Mathematics in today's blackjack games

Status
Not open for further replies.
#21
kewljason said:
It seems to me, the advanced mathematics angle has really been pushed on this site lately. Maybe it is my lack of a mathematics degree and/or background that prevents me from not appreciating this angle as much as others.

Someone visiting this site for the first time and reading through some of these threads, will conclude that to succeed in card counting today, you must play an advanced level 2, 3 or higher count, with multiple side counts, and hundreds of strategy change indices for both the primary count and side counts and that is just not so. I actually believe that in today's game the mathematics is less important than it used to be. :eek: The math guys want to run sims showing that a multiple level count will improve your results by .xx percent or two side counts will with hundreds of play variations is worth an additional .xx percent. Some of these guys, I think the majority of their blackjack experience takes place on their computer. They think Qfit's products or even their own programs are their blackjack career instead of a tool to aide in their actual play. :sad:

I am not disputing that these advanced card counting ideas don't improve performance. I do believe the value of this 'gain' is overstated, the error rate understated and that it just doesn't translate fully in actual play. I am questioning the value for the majority of players in the real world of play. In the real world there is so much education that must take place on the ground, rather than the computer screen that seems to be overlooked. Things that can result in far greater improvement to your results than higher level counts and side counts. Sniffing out and taking advantage of better games and opportunities. Learning what is and isn't tolerated at different places and tayloring your play to these guidelines and levels. Even learning these tolerance levels for different pit personnel and co-ordinating with their schedules. ;) All these advanced math techniques don't mean squat if you aren't able to apply them anywhere but your computer.

So while, you guys can fight over the merits of such gains in threads that run 15+pages, I want to go on record to say to anyone reading these threads for the first time, or the umpteenth time, whatever your level and intent, from recreational player to supplementing your lifestyle/income, to full-time player, “you don't have to be Newton or Einstein to succeed. You can still do just fine learning the basics”. :)
Forsooth Brother Kewl, forsooth! zg
 
Last edited:
#23
Sucker said:
This post, with only a LITTLE bit of editing; could and SHOULD be made into a "sticky".
Go ahead and add your little bit of editing here.
Better yet, lets turn the KJ sticky post into a wiki and we can all hack at it! zg
 

QFIT

Well-Known Member
#25
psyduck said:
to have a "mathematics" thread based on personal beliefs and feelings without any data.
Psyduck, I gather by your avatar you have a fondness for water either strained through dried dead leaves or beans.
 
#27
I know of one

Richard Munchkin said:
What's really amazing is that people have been arguing this stuff for 30 years. I would be amazed if you could find one really successful player (7 figure lifetime winnings) that would attribute it to using a higher level count.
Richard, I have heard of you! Nice to see you here!

Now with that out of the way, I know one player quite well that fits that description and sides with having the most detailed count imaginable, attributing this to overall lifetime success.
 

QFIT

Well-Known Member
#28
zengrifter said:
"Let us not confuse the menu for the meal, nor the compass for the destination." --Alan Watts
He also said “Zen does not confuse spirituality with thinking about God while one is peeling potatoes. Zen spirituality is just to peel the potatoes.”
 

MangoJ

Well-Known Member
#29
I'm very happy to talk math here in this forum, because Blackjack is a rather simple game (from the math point of view). I learned a great deal, which may - or may not - be applicable to different games. My upcoming projects is to tap into poker - profound math skills can only be benefitial.

Of course people can say that the only thing that matters is what happens on the table. I won't disagree, but that is not the point in math discussions. How do you think counting was invented in the first place ?

If you don't care about math threads, don't read'em.
 

psyduck

Well-Known Member
#31
QFIT said:
Psyduck, I gather by your avatar you have a fondness for water either strained through dried dead leaves or beans.
LOL! Dried dead leaves or beans? I think I know what your mean in both ways.

I admire your wit, humor, and of course contributioin to understanding the game of BJ! And also add your humanity as reflected in a few of your non-BJ posts I read.
 
#32
QFIT said:
He also said “Zen does not confuse spirituality with thinking about God while one is peeling potatoes. Zen spirituality is just to peel the potatoes.”
YOU ARE IT, Brother QFIT! zg

Ps - Speaking of potatoes, was your software suite at least partially inspired by Ron Popeil ("It slices, it dices, it minces, and it chops!")?
 
