Various opinions on blackjack

[mathare]

I've done a lot of reading and research into blackjack into the past few months. It's a game that absolutely fascinates me because of it (usually) uses dependent events. That is once a hand has been dealt the cards in that hand are no longer in the deck/shoe being used to deal the game altering the probabilities of other cards being dealt in future rounds of the game. This is the principle behind card counting.

I'm not going to go into card counting here but I may start a separate thread on it at some stage.

Suffice it to say that card counting is no longer the magic bullet it is still often advertised to be. Modern casino blackjack rules mean that card counting is now of limited use.

An increase in the number of decks in play, poor penetration (not dealing all of the cards) of around 50-60% and careful monitoring of betting spreads have meant that card counting just isn't worth it any more - in my opinion. It's no longer an accurate system - it's just an estimation based on incomplete information and staking according to the count can leave you suffering big losses on several hands.

Is it worth learning anything about counting? I believe it might be a nice party trick and a useful skill for keeping your brain active but I don't feel at this stage as though it would work in a modern casino environment.

It has long been said that casinos can and will ask anyone they suspect of counting cards to leave the premises, and they are entitled to do this. But the truth is I don't believe casinos are as scared of counters as they once were. They have changed the game sufficiently to basically nullify card counters yet people learn these skills at home, bring them to the casino and can't apply these skills under casino conditions. At home away from all the distractions they can take all the time they need. But in the casino it's a different story.

So if you plan to play blackjack and I've said card counting is pointless what should you do? You should learn about basic strategy.

Basic Strategy was developed by Julian Braun of IBM in the 1960s and gives the player the optimal way to play his hand for each possible dealer upcard. This was originally designed for the single deck game dealt, as it was at the time, from first card to last card. This is now not the game played in modern casinos so numerous variations on the basic strategy have been produced to cater for all the rule changes and variations.

You might not know the exact rules of the game before you get to the table and you can't be expected to learn all variations of the basic strategy chart so how does it help? It will teach you a number of basic principles that will help improve your game. It should stop you making random decisions each time you get a hand which in turn will reduce the house edge. Playing perfect basic strategy for an 8-deck shoe to typical Las Vegas strip rules brings the house edge down to under 0.5%. Random play gives the casino a reported edge of anywhere between 6% and 14%, possibly higher.

Because I'm odd like that I am trying to work basic strategy out from first principles (in Excel of course) with the aim of finding a single basic strategy that applies to the majority of modern blackjack games. My findings will, as always, be published here on the forum.

[andyp]

NIce thread mat, your results will make an interesting read! cheers

[Win2Win]

SInce we can no longer have or say 'Blackboard' the word 'Blackjack' is BANNED from this forum, as it relates to the black slave Jack, and we can not have that. The game will from now on be known as 'International Card Game for the male and female, as well as homosexual, transvestites, lesbians, christians, muslims, jehovah witnesses,................'

[MarcusMel]

Well lets call it Schwarzer Bube then :rolleyes:

[roma]

Playing perfect basic strategy for an 8-deck shoe to typical Las Vegas strip rules brings the house edge down to under 0.5%. Random play gives the casino a reported edge of anywhere between 6% and 14%, possibly higher.

It is an excellent way of going into the online casinos and coming out with (on average) 99.5% of the joining up bonuses that they give you after playing the required thousand or so hands to satisfy the wagering requirements :D

Of course it is very rare to ever come out with exactly 99.5%. Sometimes I have more than doubled the bonus, sometimes I have lost it completely, usually it is somewhere (anywhere) in between. In practice, losing and winning runs seem very marked, and the result of 1000 or 2000 hands is a long way off the average. I don't know enough about statistics to know if this is to be expected but I have wondered if it is a feature of the online casinos' "randomised" card generators which I have heard are not truly random but programmed to simulate randomness.

It will be interesting to see your results :)

[presto]

i have a few books on blackjack, and have practiced card counting / stake managment for hours on end.

though to be honnest the 0.5% possible edge is not great for the work involved. i also think it would turn a night at the casino into a mathmatical bore (though that is probably matt's idea of a good time :D ). and you probably wouldn't notice much difference (short term) in results form basic stratergy than counting.

i know many may have seen the documentaries on the MIT teams, but i think i heard somewhere that casinos actually make tones of cash from card counters (overall) as for every MIT maths genious - there wuld be 500+ wannabe counters who lose cash.

the only real way to profit from blackjack IMO is online casino 'scalping' like roma mentioned.

