Cognitive Domains of Blackjack: Implications for Cover

As a specialist in brain-behavior relationships, and as a newcomer to blackjack, I've become fascinated by the cognitive task involved in playing the game properly. Perhaps there is some general interest as well; I noticed a number of posts addressing the issue of IQ, with debates going back and forth between the idea that you have to be a genius to play blackjack, and the idea that anyone of average intelligence can play if they have a certain type of personality that seems quite difficult to measure. A similar question arises when we take time to consider the inner workings of the "ploppy" -- for example, when we discuss ways to convince recreational players and gamblers to boycott 6:5 blackjack and similar games. I don't intend to diverge into psychology; this will be strictly nuts and bolts.

First, some definitions.

Intelligence: Physiologically speaking, intelligence refers to information holding capacity, speed of processing, and the ability to cross-link concepts (e.g., forming abstractions).

Cognition: skill at specific tasks, such as learning the definition of a word, adding two numbers together, or rotating a geometric figure in your head.

Working memory: the ability to hold data in memory while performing operations upon it, or while attending to other tasks. Long term memory is context dependent; working memories are not. Working memory is a function of the front part of the brain (the frontal lobes).

Intention: the ability to inhibit short-term impulses in favor of long term goals; also, the ability to ignore novel stimuli. Intention is also a frontal lobe function.

I see four primary cognitive tasks associated with playing blackjack:
1. Determining an appropriate bankroll.
2. Learning a system (memorizing tables, learning to count).
3. Keeping count.
4. Determining a bet size.

Clearly, these four tasks tap into a number of different cognitive domains. The only one that is directly dependent upon intelligence is #2, learning. Thus, a certain amount of information carrying capacity is required. It has been suggested that an IQ that is between mean and +1 SD (i.e., between 100 and 115) is sufficient for this task. That sounds reasonable.

The ability to keep a running count is not really a function of intelligence; this is a classic example of working memory. In fact, this is a very difficult skill to measure with an IQ test; you can pull off a reasonable IQ score and still not be all that good at working memory. Also, working memory is separate from math skill, and it's also distinct from context dependent memory. So even if you have a photographic memory, that doesn't mean you will be a good counter.

I am undecided with regard to the bankroll calculation. It seems that bankroll is a critically important issue -- even if you can count perfectly, I can see how insufficient bankroll will sink you -- and I can also see that it would require considerable cognitive skill to master. The question is, to what extent are general guidelines sufficient? With regard to learning a counting system, I can see that it's not really critical to understand why removing a five from the deck gives the player an x% advantage; all you really have to know is "add one." On the other hand, I suspect sometimes that a deeper understanding of bankroll requirements is necessary, and that rules of thumb won't necessarily carry the day.

Personally, I find the dynamics of bet sizing to be the most interesting, and the least explainable by IQ or cognition. Choosing a bet is done on they fly, and rules of thumb are hazardous at best, since this task involves consideration of advantage and bankroll size, as well as host factors like risk tolerance. It is here that players have the most to lose from the common decision-making errors people face in daily life: what are the assumptions I'm going in with? And to what extent does my emotional state affect my thinking?

The act of placing a bet is an emotionally-laden experience. What first comes to mind for most people is the negative emotion associated with the risk of losing money. However, placing a bet can be a very stimulating experience, based on one of two theories:
1. The strength of intermittent reinforcement paradigms. We know that if you reinforce a rat every times he pulls a lever, he will pull the lever a lot. But if you only reward him intermittently --say, every tenth or twentieth time, he will REALLY pull the lever a lot. You've seen people play slots; you know this is true.
2. The dopamine hypothesis. Theoretically, there is a group of people out there who have a deficiency of some sort in the D2 receptor, a molecule in the reward center of the brain (the nucleus accumbens). These people, in an effort to get a sufficient dopamine "fix," might artificially stimulate the reward center with drugs that increase the amount of dopamine in the brain, such as nicotine. They might also participate in certain risk-taking activities, which also apparently tickle those D2 receptors. It is thought that gambling stimulates these receptors.

One way or the other, I would proffer that the act of placing a bet stimulates very fundamental parts of the brain, in fact areas that are similar to, and probably connected with the sex drive for example. And, as with the sex drive, considerable focus is required to make good decisions.

Thus, the act of choosing a bet depends on two domains:
1. Cognition, namely calculating the true count and recalling a bet multiple.
2. Intention, namely making a conscious decision to make the right bet in spite of emotional influences one way or the other.

Both seem to be very important.

The cognitive process, which is a left hemisphere function, requires considerable understanding of the interplay between advantage, bankroll, and variance. Near as I can tell, there are no truly universal rules of thumb in this arena, judging from the number of posts about bet sizing and ramps, and the detail and complexity of the answers.

