to see my Java code then let me know and I'll send it to you. Also, I'm not entirely sure I understood the system instructions. Here's my interpretation:

1. If bankroll <= $2501, then stop, otherwise...
2. If bankroll >= $12,501 go to step 5, otherwise...
3. If there have been at least three consecutive come-out sevens, then bet $400 on the Don't Pass.
4. After the round is over, go to step 1.

5. If the Lay 10 bet is not working and the bankroll would be >$7501 after laying the 10 for $11, then Lay the 10 for $11, otherwise stop.
6. If the Hard 10 bet is not working and the bankroll would be >$7501 after putting $1 on the Hard 10, then put $1 on the Hard 10, otherwise stop.
7. After the Lay 10 or Hard 10 bet is resolved, go to step 5.


@JerryLogan: "Are video poker machines included?" Answer: Read the rules!

Weaselman: (paraphrased) "If a system has a greater than 10% chance of showing a profit then it could win your challenge, since you're offering 10:1 odds, even if the overall EV was negative. How will you guard against this? Okay, I understand you better now, and I'm sorry I barked at you. The short answer is that I just have to make sure that we play the game long enough that a negative expectation is a virtual certainty. In theory I could say that we run the sim 10 times and the majority wins, but (1) that makes the rules way too complicated for my taste, and (2) if I felt I had to do that, that would mean the sim didn't include enough rounds in the first place. The way to avoid the 10% chance of winning is to make the sim just like the long term in the casino, where the player always loses. Does that answer adequately?

thecesspit: "This is the way I would try and 'break' the system. You'd then add in X new rolls (where X is the number left from the previous life time). The idea being to limit losses while making several outlier shots at big wins. Once your ahead by enough, eke out the rest of the lifetime making patty-cake flea bite bets. That was essentially MathExtremist's idea -- build a big bankroll and then bet like a flea. But it didn't work, because the improbability of building the large bankroll + the large number of rounds played will still grind you down.

Weaselman: (paraphrased) "Bet the whole bankroll on #0 in roulette. If we lose, we've eaten up a whole life, reducing our exposure to the house edge. If we win, we sit out the rest of the lifetime. (The system could say to wait until #13 hits 359,999 times in a row, as an excuse to sit out.) Yes, we'll have to run additional lifetimes to compensate for the rounds we're sitting out, but in those lifetimes we have a chance to lose on the first spin again which will eat up another whole life, reducing our exposure to the house edge again." This is very clever, but it doesn't work. A lifetime is normally 360,000 rounds, and there are normally 1000 lifetimes. If you go bust on spin #1 of a lifetime, you do indeed eat up that lifetime. But if you win on spin #1 and sit out for the rest of the lifetime, then we've got 359,999 rounds to make up in extra lifetimes. Since your system has only one lifetime per spin, win or lose, a win on spin #1 of your first lifetime means that we're gonna simulate 359,999 more lifetimes. And for each of those lifetimes, when you win on spin #1, we add another 359,999 lifetimes to the number that we're ultimately going to run, up to a maximum of 360,000,000 one-round lifetimes. The EV of this system is -$320.

MathExtremist: (paraphrased) "Variation of Weaselman's system: We bet the farm on a single number on the first spin, but if we win then rather than sitting out, we go play Craps making a minimum bet on the Pass Line for the rest of the lifetime." I ran this 26 times and on the 26th it showed a positive EV. The system is still an overall loser, of course, but this is still too close for comfort. I bumped up the number of lifetimes to 5000 and then the closest it came to a positive EV in 100 trials was -$51, which is still too close. I could increase the number of lifetimes even more, but I'm thinking it might be better to step back and try to see the forest instead of the trees.

First of all, this is just a clear attempt to *win the challenge* rather than *proving the worthiness of a betting system*. So rather than fiddling with the parameters so it can't win, I think the better approach would be to just disqualify this kind of chicanery in the challenge terms. But I'm at a bit of a loss as to how to do that. One idea is to say that the system has to have an overall positive EV *and* it has to win at least 25% of lifetimes. In that system I just ran, it typically won fewer than 3% of lifetimes.

Another idea is to do away with the concept of lifetimes, because critics could still say that a lifetime is too long, and that their system shows a profit on weekend trips. So I could say our base time period is a ten 10-hour weekend trips, and we average 100,000 of those. That still leaves me with a similar problem that I was trying to solve, though.

The last idea is to go back to just running a bunch of hands one time, but make it less than a billion. I haven't run the numbers, but I'm guessing that 10 million would be nearly impossible to beat. However, even though 10 million is pretty far from a billion, it would probably invite the same kind of criticism from skeptics who would say that "A system might not work for 10 million hands, but it could be great in the short term."

Thoughts?

Quote: MichaelBluejay

MathExtremist: (paraphrased) "Variation of Weaselman's system: We bet the farm on a single number on the first spin, but if we win then rather than sitting out, we go play Craps making a minimum bet on the Pass Line for the rest of the lifetime." I ran this 26 times and on the 26th it showed a positive EV.

I win!

Quote:

First of all, this is just a clear attempt to *win the challenge* rather than *proving the worthiness of a betting system*. So rather than fiddling with the parameters so it can't win, I think the better approach would be to just disqualify this kind of chicanery in the challenge terms. But I'm at a bit of a loss as to how to do that.

I'd start by suggesting that "Bet the table max on your first bet" isn't a system that anyone would pay for or can even afford to play. But more directly, how about having session bankrolls instead of just one lifetime bankroll. I probably don't use more than $300-500 in initial buy-in, mostly because I don't like carrying wads of cash or setting up credit lines. If I do that 10 times a year, and lose it all each time, I might be down $3k-5k for the year. But if I bust out my $300, it's not like I'm going to quit forever.

So rather than allowing someone to game the system by quitting early, just start with a trip bankroll of $1000 or whatever the system vendor suggests, and then force the issue by saying "okay, let's play a lifetime's worth of trips". 20 trips/year for 30 years seems reasonable - I take far fewer trips/year, but I don't plan on stopping after 30 years... Most trips are broken down into sessions and if the system uses a session bankroll instead, that's even better (e.g. more trials). But don't allow a lifetime bankroll cap - that needs to be effectively unlimited in order to do a proper test.

Quote: MichaelBluejay

If you go bust on spin #1 of a lifetime, you do indeed eat up that lifetime. But if you win on spin #1 and sit out for the rest of the lifetime, then we've got 359,999 rounds to make up in extra lifetimes. Since your system has only one lifetime per spin, win or lose, a win on spin #1 of your first lifetime means that we're gonna simulate 359,999 more lifetimes.
And for each of those lifetimes, when you win on spin #1, we add another 359,999 lifetimes to the number that we're ultimately going to run, up to a maximum of 360,000,000 one-round lifetimes.

