Basic Strategy flawed for tournaments?

As this is my first post - I'll give some background -

I have played in three tournaments, getting second in one - failing to advance in the others -

I have a strong math background - including an undergraduate major and graduate level courses - specializing in methodolgy, probability, statistics, and such -

Am an ok Blackjack player, use basic strategy and can use a simple high/low card count -

now - my issue -

basic strategy was developed using computer simulations of millions of blackjack hands - it is optimized to produce maximum expected profit over a very long term series of hands - say - 1,000,000 -

to do this - Basic Strategy sometimes sacrifices the expectation of WINNING a hand to maximise expected PROFIT - this would be in some double down and splitting situations - such as doubling a count of 11 against a dealer 10 - doubling down INCREASES your expected profit but DECREASES your expectation of winning the hand-

in tournament play - of course - one does not play 1,000,000 hands per round - more like 20 - and the goal is NOT to maximize profit over the long run - but to qualify to advance to the next round - so - basic strategy is not designed for this situation - the goal in tournament play would be better stated as optimizing your play to produce the greatest number of 20 hand series with a profit - i.e. - doing well in the most series - than maximizing long term profit -

I expect a strategy geared to winning the most hands and/or minimizing the number of losing hands - would prove superior in tournament play - compared to basic strategy maximizing long term profit-

given the coments I have seen in the strategy forum - most players recognize that you need to vary from basic strategy on certain hands - but has anyone looked into the fact that basic strategy is likely not an optimal strategy for tournament blackjack on individual hand play - where double down and split decisions are made? Has anyone tried to crunch numbers to develop a strategy optimized for tournament play? - Any interest in doing this?

 

RKczek

U hit upon the paradox. Given u're math background, u're certainly more grounded in u're analysis. I would purport a study of Ken Smith's archived articles would enlighten u some, re tourney play. As far as numbers, the one and only S. Yama would test and tempt u're discipline.

More to an esoteric player. Numbers are the foundation. The basic. extrapolate from there, by playing. Yama has stated. if I'm not mistaken,(but could be) early rounds of tournament are best played, basic strategy. U're notion of maximizing wins and minumizing losses, I find accurate. But what 45% of hands out of the 20 do you win.

Most experienced,(not I) playas say stay close to the end, to make that play that is needed. How do u know? U play. Outside of numbers, it comes down to what do I have to do to win or place. A statistical analysis does no good, if by doing it, leaves u out of a shot at the win, or place, or show. That's why at the final two-three hands u see some of the most nonsensical moves imaginable. i'm sure as always someone else will give u a more clear, more easily understood answer, but that's my two cents.

 

reply to noman

the question is how do you maximize the odds of getting to the last few hands in a position to win the table - or qualify - ?

if you use basic strategy in the early hands of a round - you are not maximizing your chances of winning each hand - and winning each hand - or avoiding losses - is what will get you there - it doesn't matter what your long term profit potential is - if you lose too many hands to be in position to win at the end of the round -

I would suggest a real easy change to basic strategy for tournament play - never double down on a nine or lower - never double down against a dealer 7,8,9,10,A - unless you need to exceed a max bet limit - this would give you the best odds of winning each hand - better than basic strategy - where this double down situation comes up - and isn't that what you want in tournament play - to maximize your chance of winning on each hand - ??? - in your playing strategy - ??

likewise - changing basic strategy for splitting may let you reduce odds of losing money on certain hands - when not losing is a good option (say you see the chip leader bust out on a big bet - if you can avaoid losing your bet - you can benefit - so splitting may help you reduce the odds of losing - even though basic strategy says this is not a good play -

bottom line is basic strategy is not about winning and losing hands - it is about maximizing long term profit - in tournaments - shouldn't you use a strategy that maximizes your odds of winning each hand?? -

 

Tourney Strategy

Kuzeck (sp) excuse me.

One must accept most of u're points.

However, in most of the accepted, currently favorable top "bibles" on how to play, the format is to get the most money on the table.

U're first post suggested the short term improbabilities, against the gabazillion plays. True, if one plays every waking hour for eternity, then by all means play the true numbers and look to any of five books to see what the true numbers are.

In the short term, the Ken Smith, among others, theory of correlating your bet, to the other players, really does give u the best "chance."

While one has to gauge their own personal true, or preceived outcomes over time, the correlation theory is, that on any particular hand all the players will have the same outcome, win or lose against the dealer.

If u want to fight the deviation, abnormality, variance, "run of bad luck" then by all means, just play the statistical edge.

