In Hiding: Response to "OK, one more step" post.

IH: I�ve delayed responding to your post in order to allow time for Roger Harris to respond to a suggestion I made to him in another thread on this site. I asked why the computer-generated results of his software program differed from those previously posted by Norm Wattenberger. He stated that the game rules were different, and that this accounted for the difference in results. Since the difference in results was (IMHO) more severe than I thought it should be, I suggested that he make minor alterations in his program so that he was playing by the same rules as Norm�s program, and then compare the two sets of results to see if they were identical. A week or so has passed without a response to my suggestion.
I�m famous for my lack of skill at designing or interpreting computer-generated data, but have had some experience using two different programs designed by �unbiased� programmers. Both programs calculate total amount wagered, allow for the establishment of a progressive betting system, and generate data that supports my claims comparing flat and progressive betting. Both Roger and Norm claim that these programs are �flawed�. There�s been no serious effort by �expert� blackjack software programmers because they say, �Why bother? Progressions don�t work.� And due to my lack of computer skills, it�s been impossible to implement programs designed by Norm and Roger so that I can personally verify their predicted results.
So much for background information...
Incidentally, I�ve never claimed that my progressive system will result in long-term (billion hands) profit, but through the analysis of dozens of short term computer sims and manually-dealt sims, I�m convinced that my progressive bettor loses less than a flat bettor, and casino comps added to a lower loss rate can make the game a positive outcome experience.
Opinions vary among the �experts.� Wong states at the top of the Progressions Page: �Progressions can be fun, but they don�t change the casino�s advantage.� Roger and Norm claim that my progression causes a player to lose MORE than a flat bettor. I claim the opposite... and so it goes.
Also, I�m not totally convinced that any computer simulation can accurately predict the outcome of actual casino play, essentially because I don�t see any way that one can be positive that random number generated shuffles are the same as dealer or shuffle machine shuffles. I make this statement after reviewing the results of tens of thousand of manually-dealt hands, and seeing the drastic difference in data generated for manually-dealt play for flat and progressive betting.
Perhaps this fact is what most upsets several other posters on this site.
I certainly don�t enjoy not being �blessed� by the computer guys, but there�s little I can do about it at this point. They throw out my manually-dealt results because the samples are too small to be relevant, they reject my sim results because they say the programs I�m using are flawed, and they don�t make their sim programs available to me in a way that I can reproduce their results (nor would I have any way of knowing if their programs are �flawed�).
I hope I�ve answered some of your questions.
Cheers!
Walter

If you weren't so adamant in your conviction, I'd probably not bother with all this. Obviously, folks a lot more mathematically savvy than I, are convinced this is all a waste of time. And I'm certainly not a math wizard nor an important poster on this site. (Since this seems to have become personal, maybe I should have continued this by email. But I'll make this one last post on the matter.)

I don't know much about sims either. But I'm smart enough to know that if two different sims are being compared, the comparison doesn't mean much unless the same game rules are programmed. If you use two different strategies for the SAME GAME (with accurate programming), then you can arrive at a fair comparison. If you program games with even slightly different rules using the same strategy, I'd be suspect if you didn't get two different results. A lot of the guys on this site use similar commercially available software and occasionally compare results. If two different people arrive at different results, someone did something wrong.

I got into BJ in the early 80's. For twenty years I've been told that the experts have done billions of sims with progressive strategies and have proven they don't work. Conversely, just over the past several weeks or so, I've seen it said that progressive strategies "don't increase the EV of the game". That phrase would lead me to believe that some progressive strategies might be on a par with good old fashioned flat-betting Basic Strategy!? But I've asked the question and no one seems willing to clarify the point. Maybe that's too great an admission.

Or maybe, the big boys have been over this so many times, they're tired of it and don't care to talk about it anymore. So we just have to settle for that line of text that appears in all of "the books" that tells us progression methods don't work.