I don't see how that follows from the rules. If I bust on the first bet, there is no sit out, so, no additional rounds should be added (I could only see additional lifetimes mentioned in the rules in connection to the sit-out requirements, nothing about busting early).
If my first bet wins, then I do sit out, the whole lifetime, so the whole new lifetime (yes 359,999 rounds) is added to the end. If I win 28 times, that'll cost me 28 additional lifetimes, not 360,000,000.

Can you show where exactly this contradicts your rules, and where do your rules say that if I want to sit out a (part of) *one* lifetime, you will add 360,000 more to the list?

Quote:

First of all, this is just a clear attempt to *win the challenge* rather than *proving the worthiness of a betting system*.

Isn't that what you asked for?

Quote:

So rather than fiddling with the parameters so it can't win, I think the better approach would be to just disqualify this kind of chicanery in the challenge terms. But I'm at a bit of a loss as to how to do that. One idea is to say that the system has to have an overall positive EV *and* it has to win at least 25% of lifetimes. In that system I just ran, it typically won fewer than 3% of lifetimes.
Another idea is to do away with the concept of lifetimes, because critics could still say that a lifetime is too long, and that their system shows a profit on weekend trips. So I could say our base time period is a ten 10-hour weekend trips, and we average 100,000 of those. That still leaves me with a similar problem that I was trying to solve, though.

Just get rid of the bankroll limitation. It is not only unnecessary, but also harmful in this case.

"@JerryLogan: "Are video poker machines included?" Answer: Read the rules!"

OK sorry I didn't see that it's offered for BJ, craps and baccarat. But you mentioned that you were out to prove that ANY betting system cannot win, and we have a professional gambler amongst us (I think he's made one post so far) who claims to have won nearly a million dollars with his self-developed system, and he says he plays only -EV vp machines.

Wouldn't that be something to consider to put your hands around, rather than a bunch of questions and comments from people who will never become a part of the challenge anyway, especially with theoretical strategies and not true-tested ones?

I like the idea of 10,000 weekends... it would be "short term" enough for the people who can't understand that the short term is part of the long term, and also show what would happen is 10,000 people brought their magical system. I also suspect for 10,000 weekends it'd be much harder to game the system to be one shot....

However, a quick analysis of 100,000 single $5000 bets on one number in Roulette shows that there's a 7.2% chance that you get 2778 winners or more... which results in a positive result.

You need some policy where the system can't make single bets over whatever course of events you are looking at.... how about a weekend is not "over" until a minimum number of -bets- are made, or the player busts?

Quote: MichaelBluejay

Ayecarumba's system. Ayecarumba didn't give instructions for when to start betting again after a "stop". If the idea was to give up for that lifetime, that would clearly fall under the category of "trying to exploit the challenge" instead of "trying to prove a winning betting system". After all, what system seller says that if you start losing you then have to stop playing *for your entire lifetime*?! But it would indeed be allowed under the current rules. All the challenger has to say is, "If bankroll drops to $2500, then the table is cold and we have to wait for 50 winning Pass Line bets in a row before we start betting again," which would effectively end the lifetime without making any more bets. However, I don't think it matters if we allow this exploit, because it's not really an exploit. Remember that we bet additional lifetimes for hands that were sat out, so ending lifetimes early doesn't ultimately reduce exposure to the house edge, because the full number of hands are going to eventually be played.

I did run a sim and results showed only a ~3% win rate and an EV of -$2600, and that's *without* adding in the additional lifetimes for sitting out. Then again, I'm not entirely confident that I programmed it correctly, because there's lots of room for error. If anyone wants to see my Java code then let me know and I'll send it to you. Also, I'm not entirely sure I understood the system instructions. Here's my interpretation:

1. If bankroll <= $2501, then stop, otherwise...
2. If bankroll >= $12,501 go to step 5, otherwise...
3. If there have been at least three consecutive come-out sevens, then bet $400 on the Don't Pass.
4. After the round is over, go to step 1.

5. If the Lay 10 bet is not working and the bankroll would be >$7501 after laying the 10 for $11, then Lay the 10 for $11, otherwise stop.
6. If the Hard 10 bet is not working and the bankroll would be >$7501 after putting $1 on the Hard 10, then put $1 on the Hard 10, otherwise stop.
7. After the Lay 10 or Hard 10 bet is resolved, go to step 5.

That's pretty much it, except for laying max free odds if a point is established in #3. Stop means the table is cold. Stop your losses, or protect your winnings (for the purposes of the challenge, effectively for the rest of that lifetime. It is understood that extra lifetimes would be required to make up the sit-outs, but hopefully, stopping losses at a relatively high bankroll will allow the averaging to wash them out.

Quote: MichaelBluejay

Another idea is to do away with the concept of lifetimes, because critics could still say that a lifetime is too long, and that their system shows a profit on weekend trips. So I could say our base time period is a ten 10-hour weekend trips, and we average 100,000 of those. That still leaves me with a similar problem that I was trying to solve, though.

The last idea is to go back to just running a bunch of hands one time, but make it less than a billion. I haven't run the numbers, but I'm guessing that 10 million would be nearly impossible to beat. However, even though 10 million is pretty far from a billion, it would probably invite the same kind of criticism from skeptics who would say that "A system might not work for 10 million hands, but it could be great in the short term."

Thoughts?

One of the foundational concepts of a "real" system would be that it would win, regardless of the number of trials. One billion trials should produce 10,000 times the profit as a 10M trial run. I think a rule that, "75% of trials must produce a profit of at least $1" is generous (it should be 99.9%) and reasonable. It should protect you from schemes designed to game the rules.

The minimum profitable trials could be set even higher without dissuading serious challengers. Anyone hoping to sell a system is going to make bigger promises, since they will not get many takers if their catch line is, "Buy my system and you may only lose 25% of the time!"

System sellers would probably say that there was not enough DISCIPLINE or you didn't apply the proper bank roll, or using a computer doesn't simulate the real world, you damn Math Guys (LOL LOL).

It's all smoke, mirrors and selling the idea that any failure is the PLAYER'S fault.

Quote: Ayecarumba

One of the foundational concepts of a "real" system would be that it would win, regardless of the number of trials.