To be real, in tourneys, that short game,(my belief only) makes the .004 advantage of one play over another meaningless, since it is based on the gabazillion simulations.

Just because it's fifty-fifty that a coin will come up heads or tails, doesn't mean it will come up heads fifty percent of the time out of twenty, or twenty-five trials.

Again, one can be armed with the best information for a particular play, but observing what u're opponents hold, what the dealer holds, what your chip difference is and what possible dealer outcomes are, or could be, matter more than strict statistics.

Now I'll iterate for you, as many on this site already know, I'm not the sharpest pencil nor theoritcal mathematician, but if there were true certainty, thru numbers applied to black jack, live table play or tournaments, one special genius would be winning everything.

Books have been written about those or groups of those who have come close, but........

 

Basic Startegy Flawed PERIOD

UNHH!!RKuczek brave, you very close to learning that no mathmetician develpoed the BS; nor did they find the flaw.:flame: At least two authors on math
one non math ended up with near same BS. Knowledge of the flaw and how it works against the players, varifies those modified basic strategies. Disclosier
of same will damage players even more than the "so called gurus" did.:flame:

 

Tools are needed

TOOLS! We like TOOLS.

 

Reply2 from RKuczek

I agree that you play your hand to do what you need to do to win - and that you're playing against the players - I learned that in the first round of the first tournament I played in - I started it thinking that I should play good basic strategy - use a card count - and let my good play win the table - the results were predictably pathetic - I was a distant fourth - and only because two players bombed out with suicide bets where they went all in and lost - I finished that round thinking - "it's all about money management" - rebought - and finished second in the tournament -

but no matter what you bet - or what your goal on that hand is - you still need to make good playing decisions - especially where doubling and splitting is concerned - example -

you are playing in a semi-final round - where only one player advances to the final table - Blackjack pays 2-1, and you can draw only one card to split Aces - player's cards are dealt face up - it is the final hand and there are you and one other player left - everyone else has bombed out - you have $280 in chips, the other player has $150 -

Now - the situation - the other player bets $150 - everything he has - you bet $140 - you get a pair of Aces - the Dealer has an Ace showing - the other player Blackjacks - he wins $300 - and now has finished with $450 -

you are holding $140 - and have $140 bet - and need to win $170 to win the table and advance - what do you do??

Basic strategy says 'never double on soft 12' and 'always split Aces' - if you split - you have a chance to win both resulting hands - and win the table - but that is very much the wrong play - here are the numbers:

Splitting a pair of Aces against a Dealer Ace: Outcomes:

Win both hands (+$280,win the table): 24.71%
Win one hand, lose the other ($0, lose the table): 57.16%
Lose both hands (-$280, lose the table): 18.13%

Doubling a pair of Aces against a Dealer Ace: Outcomes:

Win the doubled bet (+$280,win the table): 34.59%
Push the doubled bet ($0, lose the table): 4.69%
Lose the doubled bet (-$280, lose the table): 60.71%

Do you want a 35% of winning the table? or a 25% chance?

Right play is clearly to double down in this situation - it is your Maximum Winning Probablity Strategy play - Splitting is a Minimum Loss Strategy/Maiximum Profit Strategy play - a big difference in the probablities - of winning a particular hand -

Given that basic strategy maximizes profit over a very long series of hands - it is very unlikely to be the optimal strategy for maximizing the number of 20-hand series you make a profit on - which is what you want in tournaments - and basic strategy can be a very bad strategy for playing individual hands -

I think tournament blackjack calls for a player to have four base strategies - a maximum win strategy - optimized to maximize the probablity of winning each hand - a minimum loss strategy - minimizing the probablity of losing each hand - and a prefered decision strategy - which gives you the best play for splitting or doubling when you need to get more money on the table - and - a fourth strategy - which combines the max win and min loss approaches to give you the greatest differential between hands won and lost in a short series of hands - the player needs to switch between strategies as needed -

 

Make the Easier Play

Your bet should have been 175 and not 140.If you made the right bet your playing the hand becomes easy.Just win your hand and no DD is needed.

 

Truer words were never spoken

Truer words were never spoken

Chips

 

RKuczek

By god, I think you've got it.

Whose basic strategy?
What's the right bet? When?
On and on.
You'll only know, what u need, what u want after playing and reevaluating u're play after, win or loss.


Sorry, all!

But, if a golf shot, by most pros standards calls for a seven iron fade and the caddy also recommends it, but your seven iron play sucks, you go to your six runner or eight high and hard.

If u're backhand in tennis sucks, u go to position urself to give urself the best opportunity for a forehand return. If u can't get there or miss it, for u it was still as good or better a choice, than ur lousey backhand attempt.