Last edited:

Gamblor

Well-Known Member
#33
MangoJ said:
I'm very happy to talk math here in this forum, because Blackjack is a rather simple game (from the math point of view). I learned a great deal, which may - or may not - be applicable to different games. My upcoming projects is to tap into poker - profound math skills can only be benefitial.
.
Hey mango, just curious, how do you plan on attacking poker mathematically? I'm thinking from a strictly probability theory approach it might be a dead end? Not sure how you would account for bluffs, bluffers bluffing bluffers, knowing someone is a bluffer, profiling and bracketing players, etc., Are you thinking along the lines of game theory or decision theory? You tell me if calling $2 pre-flop bet at a 1-2 NL game with 67 offsuit is a good decision with 7 players calling and your on the button at a table full of loose players.

Also not sure if BJ is necessarily all that simple mathematically - with probability theory it is, but maybe there's novel ways to apply other fields of mathematics to BJ.
 

MangoJ

Well-Known Member
#35
Gamblor said:
Hey mango, just curious, how do you plan on attacking poker mathematically? I'm thinking from a strictly probability theory approach it might be a dead end? Not sure how you would account for bluffs, bluffers bluffing bluffers, knowing someone is a bluffer, profiling and bracketing players, etc., Are you thinking along the lines of game theory or decision theory? You tell me if calling $2 pre-flop bet at a 1-2 NL game with 67 offsuit is a good decision with 7 players calling and your on the button at a table full of loose players.
I don't know much about poker maths, but I'm certainly not naive. I would need to break it down: If you would know the hole cards of your opponent, you will beat him to death in the long run - just by playing the odds. You would play purely by EV considerations. Ideally you would seize your bets in a way that it doesn't matter what option your opponents will choose, for each decision you would want the same EV. Let's call this basic strategy. The point is, this "basic" strategy is more complex than a naive approach (bet when in advantage, fold when in disadvantage) as it immediately gives you the correct bet size: Even against a player (which in turn would know my hole cards) my play would be optimal, meaning that I will either beat the player (or break even).

Of course this is not poker, you don't know the other players hand, but that is a first step a poker player (or a bot) should master. The beauty is, this step is fully deterministic, and can in principle be solved like any other blackjack strategy problem (not saying that it can be solved by "counting").

Second step: Since we don't know the opponents hand, we need to apply the "basic strategy" to any probabilistic distribution of opponents hands. So we can play optimal, as long as we are given an idea of how the opponents hand might look like. Our performance would then depend on how well we can estimate the opponent - but once we get an estimate we play optimal.

vvv HERE BE DRAGONS vvv

Third step - and that is the most difficult - estimating opponents hand distribution, based on the play of this hand, and past performance.

Fourth step - disguise own play. Make cheap cover moves, or establish a certain image.

Fifth step - ...

...

Last step - randomly switch between marginal strategy models to make you untrackable.


To your original "question", as I said, I don't know much about poker. But simply calling seems to me the worst of all options. I would probably raise, either most of them will fold (which then seems as a good investment into the pot). If instead most of them call, they will call anything on the flop if I hit.
 

Coyote

Well-Known Member
#36
This site has the best of both worlds for me! One being the Mathletes who go over the game in such detail that it is hard for me to comprehend, but I do appreciate their minds. The other is where some of us have a belief that those who have gone before us have figured this stuff out and all we need to do is follow their path to get to the treasure at the end. Either way, I am having a great time!

Maybe it is good to have all these conversations of theory and math in order to keep the lazy APs at bay.

Best Regards to you all,
Coyote
 
#37
There comes a point when the math behind the game is one of the least important factors in success, especially as you diversify and realize that counting cards isn't the final word in advantage play. I don't consider myself to be especially-gifted in the arena of mathematics (despite my 100% score on the HSCT math exam, which despite me shamelessly bragging about is likely unimpressive at best), and I feel like the only reason I've experienced any sort of success that I have is due to my willingness to work hard and my ability to interact comfortably with strangers. The latter is, in my opinion, one of the most important traits of the successful advantage player.

That said, I wouldn't have the opportunities I have had without the work of those that have obsessed over the math and perfected the strategies. If it weren't for James Grosjean or Don Schlesinger (worlds collide!), I would have no hope as a successful advantage player. I feel lazy profiting from the hard work of others more skilled than myself, but I hope to contribute on my own someday, most likely in a way that does not involve an in-depth understanding of esoteric mathematics, as I have little of that.