Actually, a "real" (that is, profitable) system would be more likely to win as the number of trials increased, not less likely as is the case with casino games. Perhaps that's a way to proceed - do trials of:

100
200
400
800
1600
3200
6400
12500
25000
50000
100000
200000
400000

By the time you get to 400k, you're at roughly a lifetime of play. The results of the system will converge on the house edge (obviously) and that's going to be negative. What you'd expect in a positive, profitable system is wide percentage swings in the smaller trial runs and convergence on a +EV toward the bottom of the list. What you'll get is wide swings in the smaller runs and convergence on a -EV instead.

FWIW, I think, that Wizard's original challenge was way better (not taking into account procedural changes - like the arbitration clause, signing of the contract before disclosing the system, the time limits etc., that all make sense), it is simple, easy to explain and understand, and right to the point - that is, it doesn't leave an impression that in reality only *some* betting systems (those confirming to a relatively strict set of rules) do not work, while the others still have a chance.
If someone says a billion is too many, you can always negotiate a lower number for that particular case (you have a lot of wiggle room), and/or, you can explain to them (at that point) that that is roughly how many bets two thousand people who buy the system will make in their lifetimes (and also, that a winning system should be expected to produce more profit as the number of rounds increases, like MathExtremist pointed out).
But trying to spell all that out in the contract, just makes it look more like one of those infamous mortgage applications that are worded specifically with the purpose of concealing the information rather than conveying it.
Oh, and also, despite its simplicity and straightforwardness, Wizard's system is actually a lot harder (impossible, AFAICS) to beat or exploit.

Why fix what's not broken?

Quote: weaselman

FWIW, I think, that Wizard's original challenge was way better (not taking into account procedural changes - like the arbitration clause, signing of the contract before disclosing the system, the time limits etc., that all make sense), it is simple, easy to explain and understand, and right to the point - that is, it doesn't leave an impression that in reality only *some* betting systems (those confirming to a relatively strict set of rules) do not work, while the others still have a chance.
If someone says a billion is too many, you can always negotiate a lower number for that particular case (you have a lot of wiggle room), and/or, you can explain to them (at that point) that that is roughly how many bets two thousand people who buy the system will make in their lifetimes (and also, that a winning system should be expected to produce more profit as the number of rounds increases, like MathExtremist pointed out).
But trying to spell all that out in the contract, just makes it look more like one of those infamous mortgage applications that are worded specifically with the purpose of concealing the information rather than conveying it.
Oh, and also, despite its simplicity and straightforwardness, Wizard's system is actually a lot harder (impossible, AFAICS) to beat or exploit.

Why fix what's not broken?

I think BJ's goal is to have a challenge that is closer to the "reality" of a single player's experience. A single player cannot possibly play 1B rounds, nor comprehend how large a number that really is. I agree it is difficult to analyze and execute changes, but appreciate the effort to educate the public.

Quote: Ayecarumba

A single player cannot possibly play 1B rounds,

and a thousand lifetimes?

Quote: weaselman

and a thousand lifetimes?

Well, I did say it was difficult... Personally, 1,000 lifetimes is more "real" since I can accurately visualize a roomful of that many individuals. As for how much action each of them take over a lifetime, I have no idea.

I go back to my suggestion to have a minimum qualifying percentage of winning "lives" / "rounds" / "sessions" (or whatever is determined to be the acceptable division of minimum total decisions). From a potential customer's perspective, I think this is what they really want to know, "Does it work for me?"

System promoters should not hesitate to agree to an 80%, or 90% "winning lives" percentage minimum.

Quote: thecesspit

..... how about a weekend is not "over" until a minimum number of -bets- are made, or the player busts?

Maybe I'm still missing something.

Isn't the factor that you have been exploiting that the player can bet the entire bankroll on a high-odds bet as the first wager? Lose, go to another lifetime; win, try to hold on and hope you come out positive overall for 1000 lifetimes (hopefully with a better than 10% chance of that happening).

If that's it, doesn't letting the lifetime end when the player busts (maybe after one wager) still allow this exploitation?

MathExtremist:

"I win!"

-- Well, actually Weaselman beat you to it. I mis-programmed his original idea before your modification. Once I re-ran it then he won, so he was first in line. See below.

"I'd start by suggesting that "Bet the table max on your first bet" isn't a system that anyone would pay for or can even afford to play."

-- I agree that it's not a system anyone would pay for or could afford to play. But I don't know how to disallow it. If I said you couldn't bet everything on your first round, people would just bet everything on their second round. If I said you couldn't *ever* bet the farm, people would then bet half. If I limited bets to 1/4th the initial bankroll, then suddenly it looks to the public like my rules are really restrictive and maybe a betting system could work if bets were unlimited. My main goal is to convince the public who reads my page about the challenge that betting systems don't work, otherwise someone would have accepted it. So I do need the rules to be as simple as possible, and as non-restrictive as possible.

"But more directly, how about having session bankrolls instead of just one lifetime bankroll.... Don't allow a lifetime bankroll cap - that needs to be effectively unlimited in order to do a proper test."

-- I see that this would limit the number of rounds a bettor could eat up would be limited. But in the real world nobody has an unlimited bankroll, and I'm trying to make this as "lifelike" as possible. Still, I need a solution, so let's explore this idea. How long should a session be, and what should the starting bankroll for a session be? 120 rounds and $500?


Weaselman:

" If I bust on the first bet, there is no sit out, so, no additional rounds should be added."

-- Yes, you're right. I mentioned that but then forgot to limit the number of lifetimes added. I reprogrammed it and it showed a positive EV in 9 out of 25 trials (36%). That's more than enough to exploit the 10-1 challenge payoff, so I conclude that you won my $1000 reward. I can pay you by PayPal, or I can meet you if you're in Vegas.

For those who didn't quite follow this, the clever part of Weaselman's system was that it reduces the number of rounds from 360,000,000 to just over 1000. Because rather than playing 360,000,000 rounds per lifetime, he plays only *one* per lifetime. I did find that I can thwart this exploit by averaging 1 million lifetimes rather than 1000. That puts the number of rounds at just over 1 million, which is exceptionally hard to beat, even on a single-zero wheel (which I just simulated).

Bluejay: "First of all, this is just a clear attempt to *win the challenge* rather than *proving the worthiness of a betting system*."
Weaselman: "Is that what you asked for?"

-- Absolutely. My point was that I now need to focus my attention on how to thwart the exploits.

JerryLogan:

JerryLogan: "Are video poker machines included?"
Bluejay: "Read the rules!"
JerryLogan: "OK sorry I didn't see that it's offered for BJ, craps and baccarat."