U use different bait, different depth, different location indicators for a bass tournament than u do for trout. And just because everyone else says a double hitched, swirly fly is best,(most of the time) for stream trout fishing, doesn't mean you can't use a matza ball with oregano if u so choose.

 

Well done Noman

Thinking outside the bun...er matza.

One needs to go with their strenghts.

 

Split those ACEs

RKuczek:

First of all, I agree with Joep that the right bet in this situation would have eliminated the need to double your bet. That being said, let’s look at the situation as you presented the bets and cards.

I tried, but I don’t understand how you came up with the 24.71% and 34.59%. I have calculated quite different numbers.

My calculations are based on a 6 deck shoe, dealer hits a soft 17.
With the dealer up card as an Ace, after the dealer checks for blackjack, the dealer probabilities (based on “wizard of odds”) are:
Hit to 17 = .082842
Hit to 18 = .206593
Hit to 19 = .206757
Hit to 20 = .207262
Hit to 21 = .095264
Bust .....= .201281

Splitting ACEs - the probability of winning both hands:
Dealer hits to 17: 7/13 * 7/13 * .082842 = .024019
Dealer hits to 18: 6/13 * 6/13 * .206593 = .044008
Dealer hits to 19: 5/13 * 5/13 * .206757 = .030585
Dealer hits to 20: 4/13 * 4/13 * .207262 = .019622
Dealer hits to 21: 0/13 * 0/13 * .095264 = .000000
Dealer bust:.....13/13 * 13/13 * .201281 = .201281
Total:................................................ .319515 or 31.95% (not the 24.71% you show)

Doubling a pair of ACSs - the probability of winning that double-down:
Dealer hits to 17: 4/13 * .082842 = .006372
Dealer hits to 18: 3/13 * .206593 = .047675
Dealer hits to 19: 2/13 * .206757 = .031808
Dealer hits to 20: 1/13 * .207262 = .015943
Dealer hits to 21: 0/13 * .095264 = .000000
Dealer bust:..... 13/13 * .201281 = .201281
Total: .................................... .303079 or 30.31% (not the 34.59% you show)

My conclusion: Splitting ACEs against a dealer Ace (as opposed to doubling the soft 12) is the right play. It only gets you an additional 1.64% but it is the better choice. It confirms that basic strategy is correct for this situation and Wong’s statement that when you need to double your bet, always split the ACEs.

Now, I’m not a mathematical genius. If you or someone else can show me an error in my calculations or thought process, I’m listing.

 

Reply from RKuczek

Joep,CHips -

you're right about the bet size - but then the example wouldn't have worked as well - the point I have been trying to make - is that after you place your bet - which is 'betting strategy' and I think all the comments I have seen on this web site about sizing bets - etc - have been very good - which is why I am hanging out here - you still need to win your bet - that's your 'playing strategy' - and that's were knowing the real odds for each possible play comes in - in the example I gave - the best 'play' - would be to HIT the pair of Aces - as that's your best chance to win the hand - if you need to get extra money on the table - then your best play is to double down - it would be possible to work out an optimal set of strategies for tournament Blackjack - giving the best play for each situation - that comes up after you employ your betting strategy - no matter how well you bet - you still need to win the hand -

I was just pointing out - that without a set of 'tournament strategies' - you either end up playing basic strategy or guessing - and guessing isn't good - and basic strategy is not designed for tournament play - in a tournament - even in the early hands - as per a comment made in a message from mankeysystem - you do not want to maximize the EV of the hand - that works only if you are playing a very long series of hands - what you want is a strategy - for playing the hand - which maximizes the number of hands you win versus the number you lose, in a short series of hands - - basic strategy is actually pretty poor at that

 

Reply to Toolman

Oh Yuk - I did screw up my probability calculations - I treated the split hands as if they were independent plays - instead of both playing against the same dealer hand -- I will look over the calcs again - but I think yours are in the ball park -

Ken Smith - thanks for the comment on must win strategy - I think when it comes to hit/stand decisions - basic strategy is pretty accurate - as it all comes down to win or lose - but when it comes to split versus double down - and when splitting or doubling down may provide an opportunity to swing the chip count your way - that is where I question the appropriateness of basic strategy - where it advises on splitting or doubling - also - still disagree that maximizing EV is your goal - your goal is to win more chips than your opponents - more often - and I don't think maximizing EV necessarily does that - I am looking for some good blackjack simulation software - and when I get something useful - or can program some - I will run off different strategies and let everyone know if there is anything to what I am saying - maybe - maybe not -