[mathare]

Any house edge works by grinding down your bankroll with each bet. The more bets you make the more chances you give the casino to take your money.

The effect of a known house edge over 10, 100, 1000 etc bets is something I will be looking into as I am very keen to know when it really starts to kick in

[mathare]

Because I'm odd like that I am trying to work basic strategy out from first principles (in Excel of course) with the aim of finding a single basic strategy that applies to the majority of modern blackjack games. My findings will, as always, be published here on the forum.Must admit I am really struggling with this now.

I have worked out the probabilities of the dealer reaching all hands in the ranger 17-21 and going bust for each possible upcard and they match those generated by a blackjack strategy calculator I downloaded but where I go from here is proving problematic.

The idea behind basic strategy is that the recommended play is the one with the best expectation. I can work out the expectation for standing on all player hands easily enough and I think I should be able to do doubling down as well. It's hitting and splitting that I am struggling with though.

I do have a way forward but it will take time. There is a book by Peter Griffen I need to get hold off. Arnold Snyder also wrote a good book on blackjack maths but unfortunately that is out of print now. I also have copies of books by gambling maths gurus such as Malmuth and Sklanksy that might help.

I have also lost sight of why I started doing this in the first place. I think the question I wanted answering is "Is online blackjack worth playing?" The way I was going to address this was not only to calculate optimal basic strategy for the most common online rule sets but to then see the effect of the calculated house edge on a large number of bets for a large number of gamblers to see what the chance of profiting was. I had also planned to use the same set of rules without the shuffle after each hand to see if the rules could be used to make a profit live where a dealer would shuffle only when the cut card was reached.

There are a number of non-trivial challenges in doing all of this and I don't expect to make much in the way of progress for a while yet, not least till I can understand the maths behind calculating the expectation for hitting and splitting hands.

[presto]

one factor that i would be wary of when using XL over real play would be, the dificulties in conting.

XL never forgets the count, nor miscalculates - but in real play you would have to consider the human error factor.

i have a book by 'stanford wong' who came as a reccomended author (havnt read it yet) - but he has a site 'bj21' that is supposed to be a good resource.

[mathare]

Snyder's site seems good too.

One of the things I want to test with Excel is the idea of the count. Suppose you can keep the count 100% accurately (which as you say Excel will do) - how much does that help?

I then want to try and build in a human factor. Say you drop the count once every X shoes and have to revert to either 0 or not counting for the rest of the shoe - what does that do? What about if you're only 98% accurate or 95% or any other figure you care to choose? How accurate with the count must you be before it starts to actually cost you?

[thai-swe]

Hasnt some casinos aouto shuffel after every round?

[mathare]

Some real life casinos do use auto-shufflers and randomisers, yeah. The conventional wisdom seems to be avoid tables using these devices like the plague, which kinda implies also avoiding online blackjack.

However, I don't think that the continuous auto-shufflers used in some casinos (I know the Bellagio has some) can give a truly random shuffle. The dealers normally deal a few hands and when the discards get to about half a deck or greater they will bung them back into the auto-shuffler and deal the next hand. In a truly random shuffle the next card out is equally likely to one of the cards recently dumped back into the machine as one that was already in there, but this can't be the case if the next cards are already available for dealing. So one thing I will be looking at (hopefully) is shuffles that are only a certain percentage random, say 90% random, and seeing what effect that has.

Also no human shuffle will be truly 100% random. Cards will clump and cluster. Some casinos have games with twice as many decks as normal along with a shuffle machine. While the dealer is dealing from one shoe the machine is shuffling and randomising an equivalent number of decks. At the end of the round where the cut card shows the dealer swaps the discards and remainder of the deck for the cards that the machine has patiently been shuffling and the game flows faster with a more random shuffle.

[wolves pete]

Have you got any further with your quest Matt?

[mathare]

Have you got any further with your quest Matt?Not as such. I have been working on a few other projects. I have a couple of blackjack studies in the pipeline but the shuffle business may be beyond me at the minute.