So, is the truth that the best(?) progressive strategies do no harm or good to the average player? That they just "don't improve the EV of the game" as counting does? And then one might be led to believe that if the average disadvantage to the BS player is "only" around .5%, could a strategy be developed which, with just a little finagelling, might overcome that small disadvantage to allow a player to sort of play even?

All of this is really redundant for me in that I'm not about to start progressive betting anytime soon.

It's only of interest to me because (I know I'm repeating myself), for the past five years, a group of us have been using a classic counting strategy that includes a few quit-points and a modest stop/loss. The reason I'm not so quick to just disregard methods because they flow against convention is because I know there are those on this board who immediately call any method that includes these elements as superstition and voodoo. All six members of our group, however, have had five consecutive winning years by adopting these few changes into an otherwise classic counting game. Among us we've played over 1.5M hands in casinos and probably over 2.5M if you include practice sessions under casino conditions. And when we have a losing session, it doesn't wipe us out as I've recently read has happened to a few well known players who frequent this site. Nor, BTW, does it cripple us as would a losing session with an agressive progressive strategy.

I personally play 2 to 4 days a week. So ask me if I'll stop using the strategy because a poster on a web site tells me it's nonsense.

To end this, I can only suggest you need to find a way to do some accurate simulations of your method. If you're convinced your strategy is viable, you might have to pay someone to do it.

In the meantime, I'm outa here.

Good Fortune with the adventure.

I think that many players who have consistently been playing blackjack for over 10 years or so are less likely to be critical of progressions (if only for "cover"), quit points, and money management. It's the purists and the theorists who haven't had many years of playing experience that tend to be more convinced that traditional theories are the only way to go...
I've enjoyed our conversation. Don't hesitate to rejoin the debate!

Here's a serious explanation of what, exactly progressions do and don't, from someone who actually uses progressions in certain gambling situations.

Progressions..
DON'T change the EXPECTED VALUE of the game you play. If you put $N dollars of action on roulette with it's 5.26% disadvantage, your expected result is -$(N * 0.0526).

Progressions..
DO change the likelyhood of being AHEAD or BEHIND after a short period of time. You can have a progression system that gives you a 99% chance of being ahead after one hour.. of course, if you hit that 1%, expect a huge loss. Adventurous ones can use a different type of progression, one that gives them a 1% chance of having a huge win or a 99% chance of having a moderate loss.

Progressions..
DON'T change the inherent odds in the game - if the roulette wheel gives 53 straight reds, the odds of #54 being red are 18/38. No more, no less.

Progressions..
DO change the inherent odds in the "meta-game". If the object of a "session" is to win X dollars before losing your session bankroll, you can fine-tune any progression to give you from 1 to 99% chance to win $X. Of course, the bankroll you need to "win" the meta-game (IE to meet your session goal) will vary wildly, and of course, the odds of losing versus the chance of winning will always yield an expected result dead-on with the theoretical disadvantage in the house game (IE if you have a $11,000 bankroll, you can design a progression to have a 99% chance of winning $100, but your 1% will lose your $11,000, giving you an average result of -$1,100, which when divided by the average number of trials, will be almost identical to the disadvantage of each hand)

This last point bears repeating. You can NEVER take a negative expectation game, and design a progression that will give you a positive expectation.

M.

since it's here, I disagree on several points:
1. I don't believe that any progression will work with a game of "independent trials" such a roulette -- especially when the player starts with a 5.26% disadvantage -- but this is a BLACKJACK page -- a different game than roulette, and I don't think that the same principles apply.
2. Games that give you a 99% chance of winning and a 1% chance of losing it all back are based on NEGATIVE progressions, and the only reason they don't work is because table limits can dissalow the continuation of a negative progression. A positive progression player is MUCH MORE LIKELY to not suffer the same results as a negative progression player, because his bet increases seldom reach house limits -- and when they do, he's at the TOP bet in his progression, and can only lose what he won on the previous hand.
3. The "theoretical disadvantage" faced by positive progression blackjack players will (IMHO), depending on the nature of the progression, beat the anticipated long-term loss experienced by a flat bettor... he may not be a lifetime winner, but he'll lose less than a flat bettor.
I agree that you can't take a negative expectation game like roulette, with it's 5.26% house advantage, and turn it into a positive expectation game -- but (IMHO) it MIGHT be possible to play a negative expectation game with a .05% house advantage and reduce the house edge, depending upon the system employed. Just my opinion...