The intentional process, which is a frontal lobe function, may be even more important. Many posters, for example, have mentioned the importance of "guts," which can be defined for example as the ability to push a big bet out when the count calls for it, even if you are afraid of losing. It might also refer to the willingness to grind out a positive shoe, even if you've lost ten hands in a row.

People are often surprised when they find out that there is a part of the brain that's specifically designed to give you that kind of guts, but there it is. And it works pretty well for some people. The weakness of the frontal lobe is reflected in the games people play with the decision making process. Most of these "games" have to do with a fundamental property of neural networks: a tendency to make generalizations.

For example, consider these classic errors in decision making:

1. Anchoring. Any time people have to make an estimate, they are heavily influenced by any starting value that anchors the debate. You see this in negotiations; for example, if your asking price for a piece of property is ridiculously high, you will probably wind up with a better initial offer. In blackjack, anchoring bias might result, for example, in a tendency to bet more if everybody at the table is betting more. It might cause a new player to question his strategy of playing $5 tables, if all he can find at his hotel are $10 and $15 tables.

2. Representativeness heuristics. We sometimes tend to make decisions based on preconceived notions, templates, or rules of thumb that we've acquired over the years. This is a natural tendency, and to a certain extent it is probably impossible to avoid this altogether; nevertheless, you know what they say about assumptions. One classic example of bias based on representativeness heuristics is a failure to appreciate regression toward the mean, as in getting overconfident if you win a bunch of money, and then becoming annoyed if you go on to lose several sessions in a row. This also leads, for example, to the idea that if you've lost ten hands in a row, you're bound to win the next, because it would be real unusual to lose eleven in a row. The error is in applying a template (a prolonged losing streak) to a situation that does not match the template (playing a single hand of blackjack).

3. Framing. Framing bias refers to a tendency to change your answer if the question is rephrased. An example would be the dilemma a basic strategy player faces when he has a natural and the dealer is showing an ace. Basic strategy players might know intellectually that the expectation is greater if they don't take insurance. But they are easily influenced at the table by people urging them to take the sure thing (even money).

More than anything else, the essential function of the frontal lobe is performing the calculus of expected value. I once saw a bumper sticker that defined stress as the feeling that arises when you choose not to beat the living crap out of some guy who desperately needs it. Thus, the calculus is as follows: will I be happier if I beat this guy up now, or will I be happier if I don't have to go to jail later? A person with a weak frontal lobe has difficulty with this sort of calculation and tends to act impulsively, without regard for future consequences. This is the same thought process as knowing that it is better to have a 30% chance of winning $100 than it is to get a guaranteed payment of $25. (The kicker is, while most people are able to control their violent tendencies, very few would be willing to pass up the $25 sure thing.)

Thus, on the matter of bet sizing, there are really three challenges with regard to intentional or frontal lobe function:
1. Overcoming the emotional pull of quitting or gambling.
2. Overcoming certain mindsets that can lead to various heuristic errors.
3. Accurately performing expected value calculus.

Thus, putting this all together, we find that there are three domains that are used in blackjack:
1. Intelligence (information carrying capacity)
2. Cognition (doing addition, understanding statistics)
3. Frontal lobe function (emotional control and working memory)

Of these, intelligence exhibits a threshold effect; in other words, you have to be smart enough to get the basic concepts but beyond that intelligence is not going to carry the day.

Cognitive ability, with regard to math skills, is probably very important with regard to bet sizing and bankroll calculation.

Arguably, the most important domain is frontal lobe function, which subserves working memory (keeping the count while doing other things), screening out distractions, and performing the emotional calculus that balances fear and excitement against EV.

I know some really bright ploppies. I don't happen to know any with superior math skills, but I know some with extraordinary intelligence, people who clearly know how to count, but for some reason just...don't! Probably the one thing that uniformly separates ploppies from AP's is impulsiveness.

Interestingly, most people know their intelligence and their cognitive skills, because these domains are frequently tested. People seldom get their frontal lobes tested, even though tests are available.

I suppose if there's any practical value of this discussion, it will have something to with understanding the difference between AP's and ploppies. This could lead to two benefits. The first would be better cover. Consider for a moment that there are two kinds of pit personnel. The ones who learned their skill from a book could easily be thrown off by a few well-placed cover plays. The ones who learned their trade through years of experience probably depend more on their intuition to find APs. To beat that kind of pit person, you have to look, act, talk, walk, and smell like a ploppy. (In other words, we have to appeal to their representative heuristic). So we may decide, after close analysis, that, among other things, we should appear to bet impulsively and smoke a lot of cigarettes if we want to look like a ploppy

The other benefit might be a greater appreciation of the ploppy’s mindset when we are preaching to them about CSM’s and 6:5 blackjack. To wit: we will never win them over with logic.

I hope this has been of some interest. I know this is an unusual topic for BJ21.com, but I've been impressed with the number of really bright people who visit this board, and their generosity in terms of sharing their expertise. I hope that I will be able to give a little back.