John Martin

laughing through grad school
(academic stuff) (hints of life beyond
school and work)
(Flying Moose videos, photos, stories, etc.) (observations)

A study of Knowledge and Learning in Pool Playing


  1. Claims (Introduction)
  2. Context (Theory and Process)
  3. Confirmation (Discussion)
  4. Criticisms
  5. Conclusion and Consequences

Isolate the key features of the problem

First Attempt

I began this study by looking only at novices and experts, trying to ignore the stages of expertise between the two extremes, thinking that by disregarding the subtle changes that may occur between adjacent stages, clear differences would be easy to see. They were, but the explanations for the differences were not. So the answer to the question: Is what an expert is doing the same as what a novice is doing, but better/faster? was an easy "No", and a paper showing that seemed too trivial a task for a semester project. But the task of coming up with an alternative explanation, using the theories encountered in our readings, did not lend itself well to this dichotomous novice/expert view.

Up to this point, my novices had rated themselves as 0,1, and 2 on a 1-10 scale with 0 being someone who had never played the game, and 10 being a world-class player. The experts in the study had rated themselves modestly as 6s, or 7s[1] . I felt that in order to explain what I observed, I needed to broaden the scope of my study to include members from more of Dreyfus's levels of development,

At this point also, it became clear that experience, skill and theory significantly inform each other.
The experts complained about the constraints of the problems I presented, that they had to choose shots based on a small photograph instead of being able to look at and walk around a full-sized 3-dimensional table.

Second Attempt

In order to address the shortcomings of the first round of data gathering, I decided to give the same problems to four pool players who rated themselves between 3-6

My theory

I believe that Merlin Donald's Evolutionary Theory of cognitive change occurs on more levels than the macro level of cultural evolution. These patterns repeat themselves in the five stages identified by Dreyfus and Dreyfus from novice to expert.

The novices know that they should put the balls in the pocket, and they indicate their ideas for how that should be done. They typically only see one or two possibilities, and choose to attempt the easiest, clearest shots. They generally will not attempt a shot that they think they cannot make unless they have no other shot, at which point they have often already mentally decided that they cannot make it, so they lack the ability to shoot with confidence. They are afraid to take risks, and stick with what they know, and so often cannot see many possibilities. Some choose to not progress in pool, but to avoid the game, and so end their development here.

As pool players progress through the advanced beginner stage, the number of "forced" shots that they've experienced accumulates, and they are able to begin to build a record of those experiences. This record includes the inexact layout of the shots, how they shot them in the past, and what happened. Each time they encounter a similar shot, they are able, consciously or subconsciously, to access those records and adjust their present shot to be similar to what they did in past successes, or different than past failures. They are still focused on one ball at a time at this point, and are generally not concerned with trying to control where the cue ball will end up for the next shot.

As development continues, motivated learners begin to pay more attention to the shots of others and mentally take note of what works for others in situations. They try to mimic the shots of others, and ask those better than them for help in seeing and choosing shots. With the bolstering of confidence from an experienced guide, they take more advanced shots, and are more likely to review both missed and made shots in order to learn from them. They may look at some missed shots as "successful" if the ball came close to doing what they wanted it to do. They may also begin to connect that if they can control where the cue ball ends up, their next shot may be easier.

After this, players may have enough confidence to look for and take "risky" shots on their own. They may refer to third party instruction via books, videos, of formal courses. They may spend time on rented tables practicing their breaks or setting up specific shots that they want to learn. It's at this point that they will, on the shots that they are confident of making, begin to focus on controlling the cue ball and setting up concurrent shots. They generally are pretty accurate with straight shots of less than the pool table's width, but still have difficulty with bank shots and full table length shots.

Somewhere around Dreyfus's "Proficient" level, balls begin to appear to the players as clusters or groups to be dealt with, rather than as single balls. Players will actively design their shots for multiple purposes: first to make the target ball, and secondly to control the path of cue ball to break up other balls that are clustered too tightly together, are too close to the rail to afford an easy shot, or are in the way of other shots they wish to make. This use of other balls is a significant leap in method, and the venture into their use follows a pattern similar to the player's development of approach to the single ball.

The theory I'm suggesting melds Dreyfus and Dreyfus together with Merlin Donald in a way suggested by this graph:


In this paper, I consider the differences between expert and novice thinking in the realm of a small study of 8-ball pool theory. Specifically, this paper will:


Six adults agreed to participate in this study, to answer background questions, and consider pool theory problems, and to indicate their solutions while having their voices and gestures videotaped.

Two 8x10 color digital photographs of an 8-ball pool game immediately after break, with no balls down were shown to the participants, one photo at a time. Participants were asked to answer the following questions verbally and with gestures as they looked over each photograph.

  1. If you were playing 8-ball (bar pool) and it was yourshot on an open table, what would you shoot at? (which ball to which pocket)
  2. Why would you take that particular shot?
  3. Where would you leave the cue ball?
  4. What would your next shots be?
  5. If you had to shoot at the other set of balls(stripes or solids), explain your shot sequence?

Participants were videotaped over their shoulder as they verbally and visually (gesture) discussed each problem.


It seems that the nature of thinking changes between experts and novices in the realm of pool theory. It's not a case of experts doing "more and faster" of the same thinking as novices. The following behavior and thoughts were observed:

Novice Expert theories (confirm or counter)


The problems were presented on as large color photographs of a open pool table, just after a break, and therefore out of context to how pool players typically address such problems (2-D vs. 3-D; unable to walk around the table to get a sense of space, and much smaller, so unable to see as accurately as in a "live" game. As Dreyfus and Dreyfus suggest in "Five levels of expertise", due to these constraints, experts fell back to pre-intuition levels of seeing and decision-making, so the intuitive aspects of skill acquisition were not observed. (Lakoff: decision-making is grounded in body and physical context).

The domain of pool may be socio-culturally male-biased. Gender differences are not addressed or accounted for in this study. A larger, more gender-balanced study might offer different evidence.


The knowledge structures and Experts


None solid yet.


[1] This was interesting to note. I would rate them on a linear scale as 8s or 9s, but they seemed to rate themselves on a tiered pyramid-type scale, where for example, there can be only one 10, ten 9s, a hundred 8s, a thousand 7s, etc.