Walter,

Regarding your beefs with my post:

I don't believe that any progression will work with a game of "independent trials" such a roulette -- especially when the player starts with a 5.26% disadvantage -- but this is a BLACKJACK page -- a different game than roulette, and I don't think that the same principles apply.

Here the problem is what we define as a progression "Working". When asked to do what a progession CAN do (as I defined in my post), a progression works for a positive, negative or balanced game. When I use a progression in a blackjack tournament where blackjacks pay 2-1, I have about a 2% edge, yet I use a progression to give myself as much a chance of being up money in a short period of time as possible. The progression works in that regard, but it does nothing to change the odds of the game! If the blackjack tournament was a single deck game, shuffled after every hand, I would still use the progression and it would still "work" for what it is capable of doing.

It doesn't matter what the game is - blackjack, baccarat, mah-jongg or pari-mutual Jai-Alai; Progressions do NOT change the expected value of your wagering. Dependant or independant trials. Regardless of what the game is, the expected result of your however-many-trials is equal to the sum of the expected value of each wager. However, with expected result comes variance. Craps has very high variance - Pai Gow has very low variance (lots of pushes, only even money wins and losses). Variance is what you can affect with a progression. Therefore, the more you play a given game, the more the effects of your progression fade away as your measured results get closer and closer to your expected results. If you can refute this, you have a Nobel Prize on your hands for breaking the Law of Large Trials.

Games that give you a 99% chance of winning and a 1% chance of losing it all back are based on NEGATIVE progressions, and the only reason they don't work is because table limits can dissalow the continuation of a negative progression. A positive progression player is MUCH MORE LIKELY to not suffer the same results as a negative progression player, because his bet increases seldom reach house limits -- and when they do, he's at the TOP bet in his progression, and can only lose what he won on the previous hand.

A table limit does not enter into the equation I described. The progression I described is simply stated in terms of "winning the session" (winning $X) vs "losing the session" (losing your session bankroll). Whether you want a 1, 2, 5 or 99% chance of winning the session, you can design a progression to do just that. Even with no table limits, a Martingale will not give the player a positive expected value - for that (mathematically anamolous case) you need an infinite BANKROLL.

The "theoretical disadvantage" faced by positive progression blackjack players will (IMHO), depending on the nature of the progression, beat the anticipated long-term loss experienced by a flat bettor... he may not be a lifetime winner, but he'll lose less than a flat bettor

The only way in which any sort of progression might offer any sort of improvement on the expected value compared to a flat bettor:

1> If there is a correlation between previous trials and future trials (as exists in blackjack, as does not exist in roulette), and

2> If there is a correlation between the progression and whatever factors affect the change in expected value in 1> above.

For instance, the only way a NEGATIVE progression in blackjack (not positive, as you suggested) offers any manner of improvement over flat-betting is because the expected value change between rounds in blackjack is correlated to the effect of removing cards from a finite shoe and because the correlation between winning/losing hands and their likelyhood of having cards which will change the expectancy between hands.

If you lost a hand of blackjack, with all other things being equal, your EV on the next hand goes up microscopically. If you won a hand of blackjack, with all other things being equal, your EV on the next hand goes down microscopically.

To give you an idea of the scope here, a 1D game with good rules (S17, no DAS), if you use the following negative progression:

1-1-1-1-48

where you reset the progression if you win a hand OR they shuffle the deck, and they deal 5 rounds to you before shuffling, you will BREAK EVEN. And that's a 48-1 spread on single deck, a spread far more than big enough for card counting to make money on just about any game under the sun.

M.