-- No, that is NOT what the rules state. Again, read the rules. And by the way, the reason VP isn't covered is that the whole point of this exercise is to show that BETTING SYSTEMS FOR TABLE GAMES don't work. It's the same reason the challenge doesn't try to mow your lawn or cure cancer.

thecesspit "I like the idea of 10,000 weekends... it would be "short term" enough for the people who can't understand that the short term is part of the long term, and also show what would happen is 10,000 people brought their magical system."

-- It's not 10,000 weekends, it's 100,000 weekends.

AyeCarumba: "That's pretty much it, except for laying max free odds if a point is established in #3."

-- Yes, thanks, I forgot to mention that in my post, but I did include it in the sim.

Weaselman: "If someone says a billion is too many, you can always negotiate a lower number for that particular case."

-- But that's not the point. Again, I'm trying to show readers that my challenge rules are realistic. A fair number of them would suffer from the misconception that my challenge is flawed because it requires too many hands than anyone could play in a lifetime.

Conclusion

Maybe it's best if I offer challengers a *choice* of rules. Based on what I've learned, here's what I suggest letting challengers pick from:

(1) Average profit of at least $1 and at least 25% winning sessions, averaged over 1 million sessions. Starting session bankroll of $500.
(2) Average profit of at least $1 and at least 25% winning lifetimes, averaged over 1 million lifetimes. Starting lifetime bankroll of $5000.
(3) Any profit over 100 million rounds, starting with an unlimited bankroll.

Discuss. :)

"And by the way, the reason VP isn't covered is that the whole point of this exercise is to show that BETTING SYSTEMS FOR TABLE GAMES don't work. It's the same reason the challenge doesn't try to mow your lawn or cure cancer."

OK, but are you saying you believe there's a system that can beat -EV machines, or that you're not interested in understanding what he's done because it's not a table game?

Unless you have some already, I don't see that you'll get any takers. All I see is people coming up with all kinds of ideas and angles, and people interested in stats, analysis, and optional options. Is this thread just for discussion or is anyone actually going to do this? That's why I offered the info on Singer (prior to understanding the challenge was only for table games). He's developed strategy, he's already played it many many times, and he's beaten the negative machines. Are you just looking for theoretical systems or actual ones?

Damn, weaselman beat me to the punch there,... Nice work, fellah, I didn't quite see the loop hole in time.

100,000 weekends I'd have to punch into the spread sheet I was playing with, but even with 10,000 it was only a 5% chance you could use an approach like Weaselman was trying (reducing the number of bets to try and ride the variance).

@Doc :: It's not the busts that allow the exploitation, it's the winners being one-and-done. Basically you reduce the exposure to a set of one off bets, so instead of a weekends betting, it's a single bet. It was interesting to see that even over 10,000 bets on a single number, there was a chance you could be ahead. I knind of new it from looking at Craps in the past (a player can be a head over a large number of pass line plus x2 odds over life time).

Actually, it *is* the busts that allows the exploit, not the wins. There are normally 360,000 rounds per lifetime, and normally 1000 lifetimes. When you bust out on the very first hand of a lifetime, you've just eaten up an entire lifetime -- 360,000 rounds!

When you win the lifetime is also over, but that eats up only one round, because we add the remaining 359,999 rounds for that lifetime to to additional that have to be simulated. Since this scheme has only one round per lifetime, win or lose, we're basically going to simulate another 359,999 *lifetimes*. And each of *those* lifetimes will last only one round, and the overwhelming majority of them will be eaten up by the bust on the first round.

So the bust is the key.

Quote: MichaelBluejay

" If I bust on the first bet, there is no sit out, so, no additional rounds should be added."

-- Yes, you're right. I mentioned that but then forgot to limit the number of lifetimes added. I reprogrammed it and it showed a positive EV in 9 out of 25 trials (36%). That's more than enough to exploit the 10-1 challenge payoff, so I conclude that you won my $1000 mini-challenge. I can pay you by PayPal, or I can meet you if you're in Vegas.

Yay! :)
Thanks, it won't be necessary. I am not in this for the money. I just love winning :)
This challenge fascinated me somehow with its complex relationship to gambling. It's like trading derivatives on the stock market, but reversed. When you trade stocks you accept some risks for certain chance of return. Trading options has higher risks, but promises larger returns, then you can do something like sell option futures, which is even riskier, but can make you millionaire in a month if you have enough guts and intuition.
And here we have it the other way.
You can just play roulette, which is risky, and -EV, but you can get lucky if you bet big. You can accept the challenge, where your risk is caped at $1,000, and the "house edge" is better than the roulette itself. And then there is this "challenge on the challenge", where you don't have to put up any money, and it's a sure win.

You may be on to something here ... Perhaps, if you can get the wording of your contract, and the terms of the challenge just right, you could run a perfectly legal gambling enterprise this way :) Just make up a set of rules, that can be beaten with a certain probability, say, 9%, and put up a challenge that no betting system can beat it with 10:1 odds, and enjoy $100 from each $1000 bet.
Or alternatively, make the chance of winning 10%, pay 10:1 odds, and just collect a $100 administrative fee.

Quote:

-- But that's not the point. Again, I'm trying to show readers that my challenge rules are realistic. A fair number of them would suffer from the misconception that my challenge is flawed because it requires too many hands than anyone could play in a lifetime.

I get that. My point was that, to me, it seems that Wizards rules do a better job at that. There is only one possible point of contention there - the billion - that can fairly easily be explained away (a billion is a thousand million, it is fairly realistic to play a million rounds one lifetime, so, the same "thousand lifetimes" argument can be used you are using), while your variant has a whole bunch of them - the bankroll limit ("if it was $10,000 instead of 5000, maybe then my system would work?"), the number of bets per "lifetime" ("most of the failed "lifetimes" were just a few bets short of winning big, in real life, you'd just make that extra bet per weekend and win"), the extra lifetimes for sitting out ("my betting system does not allow any additional rounds, if you lose X, you have to stop, this is called DISCIPLINE"), the contention over how many hours per weekend one plays ("my system is targeted for light gamblers, who only play 3 hours per month"), etc ...

That's very generous of you to decline the reward money. I did suspect that anyone clever enough to figure out how to win had already applied those skills to making plenty of money in life.

About the idea that the "billion hands can be easily explained away", I just have to disagree. I think many people scoff at a billion hands as being "unrealistic" no matter how good the explanation.

