Wizard is wrong on his Martingale analysis

Quote: Wizard

If you have a positive EV game, then a betting system couldn't lose over the long run.

The problem is, the original poster is confusing a positive EV game with a positive EV event.

He has taken a positive expectation event, (the occasional double down), applied it to a session running on a negative expectation game, and compounded it with a Martingale.

Quote: Wizard

Not to say this is the case with the original poster. If you have a positive EV game, then a betting system couldn't lose over the long run.

If I can assume a finite bankroll, I'll challenge this statement:-)

I disagree with you -- it is too simplistic to say that any overall house advantage would automatically mean that a betting system will lose, in a game with occasional variable odds. Remember, with Martingale you can lose 80% or more of your overall bets and still come out a winner. You just don't want negative streaks. You calculate the probability of a long negative streak (which will kill Martingale) based on the overall house advantage. However, that is not correct in blackjack where occasionally bets are made with better odds. Perhaps if the house advantage is big enough, you may still come out a loser, but you would rather have a game with variable odds than with consistent odds to break streaks. Lets say that you are in the 10-20-40-80-160 progression, the ability to place another 160 down with better than 50% odds (if you get 11 against a 6, for example) helps to break streaks. You can say, well that is factored into the overall odds of a 51-49 house advantage with perfect basic strategy, but I would say, that is correct when your bets are even each time, but not correct with a progressive betting system where you don't care how many wins you get you just care about not having long losing streaks.

Overall, practically you might hit the table limit or run out of money before you get enough "favorable" bets to make a difference, but theorectically I would like to see the math in a situation where there are variable odds (even occasionally) not consistent odds in a progressive betting system. That is my point.

I play Martingale because I walk into a casino about once a month with $1000 and don't care about losing it, and don't care about walking out with $700 or $1200 by playing straight bets, but care about occasionally hitting it big. Yes, I have busted many times with my $1000 after being up and then crashing. Then I walk out. But last month, I walked out with exactly $25,000 from my $1000 using Martingale. Many times, $3000 - $7000. Over time, if I played once a month for 10 years with my $1000, then added up all my winnings and losses, I would probably come out behind, unless you count. But it is a lot more fun -- playing straight Martingale 90% of the time you are up at some point (unless you hit a bad losing streak right at the beginning) and that is what is fun. According to you, no matter what in the long run, you will end up a loser with gambling. But would you want to win or lose $200-300 each time or lose your grand 75% of the time but almost always be up (before you crash, but that only takes 5 minutes) with a lot of chips on the table and occasionally walk out with LOTS of money? That is what gambling is about, entertainment. This is a lot more fun. And not as doomsday as you state -- for the reasons above.

By the way, I count too and quickly adjust my Martingale (keep high bets out) when the count is favorable and stop at only 3 progressions when the count sucks. The dealers can't figure out I am doing it, as I am always adjusting my bets and with a busy table they can't remember what progression I am on. That is probably the best way to play. Maybe that is why I am ahead, as I can get to 50% odds or perhaps more by counting...

Quote: andraaklly

However, that is not correct in blackjack where occasionally bets are made with better odds.

By the way, I count too and quickly adjust my Martingale (keep high bets out) when the count is favorable and stop at only 3 progressions when the count sucks.

1. You don't know in advance when the bets with better odds are coming (or when the bets with worse odds are coming), so the martingale does not help you.

2. Straight counting is better than the counting + martingale, because you'll have less money out in poor counts.

Just last night I lost 19 straight hands with an advantage while playing blackjack. I wong in at a true count of +4 in an 8 deck shoe, play 9 hands before it ends and lose all of them. The next shoe that I walk up to and backcount skyrockets to an amazing true of +7 and I lose 10 more straight when I jumped in. Not to mention there were 3 doubles and one pair split. Lost all of them.

You are right that you don't know in advance, but it doesn't matter. Mathematically I believe (correct me if I am wrong) that with the same total house advantage of lets say 51-49, you would rather not have each bet be 51-49 like Roulette when betting progressively, but have some bets be 60-40 in your favor and some bets be 40-60 not in your favor or whatever inbetween (averaging to 51-49 house advantage). You can't beat the house advantage outright, so with straight betting you will always lose in end, but you are trying with betting systems to beat streaks, where you can win money even if you overall lose more times then you win, because you are winning the bigger bets. That, perhaps, can be done in a game with variable odds and progressive betting? I just want to see the math, but it is very complicated... that is why blackjack is a facinating game.

I think one forgets here as well, as the count rises, the player actually LOSES more hands than normal, doubles become rarer, and ties become more of a factor. This has been addressed over on bj21.com as well

Quote: andraaklly

That, perhaps, can be done in a game with variable odds and progressive betting?

As long as the mean odds are in the house's favor, you lose in the long run. The variability does not help you. Among other things, the way you can tell that's true is that everyone is not betting like you do. If it really worked, people would do it...

Quote: andraaklly

You are right that you don't know in advance, but it doesn't matter. Mathematically I believe (correct me if I am wrong) that with the same total house advantage of lets say 51-49, you would rather not have each bet be 51-49 like Roulette when betting progressively, but have some bets be 60-40 in your favor and some bets be 40-60 not in your favor or whatever inbetween (averaging to 51-49 house advantage). You can't beat the house advantage outright, so with straight betting you will always lose in end, but you are trying with betting systems to beat streaks, where you can win money even if you overall lose more times then you win, because you are winning the bigger bets. That, perhaps, can be done in a game with variable odds and progressive betting? I just want to see the math, but it is very complicated... that is why blackjack is a facinating game.

You don't know when you sit down if the next bet will be a 60/40 or 40/60 hand. You don't know if the advantage split will happen when you bet $5 or $600. It all wraps up after doing the mathematics the hard way that you'll lose about 0.5% of the total amount staked.

If you count, your better to play other systems than a martingale (progressive/parlay). If you can easily calculate the edge, you are best using the Kelly Criterion (proven to the highest rate of bank roll growth you can get).

Even with the 60/40 game, a Martingale is not the best method. A Martingale would work if there was a game with very dependent trials, but as it doesn't exist in the gambling world, it's not really that interesting in terms of this board.

You are coming from a position of positive experience rather than a mathematical, logical basis. That's fine, and it's a common mistake. The Martingale is not the most horrible thing in the world. You can win for long periods of time, and it's exciting. But in the long run you will lose slightly more than straight betting, and your variance will be higher. That's an established facts. I agree, blackjack is a fascinating game, and I'm sure there are people who are ahead over a long period of time even without counting because of where they fall on the expectation curve. (I'm down over the long run).
------------------------------
If you read the first chapter of Richard Hoffer's Jackpot Nation you will find he continuously won over a period of a couple years just Martingaling and straight betting. He paid for his children's educations and his mortgage. Then he had an absolutely catastrophic loss and went six figures in the hole. So yeah, it can sometimes take place over a long period of time.