So I think there is a place for discussing the finer points of advantage play, but I agree that the application is limited. While I'll avoid the derisive "AT" tag, I'm glad that there are players toiling away with small refinements to the greater game, as I wouldn't have gotten very far in the first place without them.
 

iCountNTrack

Well-Known Member
#38
kewljason said:
It seems to me, the advanced mathematics angle has really been pushed on this site lately. Maybe it is my lack of a mathematics degree and/or background that prevents me from not appreciating this angle as much as others.

Someone visiting this site for the first time and reading through some of these threads, will conclude that to succeed in card counting today, you must play an advanced level 2, 3 or higher count, with multiple side counts, and hundreds of strategy change indices for both the primary count and side counts and that is just not so. I actually believe that in today's game the mathematics is less important than it used to be. :eek: The math guys want to run sims showing that a multiple level count will improve your results by .xx percent or two side counts will with hundreds of play variations is worth an additional .xx percent. Some of these guys, I think the majority of their blackjack experience takes place on their computer. They think Qfit's products or even their own programs are their blackjack career instead of a tool to aide in their actual play. :sad:

I am not disputing that these advanced card counting ideas don't improve performance. I do believe the value of this 'gain' is overstated, the error rate understated and that it just doesn't translate fully in actual play. I am questioning the value for the majority of players in the real world of play. In the real world there is so much education that must take place on the ground, rather than the computer screen that seems to be overlooked. Things that can result in far greater improvement to your results than higher level counts and side counts. Sniffing out and taking advantage of better games and opportunities. Learning what is and isn't tolerated at different places and tayloring your play to these guidelines and levels. Even learning these tolerance levels for different pit personnel and co-ordinating with their schedules. ;) All these advanced math techniques don't mean squat if you aren't able to apply them anywhere but your computer.

So while, you guys can fight over the merits of such gains in threads that run 15+pages, I want to go on record to say to anyone reading these threads for the first time, or the umpteenth time, whatever your level and intent, from recreational player to supplementing your lifestyle/income, to full-time player, “you don't have to be Newton or Einstein to succeed. You can still do just fine learning the basics”. :)
I am afraid I have to disagree with you on this one KJ. All the Mathematical/theoretical/computational discussions have been and are always confined to the "theory and math" section. Moreover, if you actually "count" the number of "practical" questions about card counting or casino play, you will find that the latter wins easily. Besides all that and more importantly, the stickies and posts by site admin, mods and prominent board members stress out the fact that you do not need to be a mathematical genius to count cards, and advocate the use of simpler counting systems.

Since this is a sticky, i will reiterate the " accepted general wisdom" about card counting systems with a few touches of my own personal taste.

One needs to realize that "in general most counting systems" are good systems, but some systems are better for particular rules, number of decks, penetration etc. There is no universal best counting system if you factor in all these including ease of use.
Let's discuss point by point the topics related to card counting theory.

Count Level
While in the 70's and 80's complicated Level 4 systems were popular, much simpler systems are used these days. Besides a few great individuals(raise? :)) who use halves (a level 3) system, it is widely accepted that Level 3 and higher are not needed. While there is no disagreement on the latter point, the debate over choosing a level 1 system over a level 2 system is an ongoing topic. My personal opinion about the matter is that it really all depends on several factors. If you are playing as a team, than a level 1 system is much better mainly because it allows for better playing quality control. If you play solo and your sole advantage play is card counting, with enough practice and discipline you might be better or with the 10-15% extra gain in return. This is where a good simulator that also includes possible player errors come into work. You can run the simulator to see if the errors are reducing your return to a point that is not worth it for you to use a higher level count.

How many indexes?
If only i got a dollar every time this question was asked... The obvious answer "as many as you can" however that is only a partial answer, because again it all depends on your game strategy mostly your playing strategy. If you are backcounting the I18 and Fab4 are more than enough, if you are playing through positive and negative shoes, the more indexes you add the better off you are. In fact if you compare a play-all game for I18 and full indexes, you will get ~ +0.2% increase in ev.

Side counts
The general wisdom is to reserve side-counts to pitch games (unless you are side-counting for a bonus side game). Also unless you are playing very deeply dealt single deck games(time machine back to the 70s), there is really little gain in side-counting cards and using multi-parameter tables to increase playing efficiency. My personal opinion, if you only play pitch games (like fine 2D games) , Hi-Opt II with ace-side count for betting is a really strong combination. And it is very doable with enough practice.
 
Status
Not open for further replies.
Top