I appreciate all the work you have put into this. But, gulp, on this one issue I don't see why you wouldn't take up Jerry Logan on the VP challenge. If you know that the machine they are using has a negative EV, I can't imagine why you wouldn't include it in your challenge. You might have said 'no system can beat a negative EV table game', but I know your concept should apply to any type of negative EV game. After reading all of the proposed changes, I am coming to the conclusion that your original 1 billion challenge should be maintained. If someone chooses not to take you up on it because the number is too high for them, I do not think you will lose any sleep over it.

Quote: SOOPOO

But, gulp, on this one issue I don't see why you wouldn't take up Jerry Logan on the VP challenge. If you know that the machine they are using has a negative EV, I can't imagine why you wouldn't include it in your challenge.

I've already addressed this, and I don't feel like continuing to repeat myself. I don't offer the VP challenge for the same reason I never pursued being a ballerina or ever tried eating mud. It simply doesn't interest me. For the last time, the whole point of the challenge IS TO DISPROVE THAT BETTING SYSTEMS WORK ON TABLE GAMES.

If you still don't understand, sorry, I can't help you. There's no need to discuss this further.

If someone's interested, I could do the VP challenge on terms similar to the original Wizard's rules - 1 billion rounds summed up, the betting system is disclosed before signing the contract (I promise to keep it secret) etc. In addition to the description of the betting system I will need the full pay table as well as complete rules of the particular game (I am no VP pro, so something like "the usual/customary rules" does not work, you'll have to explain to me (in writing) what exactly the rules are, and that will become a part of the contract. I will also require a week to research it after I have the rules and the paytable to make sure the game is indeed negative EV, and I am not being fooled). The amount of bet, and odds is negotiable.
Of course, I cannot make any promises regarding publication of the results on either Wizard's or Mr. Bluejay's websites, but I will agree to sign any official document confirming the results of the test, and provide the source code to be used anywhere without limitation as a proof of the betting system success or for any other purpose.

Quote: weaselman

If someone's interested, I could do the VP challenge on terms similar to the original Wizard's rules - 1 billion rounds summed up, the betting system is disclosed before signing the contract (I promise to keep it secret) etc. In addition to the description of the betting system I will need the full pay table as well as complete rules of the particular game (I am no VP pro, so something like "the usual/customary rules" does not work, you'll have to explain to me (in writing) what exactly the rules are, and that will become a part of the contract. I will also require a week to research it after I have the rules and the paytable to make sure the game is indeed negative EV, and I am not being fooled). The amount of bet, and odds is negotiable.
Of course, I cannot make any promises regarding publication of the results on either Wizard's or Mr. Bluejay's websites, but I will agree to sign any official document confirming the results of the test, and provide the source code to be used anywhere without limitation as a proof of the betting system success or for any other purpose.

I guess I don't get why Mr. Bluejay hasn't an interest in a vp effort. Gambling is gambling. However, if we can ever WAKE UP Rob Singer to get on here and respond to what you just wrote, I believe something very interesting could result. Weaselman, would you mind sending him an e-mail since he may not be reading this right now? rob_singer@q.com Thanks!

Quote: JerryLogan

Weaselman, would you mind sending him an e-mail since he may not be reading this right now? rob_singer@q.com Thanks!

Yeah, I would mind actually.
I feel, that I've done enough as it is. I am not going to walk any extra miles looking for "customers".
If you would like to attract anyone's attention to my offer, you can email/call/radio them, and give them a link to this thread.

Quote: MichaelBluejay

"But more directly, how about having session bankrolls instead of just one lifetime bankroll.... Don't allow a lifetime bankroll cap - that needs to be effectively unlimited in order to do a proper test."

-- I see that this would limit the number of rounds a bettor could eat up would be limited. But in the real world nobody has an unlimited bankroll, and I'm trying to make this as "lifelike" as possible. Still, I need a solution, so let's explore this idea. How long should a session be, and what should the starting bankroll for a session be? 120 rounds and $500?

No, nobody has an unlimited bankroll, but your lifetime bankroll isn't your session bankroll. What happens when you lose the $1000 you came with? You stop gambling early *that trip*, go home, work for another two months at your job, and go back to Vegas with another $1000. Session bankroll might be fixed, but there really isn't a limit on the number of sessions you can make under the assumption of income from work. If we're talking about realism, we have to assume that the funds for gambling come from employment income, not from some fixed pot of money that can never be replenished except through winnings.

Quote: JerryLogan

I guess I don't get why Mr. Bluejay hasn't an interest in a vp effort. Gambling is gambling. However, if we can ever WAKE UP Rob Singer to get on here and respond to what you just wrote, I believe something very interesting could result. Weaselman, would you mind sending him an e-mail since he may not be reading this right now? rob_singer@q.com Thanks!

Several reasons:
1) The purpose of the challenge is to disprove the premise that varying one's wagers can overcome the house edge. You can't vary your bets on VP in the same way you can on a table;
2) VP is a highly strategic game with millions of different mid-hand possibilities, and it would be very tedious to program hand-by-hand strategies for a simulation;
3) You don't need a betting system to beat VP anyway. You just need to find the right +EV games and play optimal strategy while flat-betting max coins.

More over, Singer's method is quite clear that he does not play basic strategy at all times. I'm not sure how easy it would be to codify his system, or if it is ultimately worth the effort. But that's between him and weaselman to discuss.

Quote: MichaelBluejay

Actually, it *is* the busts that allows the exploit, not the wins. There are normally 360,000 rounds per lifetime, and normally 1000 lifetimes. When you bust out on the very first hand of a lifetime, you've just eaten up an entire lifetime -- 360,000 rounds!

You are right. The key is effectively reducing the number of time you play... funny that, isn't that what we know already :)

I just got back to my spreadsheet. Even with 100,000 weekends, there's a better than 1% chance you could be ahead with the method proposed.

Quote: MathExtremist

Several reasons:
1) The purpose of the challenge is to disprove the premise that varying one's wagers can overcome the house edge. You can't vary your bets on VP in the same way you can on a table;
2) VP is a highly strategic game with millions of different mid-hand possibilities, and it would be very tedious to program hand-by-hand strategies for a simulation;
3) You don't need a betting system to beat VP anyway. You just need to find the right +EV games and play optimal strategy while flat-betting max coins.

1. Singer's strategy uses 6 different denominations from $1 up to $100.
2. From reading what he has put together I do agree with you on that.
3. That's just it. Singer has reported to have consistently beaten the vp machines for around 10 years on -EV games with what he has developed, and his first book is a testament on why what you say about those +EV machines is not true.

Weaselman, I'll try to contact him and get him here. For the record, I have no idea what you've done here and as such, why you're balking at contacting him. Regardless, I'll do what I can.

Quote: JerryLogan

For the record, I have no idea what you've done here

By "done enough", I meant, that I have offered to do the challenge if somebody wants to make a bet. I have no interest in making any "advertising" efforts beyond what I have already done (advertised it here).

Quote: weaselman

By "done enough", I meant, that I have offered to do the challenge if somebody wants to make a bet. I have no interest in making any "advertising" efforts beyond what I have already done (advertised it here).

Thank you. BTW, has anyone offered to bet or take the challenge?

Quote: JerryLogan

Thank you. BTW, has anyone offered to bet or take the challenge?

No. And I don't think anyone ever will.
But, let's not hijack this thread. If you have more questions for me, start another one.

Quote: JerryLogan

1. Singer's strategy uses 6 different denominations from $1 up to $100.
2. From reading what he has put together I do agree with you on that.
3. That's just it. Singer has reported to have consistently beaten the vp machines for around 10 years on -EV games with what he has developed, and his first book is a testament on why what you say about those +EV machines is not true.

Very few "betting systems" rely on only 6 different wager levels between 1x and 100x. You'd never be able to play a d'Alembert or Labouchere system on a VP game unless you were at a 100-play machine, and even then it's difficult. The payback structure of VP (lots of high pays, low hit frequency) is fundamentally different than a coin-flip game, blackjack, baccarat, the pass line, red/black, etc.

Whatever you have read appears to discuss making sub-optimal plays at VP and managing bankroll to increase the chance of a short-term session goal, at the expense of overall EV. That's not a recipe for beating the VP machines. That's just shortsighted.

"Testament" aside, nothing I said about playing +EV machines is untrue. The longer you play a +EV machine (properly), the closer your results will be to the expected positive gain. 0.5% might not sound like a lot, but on $100,000 in handle it's $500. That $500 plus the comps you'd get from 100k in action is not an insignificant amount. You don't need fancy misleading systems to achieve that - just the discipline to find the right machine and play it properly.

However, suggesting that making bad plays can be part of a winning strategy for VP is somewhat like suggesting that never splitting 8s can be part of a winning strategy for blackjack. Play Mr. Singer's style if you wish, but don't be fooled into thinking it's better for your long-term bankroll growth than playing optimally.

Quote: MichaelBluejay

MathExtremist:

"I win!"

-- Well, actually Weaselman beat you to it. I mis-programmed his original idea before your modification. Once I re-ran it then he won, so he was first in line.

I'm no intellectual property expert, but after reviewing the thread, I 'd have to say that MathExtremist deserves credit (and maybe some $$S... well maybe $... well maybe just a hearty pat on the back) for coming up with the "Go Big, then Flea" strategy which Weaselman modified to come up with the mini challenge winner.

Good job MathExtremist.

Quote: Ayecarumba

I'm no intellectual property expert, but after reviewing the thread, I 'd have to say that MathExtremist deserves credit (and maybe some $$S... well maybe $... well maybe just a hearty pat on the back) for coming up with the "Go Big, then Flea" strategy which Weaselman modified to come up with the mini challenge winner.

Good job MathExtremist.

I have not seen MathExtremists' spost where he suggested this. It was actually your "magic craps" system (and Doc's response to it), that gave me the idea - I knew that the "stop-loss" thing would not work, but seeing Doc's comment made me think about the "stop-win".

Quote: weaselman

I have not seen MathExtremists' spost where he suggested this. It was actually your "magic craps" system (and Doc's response to it), that gave me the idea - I knew that the "stop-loss" thing would not work, but seeing Doc's comment made me think about the "stop-win".

Thanks, weaselman, but my "magic CRAP" system was actually a derivative of MathExtremist's "Go Big, Flea", so I guess all roads lead back to that doorstep.... I guess this is what a shared Noble Prize feels like...hehe. Anyways, good job of nailing it down for the win.

Is there any feedback on BJ's request for input on starting session bankroll ($500?). I'm not sure it makes a difference when exposed to 1M rounds, but does any amount lower than $10k/session make a difference?

Quote: weaselman

... but seeing Doc's comment made me think about the "stop-win".

Aha! So how do I claim my fair share of the "just keep the money, Bluejay" goodwill?

I'd like to suggest some additional stipulations to your challenge rules, Mr MichaelBlueJay. Given all the work and debate that the last challenge created, I suggest that anyone submitting their system must deposit a percentage of the overall pot as collateral for the work needed to analyze the betting system. If the challenger backs out after seeing initial results or decides not to proceed, the deposit is forfeited. If the challenge does proceed, then it can refunded after the final settlement. Should the challenger fail to live up to the terms of the challenge after it is agreed upon, the deposit yet again is forfeited.

Not to sound like a loan shark or anything, but all the time and effort spent analyzing a system seems like a waste if at the last second, someone decides to back out or readjust their system. If you got a fool-proof system, you should have done your homework upfront, not afterwards.

Quote: MathExtremist

Very few "betting systems" rely on only 6 different wager levels between 1x and 100x. You'd never be able to play a d'Alembert or Labouchere system on a VP game unless you were at a 100-play machine, and even then it's difficult. The payback structure of VP (lots of high pays, low hit frequency) is fundamentally different than a coin-flip game, blackjack, baccarat, the pass line, red/black, etc.

Whatever you have read appears to discuss making sub-optimal plays at VP and managing bankroll to increase the chance of a short-term session goal, at the expense of overall EV. That's not a recipe for beating the VP machines. That's just shortsighted.

"Testament" aside, nothing I said about playing +EV machines is untrue. The longer you play a +EV machine (properly), the closer your results will be to the expected positive gain. 0.5% might not sound like a lot, but on $100,000 in handle it's $500. That $500 plus the comps you'd get from 100k in action is not an insignificant amount. You don't need fancy misleading systems to achieve that - just the discipline to find the right machine and play it properly.

However, suggesting that making bad plays can be part of a winning strategy for VP is somewhat like suggesting that never splitting 8s can be part of a winning strategy for blackjack. Play Mr. Singer's style if you wish, but don't be fooled into thinking it's better for your long-term bankroll growth than playing optimally.

I understand what you say. Singer does not use cash back or comps etc. in his reported results and according to his site, he believes that's how the optimal-only players actually are able to achieve positive results, albeit by way overvaluing the slot club benefits just enough to do so. My opinion is who knows, because as a vp player I can tell you I fib quite a bit when I get home.

I also understand what you are saying about his 5% sub-optimal plays , managing bankroll etc. But what makes it interesting is what you're saying is the same as I've seen on the vp forums many times over, yet he is the only vp player who fought back at those who called him liar & fraud by proposing a very large bet that he had and can win. He actually put up a large 6-figure cash escrow at one of the LV casinos to show he was serious and it was published, but that's when the doubters walked away. If what he does is so shortsighted, how was he able to do all this? That's why I thought it would be a great opportunity for us to find out more with Mr. Bluejay's challenge.

Quote: MathExtremist

"Testament" aside, nothing I said about playing +EV machines is untrue. The longer you play a +EV machine (properly), the closer your results will be to the expected positive gain. 0.5% might not sound like a lot, but on $100,000 in handle it's $500. That $500 plus the comps you'd get from 100k in action is not an insignificant amount. You don't need fancy misleading systems to achieve that - just the discipline to find the right machine and play it properly.

However, suggesting that making bad plays can be part of a winning strategy for VP is somewhat like suggesting that never splitting 8s can be part of a winning strategy for blackjack. Play Mr. Singer's style if you wish, but don't be fooled into thinking it's better for your long-term bankroll growth than playing optimally.

Well, playing VP optimally is hard work, and requires discipline as well. Many, many recreational gamblers don't want to make the effort to learn how to play +EV. Similarly, discipline, for some people, is utterly antithetical to what they consider to be the "gambling experience". So the defense to that is often "sour grapes"--"I don't use strategy tables or learn where the best machines are because it's all a bunch of BS anyway." And of course, this mantra is sefl-reinforcing; if you don't bother to play well, you will lose, which will provide further "proof" that "you can't win".

Similarly, Singer is appealing to so many people because he makes the assertion that it is possible to construct a "winning" strategy from playing -EV games. Those who don't want to put in the effort to play winning VP are captivated by the Singer System, which tells you that you don't need to be disciplined, to work hard, or think--just play the Singer Way and the money will roll in. Obviously appealing (like many other kinds of bullshit). Who knows, Singer may even be fooling himself. Much of his writing suggests that he actually believes what he's pitching.

Just about every claim of a winning "system" for a losing EV game--from Singer to Martingale to John Patrick to 98steps--relies on the definition of "winning" as frequency of wins, not aggregate results. A system where you win 1 unit nine times out of ten and lose 1000 units the tenth time is obviously not a winning system to anyone using common sense, but it produces a NINETY PERCENT WIN RATE!!!!! for the Singers of this world.

And their counterarguments almost always include one of the following chants 1. Yer not winning eeethur, neener neener neener 2. Yer just JEALOUS 3. I have been winning for forty skillion years and HAVE NEVVUR LOSTED. The +EV players just rely on that silly ol' math, and not their egos. But I doubt that any amount of math, logic, or reaon could ever convert any of the system believers.

You're a million miles away from hitting the proverbial nail on the head about him. It looks like you've copied/pasted all that stuff from a vp forum of loser AP's.

Three things stick out like sore thumbs:

1. You failed to address how and why, if he really is "lying" as you elude to, that he put up that much cash and published such a telling bet in LV for all to see and verify. I remember reading it when it came out. It scared the hell out of me and I had no dog in that hunt.

2. In the past he's always responded to posts like your with the same sensible reply that I personally do not have the ability to analyze, but guys like you always duck when it's really key to the whole argument. You say he wins lots of little one then loses one big one that defeats the entire purpose of the system, at least that's what I think you mean. He usually comes back with the fact that no one wants to understand that there are also large winners that occur, at a rate higher than the large losers. In a system that structurally progresses, that makes sense to me.

3. If you're so sure that he's all hot air, why are you not pushing for his strategy to be challenged here or analyzed by the Wizard or some such thing like you putting up some of that teacher's salary in a bet with him? I'd certainly take some of THAT action! I know he says it's far more complicated than learning optimal play, which already is a part of what he does.

Because you're not being an unmitigated asshole, I'll dignify you with a reply. I hope you maintain this new civil tone.

First of all, I don't really care about Singer. He's just one of the the loudest of tens of thousands of people who claim to be able to beat negative expectation casino games. He claims to have had positive results. For all we know, he has. There are no doubt some career keno players out there who are ahead for their lifetimes. This, in itself, proves nothing. I play +EV video poker. I also claim to be ahead, lifetime. This, in itself, also proves nothing, and you have missed the mark in continually calling me a liar when I say it.

The reason why Singer is wrong is that negative numbers cannot be added to produce positive numbers. Every bet he makes at a Double Double Bonus machine (his game of choice, as I understand it) will cost him 1% of that bet, in the long run--a small negative number. And despite what Singer repeatedly says, the long run is the only thing that matters, IN ASSESSING THE VALUE OF A "SYSTEM". It is pointless to examine a particular set of results, his, mine, or yours. So his claims of being ahead are meaningless, whether they are true or false. You can play at a 50% disadvantage and still be ahead.

I don't know about Singer's "challenges", other than the fact that he's continually weaseled out of them when push came to shove. If you believe otherwise, so be it.

The fact that Singer occasionally books a large win does NOT negate the fact that his total losses will outweigh his total wins. He plays NEGATIVE EXPECTATION GAMES. That means, mathematically, that he will lose more than he wins. This is, sadly, inescapable. I know that you wish that this wasn't so. But it is--in the same way, and for the same reason, that a roulette player will lose more than he wins.

The difference between a losing VP player and an advantage player is simply this: the losing player plays 9/6 (or worse) DDB, with a return of 98.9%. The AP plays 10/6 DDB, which returns 100.0%, on double points days in a casino that already awards 0.25% cash back. They are playing the same game, but one stands to lose $1.10 for every $100 bet, while the other stands to win $0.50 for every dollar bet. The losing player gives one bet to the house every time he hits a full house, because he SHOULD be getting paid 10 bets, not 9. And the fact that full houses come up once every 90 hands is why the 9/6 player is losing that 1.1%.

Singer is happy being robbed of that 1 bet every 90 hands, because he mumbles and chants and convinces himself that his "system" is making up for that. It isn't.

Please don't bother to reply with the same old diatribe about how APs don't exist, everybody is "jealous" of Singer, I'm just a rotten old lonely man who has actually lost $3 billion playing video poker, and no one has the stones to take on the great Singer, who has won a million buckazooties playing a negative expectation game, winning 80 percent of the time. We've heard it all before. What refutes Singer is the MATH, and the math is what determines the outcome of casino games. It's really just that simple, and just that irrefutable.

Quote: JerryLogan

3. If you're so sure that he's all hot air, why are you not pushing for his strategy to be challenged here or analyzed by the Wizard or some such thing like you putting up some of that teacher's salary in a bet with him? I'd certainly take some of THAT action! I know he says it's far more complicated than learning optimal play, which already is a part of what he does.

You've got it backwards. You can assert that 2 - 3 = 4, but I don't need to challenge your error by running tests or analyses - you're just wrong. The same is true if you assert that your betting system based on some sequence of -EV bets results in an overall player advantage. I don't need to challenge your error by running tests or analyses - you're just wrong.

The premise behind this whole testing thread was that the system hawker, who is ignorant of this error, feels so confident in their mistake that they're willing to put their money where their mouth is. That's why the past few pages have been about designing the parameters so that the test can't be exploited.

I've never read any of this fellow's writings, and I don't plan on it. If he claims, as you say, that he plays -EV video poker at a theoretical profit by making suboptimal plays and varying his bets, well that's just silly. I don't need to test that assertion to know that it's all hot air. If *you* feel the need to test it, that's a different question. But I doubt you'll get someone to donate the programming resources to do a proper simulation. That would likely be a paid job.

Quote: MathExtremist

You've got it backwards. You can assert that 2 - 3 = 4, but I don't need to challenge your error by running tests or analyses - you're just wrong. The same is true if you assert that your betting system based on some sequence of -EV bets results in an overall player advantage. I don't need to challenge your error by running tests or analyses - you're just wrong.

The premise behind this whole testing thread was that the system hawker, who is ignorant of this error, feels so confident in their mistake that they're willing to put their money where their mouth is. That's why the past few pages have been about designing the parameters so that the test can't be exploited.

I've never read any of this fellow's writings, and I don't plan on it. If he claims, as you say, that he plays -EV video poker at a theoretical profit by making suboptimal plays and varying his bets, well that's just silly. I don't need to test that assertion to know that it's all hot air. If *you* feel the need to test it, that's a different question. But I doubt you'll get someone to donate the programming resources to do a proper simulation. That would likely be a paid job.

While agreeing with you completely, I think you shouldn't encourage the idea of a "test" proving anything one way or the other--it doesn't. If you make the test include enough trials, the -EV advocates will just bleat that "real life" doesn't involve that many outcomes--no one sits through a billion decisions at craps, etc. etc. Then you can fall back on the math, but -EVers don't think there is any such thing as math. The Singers of this world are, as you say, self-refuting, but they try to make up for that with obfuscation and sheer volume.

I cannot conceive of a true believer seeing the results of such a test and having the scales drop from his eyes, nor do I see any of the professional frauds admitting that the results of ANY test, no matter HOW you devise it, expose those frauds for what they are. No matter how you tweak the test, can you possibly imagine any believer or fraud acknowledging its validity (when it proves the system worthless), and saying, "Oh wow, I guess I was a sucker/charlatan"?

I'm just trying to reply to this post in general - not to Jerry's post above specifically if it ends up appearing that way.

I just found it interesting from, maybe, the opposite side, that such betting challenges may actually just support how likely it may be that an Average Joe might actually "win" in just his one and only lifetime.

Like, ok, I play 2 to 3 weeks per year max on vacation because I like to go to Vegas or AC or wherever. I enjoy BJ. I've used crap like Oscar's Grind, etc, on other games, mostly craps.

Sure, my expected lifetime is a "small" mathematical statisical sample. I realize that. That's my point.

Generally, I've brought "lots" of units for my week of play, spread widely but infrequently, happy to win a "big" bet.

I realize I would probably be a loser if I lived 1000 lifetimes to 70,000 years old.

But, when one has a 100-500, or more, initial min bet unit roll that succeeds in winning one unit 90+% of the time while even just flat-betting while employing a system, in a -EV game, why not use such a "losing system"? Basically rolling the dice for one's individual expected length of playing time in one's remaining lifetime?

Why not make it a real betting challenge and let the bettor specify length of time and initial unit roll and set, somewhat anyway, fair odds from there for each indiv challenge? You set the odds for each challenge and let the guy accept or not.

What is the point of making a Mr. Obvious "betting challenge" that every sane person knows is unwinnable when freed from time and bet restraints?

Unless it is to show the likelihood of how often a losing system can win in a specific length of time with a starting roll of x times min bet?

It just seems to me, while attempting to definitively prove the obvious, which needs no proving it is so obvious, you gloss over how much more likely it is to win by playing a certain "system" for a fixed length of time with a defined initial roll.

Is it not true that with a 10000 unit starting roll using Oscar's Grind in a "fair" coin-flip game I will win only 99.991% of the time?

I don't know. Using betting "systems", as it has happened, has served me pretty well over my puny wagering lifetime.

1. Being "civil" is helped a whole lot by not being vulgar at the same time.

2. I don't disagree with your second para.

3. My understanding is that Singer has developed a method that HE is able to master as a profitable endeavor. He teaches short term methodology, not a long term approach that professes to beat the casino math. You disagree with the short term but I do not, if only because that's exactly how I play in my relative few visits each year. So this long run does not matter at all to me, and maybe not to him. He gives himself more of an opportunity, 5% more if I understand him right, to win each session than the long term "experts" give themselves.

4. Your weak assertions are starting to surface again. It's not a question of me believing his word. The published challenge bet I mentioned is an actual event that was in a LV paper's front page in big bold print, and it named the casino and executive where he escrowed the CASH. Something like $300,000 or so. If that's weaseling out of a bet at "shove time" then please do tell more. And I think I've aslked you this before but to no avail. Name a time when he walked away from a bet and with whom. It can be checked out on the vp forums or in his articles I expect.

5. I get what you're saying, play a negative game and lose over time. Play a positive game over time and have a chance to win. In fact, that's what you've come up with to counter that he has more large winners than large losers but it just doesn't make any sense. Two problems with that. I lose at a clip of about 5%-10% of what I play and I know what I'm doing. If I played at a 2% advantage I'd still be a loser, like I believe all but the luckiest of players are. Also, Singer was an AP for 5 years and he could not win. That's why he changed methods. He's actually played both ways and his critics only one. Whom should we believe in such a case?

6. If the math is what determines the outcome of casino games, then we are in agreement. So please explain how all these vp AP's are allowed to keep winning & some keep on writing about where and how much they've been winning while absolutely raping the slot clubs on non-stop invites....while BJ card counters are immediately tossed.